This post is a plea for support of a project to write a comprehensive story about Franklin Antonio. Please respond if you have material about Franklin or would like to support my sense t…

Read more

The Strategic Logic Of The Ai Arms Race

170KB ∙ PDF file

Download

As he has at strategic points in the past, Charles gives the strategic framework and the implications of AI in warfare and international relations. Read also O'Brien

Download

Read more

Entering MIT in 1961, I saw the of the first generation of AI giants Norbert Weiner and Marvin Minsky1 and others walking the halls. Weiner taught only grad students, and was famously absent-minded, reputedly having asked someone which way he was walking when they met so he could determine whether he had had lunch yet or not. Minsky taught my freshman calculus class at 9:00 M-W-F, and he was terrible — he lectured to the board, never turning to face the class. Most of us drifted over to Gian-Carlo Rota’s class at 10:00 after a few weeks, but Pat Winston stayed with him to the bitter end: Minsky supervised Winston’s PhD, and Pat succeeded Minsky as the head of MIT’s Computer Science and Artificial Intelligence Lab (CSAIL)”, and was the leader of the next generation of AI gurus at MIT.2

So, since I have been promised the “Coming AI Revolution” for more than six decades, I may be forgiven for being slow to join when the current AI revolution was announced, but I am on board now and here is my take on where we are today.

Last edited on October 22nd, 2026

Introduction

After reading Stuart Russel’s good book Human Compatible (2019), I settled on Max Tegmark’s 2017 book Life 3.0: Being Human in the Age of Artificial Intelligence3 as the best book to suggest to learn something about AI. Tegmark is a Professor of Physics at MIT, and taught a lot of this generation’s AI gurus. By the time he ascended to guru status, AI had turned from the work of Winston, who represented the symbolic or Good Old-Fashioned AI (GOFAI) period, and later Rodney Brooks, who represented the embodied or situated or behavior-based AI period, to the Large Language Models and Neural Nets operating at massive computational scale period.

The scale of the tools generated a lot of discussion about how they should be used. I in particular liked a Substack post by Juna Kollmeier which spoke in terms of utility functions.4

In the early 2010’s, Tegmark got grants of more than $10 million from Elon Musk to twice assemble more than 100 AI experts to discuss the risks of AI and ways to mitigate them, at conferences in Puerto Rico in January 2015 and a follow up in 2017. The group wrote strong position papers and got statements of support signed by over 6,000 recognized leaders. Life 3.0 documents the process supporting those two conferences, and gives an overview of AI. There are dozens of other books about AI available.5 While Tegmark remains a professor of Physics, he devotes much of his time to the Future of Life Institute he founded to deal with AI safety and governance.

The real difference between search as a tool, and AI as a tool, is that, whereas search involves key words, with results pretty much presentable on an X-Y grid, AI allows for much more complex associations.

In the same way that our brains can react to words, to numbers, to smells, to noises, to verbal and non-verbal cues and associate them with contests, memories, with objects, and with emotions, all in a complex way by virtue of the notoriously complex structure of nerves and connections in our cerebral cortex, AI uses chips designed to deal with notoriously complex mathematical relationships to associate visual images, endless series of numbers, data of all types with things like faces of people, of concepts like romance. (It was long ago called associative memory but that phrase seems not to be in style now) AI famously allows the inference of how one protein is constructed by comparison to another, analogous protein. It is understandable that some feats of current AI models appear almost magical.

My perhaps too simplified way of thinking about AI is as follows:

• The current AI a wildly more capable extension of search engines. With a search engine, we can start our projects searching for the basic ideas and nail down the vocabulary involved quickly. We can assemble facts as we proceed. It is a great boon to productivity, just like the HP12C improved on the slide rule. It is a lever and using it is not a crutch. With the search engine we we can retire our 3 X 5 cards just like the HP12C allowed us to retire our slide rule.

• An entry level AI model, too, is a lever, taking associations to a higher level, dealing with more domains and making associations that mere searches would never make. If, as a user, we take the time to understand the background or the associations that the AI model made, we are not letting the AI lull us into thinking that we understand when we really don’t. It, too, is a lever not a crutch and it can extend our basic skills and save time and energy. Agentic AIs with harnesses can release us from periodic tasks and unnecessary repetitive coding, etc. While some skills may get rusty in the process they are not necessarily losses we have to morn.

• With more advanced AI, the tradeoffs get trickier. Search engines come up short when dealing with conceptual ideas, but AI tools thrive on them (or at least appear to). Some advanced AI even appear to have cognition. If we come to depend on AI for developing our concepts and fail to develop that skill for ourselves, or if we fail to develop our own cognitive skills because we default to the supposed cognition of our AI tools, we are allowing our lever to become a crutch. If we do that, and if some of the skills which AI may make rusty are the very ones which we should be most protective of, we begin to lose We have to protect our cognitive skills, our sense of values, insight, and have to be able to identify what is not known.

But I do not fear them obsoleting me. They might easily get out of control and there are very, very real risks. And some form of regulation is needed, just like some form of regulation is needed of the magic-appearing CRISPR-based tools.

No matter how magical the feats of today’s AI tools are, and while the basic structure of the current AI mimics in the basic structure of the human brain — they both have sensors taking in signals from the world in which they operate; a central processor acting like a CPU, and a complex neural network structure allowing association much more complex than and X » Y relationships — the AIs have neural networks on the order of 4-10 levels while the brain’s neural networks are on the order of hundreds of levels. The effect of higher orders of complexity is combinatorial, not linear.

I do not find discussions of AGI and the dangers of AI taking over in the sense of becoming like HAL 9000’s compelling. I am comfortable with discussions of limits, and of risks to to cognitive development, etc., of users very compelling. Read on.

Use Cases

Protein Folding

Somewhat by accident I got exposed to the problem of protein folding more than thirty years ago because I knew David Shaw through the NY Academy of Sciences.6 David made billions of dollars as one of the first quantitative traders. He has exposed to molecular biology at his undergraduate school, UC San Diego (he graduated in 1975, the undergraduate school opened only in 1964). With his newfound wealth he decided to return to that world and to use his skill at math and comp sci to build proteins after he lost his mother, father, and sister to cancer.7 He failed after spending a lot of time and money, so far.

While today’s AI juggernaut seems to have emerged over the last four or five years, it started gathering stream at least a decade ago. There are two movies, AlphaGo and The Thinking Game, which describe why Google paid $650 million in January, 2014 to buy the AI model company DeepSeek depicted in the movie.

I describe the protein folding story and DeepSeek like this: Proteins are the factories that create life. They have evolved over billions of years. They are incredibly complex, with millions of amino acid molecules arranged in a precise genetic sequences in physical structures to accomplish their tasks. Using electron microscopy and other tools research scientists painstakingly built up an understanding of the structure of something like 200,000 proteins, some rather complete and some incomplete

And there are ways to rank or score the incomplete structures, and structures could be judged and improved upon. There are standard sequence databases and ways to share known structures, or guesses at structures, and competitions to rank what is being accomplished. This process received a lot of attention over the years. Today

Essentially, using DeepSeek’s protein folding model Alphafold, the first known proteins were used, over a very small number of years, to create first dozens and then hundreds and eventually all 200 million+ proteins. Alphafold is now in its third iteration. It is being use for drug research, among others. That is, the advanced AI solved the protein folding problem, to a great degree. I see hints, however, that the answers given may be imprecise enough that, at some level of detail in the process of drug creation, etc., tools like what David Shaw built may yet find use.

AI and Creating Something New

The question as to whether AI is creating something new is perplexing.

• If advanced AI programming tools expose numerous vulnerabilities that have lurked in system after system, is that finding something new or is it just a very efficient checklist?

• Seeing the way an AI tool hacks a simple pong-like game may appear to be insidiously clever, it is just a process not unlike evolution at work where given enough time and enough chance and a way to judge the results, computers can come up with solutions that appear inspired but the process by which they were achieved is explainable.

• If AlphaFold could use analogs to the first 200,000 proteins to suggest the structure of 200 million, we can see how the process was carried out.

• If an AI model can take all the know math problem solutions and solve some that had not yet been solved, we can understand the process and it can be argued that the AI was just applying what was already known to the new problem at hand, more effectively. The pong game had the feature that attacking from behind was an effective strategy but no one had tried it before to see that it existed. There is no blind watchmaker at work.

Many have pointed out that truth comes from asking good questions. I would change that and say truth comes from asking NEW questions. While AI might appear to be expanding the boundaries of knowledge, it starts with what is known by using answers that were already known. By definition, knowledge expands at its boundaries. To the extent that AI is an extremely effective aggregator of what is already known, it will help define the boundary but it will not extend it. It is only by asking questions that have not been asked before and answering them that the boundary of knowledge is expanded.

Juna Kollmeier likens it to the AI tool having a lot of dials which be set to establish a utility function. In her discussion she contrasts the ‘positive utilities’ of having a patient tutor, skilled in all aspects of a discipline with the ‘negative utilities (dis-utilities) of possibly allowing the student to sidestep the ‘struggle’ by which learning takes place, and ominously “unchecked opportunity for exploitation.” 8

AI in Education

I spent three days at MIT in mid-September 2026. As you would expect everyone talked about the effect of AI on teaching, learning, and testing. A 40-page report had been issued mid-August grappling with the subject.9 One of its authors, Prof. Eric So, the Sloan Distinguished Professor of Global Economics and Behavioral Sciences, described his forthcoming book, The Collision: What AI Does To Us.10

Prof. So’s “collision” is what happens when the value (‘utility’ in Kollmeier’s equation) of what AI does FOR students as they solve problems speedily and cheaply comes into contact with the cost (‘dis-utility’) of what AI does TO students as they collect and organize ideas using AI. He cites numerous studies of learning and cognition to show that using AI to solve problems makes users overconfident about their abilities because of the ease with which they can answer apparently subtle problems with AI. Getting answers without having to overcome the “friction” of solving the problem ourselves diminishes all users.

Professor So urges that we consider the threat of AI in teaching and learning from the point of view of what AI does to us. He wants his colleagues and students to tread the line separating the power of the tool and its capacity to interfere with the process of learning. He uses what is known about behavior, psychology, and cognition. The “collision” of his book’s title is the threat that that having AI tools at our disposal will interfere with our skill development if we allow it to become a crutch

Prof. So proposed a set of “So’s Rules”:

• Value the Struggle. Stated as: The student has to experience the friction of learning, has to make it past the challenges, has to experience the failures. No one should outsource what has to be overcome to learn, which is what happens when we depend on AI. It is central to the book’s worry about what AI does to us.

• Value Who You Are Without AI. He quoted many studies demonstrating that AI gives the subject a false sense of understanding without the friction of actually learning; that it depleted rather than developed the learner.

• Reinvest Your Surplus. Use the time saved or the energy freed up to be sure you understand the logic and the principals behind the answer that AI gave.

• Make AI Your Cognitive Trainer. The real value of the tools is to design them carefully and specifically to take advantage of the natural cognitive process which people express

International Affairs

Must read: Charles Ferguson piece in Project Syndicate, The Strategic Logic of the AI Arms Race, available HERE. Charles shared this pdf

For reports from the front lines in the Ukraine, literally, see Phillips O’Brien

Responding to Existential Risks

Dario Amodei Substack Post

All companies go through “Life Cycles of Startup, Growth, Plateau, Decline, and Death” which I describe as:

• Startup: Managing by Your Checkbook©;

• Growth: Managing by Your Financial Statements;

• Plateau: Kicking the Can Down the Road;

• Decline: Managing by Cash and Accounts Payables©; and

• Death: Becoming a Zombie or Closing Down.

New financial control tools, disciplines, and accounting acumen are required at each stage. Company success is absolutely determined by how well its financial management and control processes are adjusted at each transition from one stage to the next along the path of growth. This post, in book form, discusses how to manage financial control at each stage as a company grows, plateaus, or begins to die. It is what I learned over five decades of experiences in companies at all stages of growth.

This post is targeted at a company CFO or CEO. It is my attempt to describe how I applied my training to solve the problems and challenges I faced through 50 years of financial control and technology management in companies that were large and small, and in startups and in mature and in declining companies. It can be useful for anyone trying to learn how to create, manage, understand, and use the numbers describing the financial health of their activity. Many of the insights in this post have made their way into my post Mantras.

The Balance Sheet and Income Statements are the usual basic financial reports used by management. The specific line items of Assets and Liabilities, Revenues and Expenses are as defined by the CFO and accountants at the time of use. They are all changed over time and the details matter. This post talks to the way these differences come about.

The expression “The Income Statement and the Balance Sheet are constantly Talking to Each Other” often applies. It means that financial the flows of cash, revenue, and expense go through the Income Statement (P&L) and are used to build or deplete the stocks of assets and liabilities and net worth on the Balance Sheet. The conversation is carried out using the language of double-entry bookkeeping, where every dollar is booked as an asset, liability, debit, or credit. Changes in one causes changes in the other. This double-entry bookkeeping process, where the Income Statement line items and the Balance Sheet line items “talk to each other,” keeps the books of the organization in balance and tells the story of its strengths and weaknesses, success and failures. It is the way an operation or process is “kept in accounting control.”

The job of financial control is to manage the conversations as the numbers talk and assure that they tell the story of how the company works. The company’s Finance Manual is the private repository for all the details used to run its Financial Management process. It has to be maintained current at all times. The “Notes to the Financial Statements” are the public description of certain aspects of the Finance Manual.

Keeping a company or operation under “accounting control” means exploiting the wonders of Double Entry Bookkeeping to manage all the stocks and flows. Something is “under accounting control” when every transaction in the process generates entries on both sides of the general ledger and when all affects on the Balance Sheet and the P&L are recorded for every transaction. One-sided transactions create a process that is not under accounting control.

This post is really about two companies I got heavily involved in. The first is Cloud Cover Media, a great success, which is described as it made its way for Stage One — “Managing by Your Checkbook©” to Stage Two — “Managing by Your Financial Statements.” The other — Torrance Optics — is a fictional company used as a device to describe a medium-sized company as it plateaus, drifts and finally fails. This fictional story is background for what is described in Stages Three — Plateau or “Kicking the Can Down the Road”; Stage Four — Zombie or “Managing by your Accounts Payables,©”; and Stage Five — Death or “Becoming a Zombie.”

My thirty years at Citibank and the three-year creation out of whole cloth of a company called Dryad Communications that I created with a fellow MIT alum and then sold at a substantial profit are also described in other chapters of my Memoirs. For those stories, see the following posts:

Story of 30 Years at Citibank

Story of Dryad Communications

It is worth spending a bit of time reading my Memoir chapters describing the good story of a startup that was sold to SiriusXM in very successful exit and a story of decline and death in the world of SBIRs which I fictionalize as Torrance Optics:

Story of Cloud Cover Media

Story of Torrance Optics

Discussion By Stage

Stage One — Startup: Managing by Your Checkbook©

In the early days a company is almost always “Managing by Its Checkbook^©.” Options for accounting and management control are few and decision making is simple and focused on cash.

• Management time is spent on getting the product out the door and husbanding cash.

• There are few abstraction; work is all about simple goals and successes; there is very little financial process.

• Accounting is simple and focused on day-to-day activity. There is little interest in the balance sheet other than cash and maybe equity, and the Cap Table is simple - where did the operating cash come from.

• The checking account is necessarily the accounting control system and the checkbook becomes the book of account. Sales are revenues are deposits are cash, funding is the liabilities/net worth and so bank statements become the company’s cap table. There is little cash management, and little flexibility for cash management.

• Financial Control has been ceded to the bank. Cash management is just a check-by-check decision to write it or not write it.

• Being “Under Accounting Control” means balancing your checkbook (with an eye to checks in the process of collection in both directions!).

• Management becomes very attuned to how customers pay and how vendors bill. (What is learned should be recorded as the beginnings of a Finance Manual.)

• Funding comes from credit cards, personal assets, or Friends & Family investors. Eventually customer revenue begins, which becomes cash, with the only strings or constraints being the need to provide the agreed service or deliver the promised product.

• The company should be paying attention to the details of payment, delivery, inventory management, attendant costs, etc. that will be needed later when setting up contracts. The more attention paid in the beginning the fewer will be the need for catch up later. Accounting controls and revenue recognition will be needed soon enough and the earlier it is understood the better.

Stage Two - Growth: Managing by Your Financial Statements;

When the company gets funding and starts to grow, management will be forced to start “Managing by Your Financial Reports.” The Income statement talks to the balance sheet as a normal course of activity. While, much of what is done for financial management depends on the dynamics and needs of the business, there are a few general rules.

• Revenue has to be booked in the time period it is earned.

• Expenses associated with that revenue should be booked in the same time period.

• Expenses which are incurred for future benefit can, and sometimes should, be capitalized to more closely align expense with when revenue will be earned.

• Expenses which are deferred for any reason but which eventually will be paid should be reserved for.

• The presentation of the P&L should show Operating Expenses distinct from non-recurring and start-up expenses.

The challenge is to apply these general rules to every important transaction and to set up the accounting record correctly and to describe it in the company’s Finance Manual so the company builds a financial management process that best presents the business and most accurately represents the profitability and growth of the company.

• It is a time of creativity and abstraction, aggregation, and expansion.

• Process is very important.

• Accounting Controls are added at more and more points of the activity, and the books and chart of accounts begin to drive the day-by-day activities. Managing By Your Checkbook is left behind.

• The financial management process is enhanced as the business expands and becomes more complex. There are more steps in the process and increasing scale generates operational and financial risks and the need to describe and manage revenue and expense better grows.

• Revenue and Expense Recognition (as an accounting discipline) becomes important so revenue and expense for product development and cost of goods sole are reported in the same period. Business modeling is not done in the accounting books but rather as standalone analytical exercises because it is tied intimately to the processes of the business activity more than of financial control. The results of RevRec is transferred to the financial systems at the end of daily, or monthly close processes.

• New assets and liabilities are created to manage revenue recognition and manage cash flow. The chart of accounts becomes more complicated and issues of presentation and company description are constantly in flux.

• Reserve Calculations and Collectability as operational issues and accounting problems become important.

• Reward programs for staff incentives and customer growth and retention programs are designed and implemented, with process, accounting, and profit presentation issues.

• The Finance Manual becomes an important document as a record of actions and repository of institutional memory.

• Software and other R&D can be capitalized to increase earnings by deferring expense.

• Reserves can be set up to reduce risk to profit, although at the expense of earnings, and to smooth earnings flows.

• Distinctions can be made in line items for assets, liabilities, revenue and expenses.

New sources and uses of funds can be tapped.

• Seasoned investors want to invest through preferred offerings, not as partners or common stock owners.

• VC’s, Private Equity impose their own requirements· Less-demanding Strategic Investments are made by vendors, partners, or customers

• Processes to manage capital accounts and prevent abuse by insiders become necessary. (Specifically, insiders will be tempted to convert liabilities into “cashless investments”, especially in times of stress, at the expense of other stakeholders.)

Service vendors and their control systems can be outsourced:

• HR to Professional Employer Organizations (PEOs) like Tri-Net or Insperity

• Customer service to Zendesk,

• Back office to Recurly

• MIS/accounting to Chart Mogul

New processes are required for asset management, cost control, cap table management, sales and for other bonus calculations. Profitability becomes driven as much by structure as by product choice:

• It is necessary to review contracts at each stage because higher sales volumes may open up volume discounts: volume pricing may have not been reviewed since “the beginning of time”. High-volume, long-term discounts, unavailable in the beginning, might be possible at substantial savings and with the possibility of fixing costs for periods of time.

• Revenue should be used to build the business, not pay off your investors.

  Investors in growth companies want management to increase the value of their investment, not pay it back. The company becomes more attractive to an Investment Bank or an acquirer by setting up reserves rather than by showing profitability or declaring dividends.

• Resist any attempt by anyone to convert a loan or deferred payment or any other liability to capital. New capital has to come from new cash, not from a converted or re-term liability.

• Decision making becomes complex, and organizational issues begin to compound.

• Attention turns to acquisitions and divestitures or to a liquidity event.

• Assets like Intellectual Property (patents, practices, institutional memory and knowledge), should be exploited or they will be ignored and wither.

Stage Three - Plateau: Kicking the Can Down the Road;

When the company stops growing and plateaus, either because of slowing sales or product failures or obsolescence, management starts “Kicking the can down the road.” Financial Management takes a back seat and decisions are made based on protecting jobs and withholding information.

• Managers hide the truth, first from vendors, then from employees, then from investors, and finally from each other. Effectiveness slides with each step.

• As managers thrash, the Income Statement becomes a record of shrinking Revenue and reduced Margins.

• As the organization shrinks, overhead eats up margin

• The Balance Sheet begins to get distorted:

  • Bonuses are accrued but are not paid.

  • Payables begin to build.

  • Accruals pile up.

  • Reserves disappear.

• Much activity is only “satisficing” -- doing the least amount necessary -- as management spends its energy shrinking the company’s footprint. Promises are made to employees, vendors, and Investors that are harder and harder to keep.

• Changes that should be made as early as possible include:

  • Downsize facilities

  • Narrow focus to what the company is good at.

  • Reduce staff.

  • Outsource or eliminate services that are below critical mass.

  • Sell underutilized assets like slow moving products or unexploited patents, no matter what the discount.

• If changes are made too late, they will be ineffective and the company will experience:

  • Increased overhead.

  • Loss of strategic assets including specialized skills and institutional memory.

  • Loss of markets.

• Cash dries up and the company returns to the limitations of “Managing by Your Checkbook”, but now with burdens of the past.

• Staff departs or are laid off, sales decline, and working capital is depleted.

Stage Four - Decline: Managing by Cash and Your Accounts Payables©

As working capital disappears and the inevitable decline sets in, company strengths withers, management stops paying its bills. It has been backed into a world in which they are “Managing by Accounts Payables^©”. Attention returns to financial management, but now rather than on how to build more revenue, it is how the balance sheet can be used to keep the company afloat. Since the company is not bankable, and no one wants to invest or buy assets, it relies on stretching it A/P.

• When cash dries up, selected vendors become a source of funds as bills go unpaid. To survive, the company first ignores them and then renegotiates contracts with them or sets up Payment Plans for them.

• Knowing this, smart vendors manage the relationship with an eye on their own Accounts Receivables.

  • The ones who know they are important to the company refuse to ship new goods unless A/R is brought current.

  • The smart ones have built in fees that they begin to charge, increasing ther cost.

  • The devious ones begin to pad or double up invoices.

• To investors and vendors, the company has become “the hostage they are unwilling to shoot” because they realize or are told by you that enforcing contractual terms would destroy the company. It becomes a game of “chicken”, and the company almost always wins, at least until the claims climb too high.

• The reality is that many vendors are at the company’s mercy, and it is possible to delay payments much longer than logic suggests.

• Advice about “Promise a plan and stick to it,” is correct, except the company does not even have to stick to the plan since the vendor generally has no practical recourse, anyway.

  • The conversations seldom veer toward “that’s just not good enough!”. They are usually, “I will tell my manager that you called and are starting to pay.” There is in practice almost no way to compel a customer behind on its bills to answer its phone.

  • Verbatim from an email: “Per our conversation, we will expect monthly payments of ~ $1,500 until the financial situation improves. If at any point the amount you are able to contribute changes, we simply ask that you keep us informed with clear communication via email. Lastly, please confirm when we can expect the first installment.”

Rules for setting vendor payments appropriate for a company fully into managing by its accounts payable:

• Pay based on immediate, not future, need. Pay based on legal exposures (of which there are few) or enforceable financial penalties (know which vendors have them). Pay older invoices in order, but never lose control of new vendor invoices: vendors protect themselves by double invoicing and adding fees since they assume new ones are being ignored. Smart vendors inflate their receivables to position for a settlement.

• Develop rules of thumb, such as:

  • Pay just enough to be interesting and to offer a hope of repayment but not enough to harm yourself. Once you start, or have throttled back as far as you can, persist.

  • Pay as a % of outstanding’s, so send smaller payments as each vendor is paid down

  • Pay as inverse % of historical total (the more they got in the past, the less you have to pay them now).

Stage Five - Death: Becoming a Zombie and Closing Down.

Without a major change, the company has “Become a Zombie” and soon should Close down. This happens when a company is stuck with old contracts, large and growing A/P. As accruals grow and vendors are held at bay, the following starts to happen and conditions deteriorate:

• Salaries stagnate, bonuses are skipped, and employees fall farther and farther behind.

• Contract workers lose incentive as they wait for payments.

• Intellectual Property including patents, practices, institutional memory and knowledge, should be, but seldom are, sold off or otherwise exploited.

• Intellectual Property withers because there no one to exploit it or answer inquiries about it.

• Reputation deteriorates until one vendor or a group of vendors takes your company to BK Court, or vendors band together to collectively gain power and turn the situation to their advantage.

Companies in this situation can or should respond as follows:

• Pay off sole proprietors, or, as a moral obligation, tell them you cannot pay.

• Pay only vendors who can demand cash on delivery. Stop paying all the rest.

• Alert all vendors who have allowed their A/R from you to grow (your a/P) that no payments will be forthcoming. This saves both of you the trouble of dealing with the issue and allows your debtors and vendors to deal with the tax consequences.

• Try to get value from Intellectual Property

• Very little can force you to be closed and then dissolved, but at some point, it is the right thing to do.

Final Acts.

When a company is finally closed, things to keep in mind include:

• Keep a checking account open in the company’s name no matter what, to be able to transact business and accept returns of deposit, settled A/R, etc. even after you are closed down.

• Licenses and accounts are closed by their issuers but you have to alert them to do so.

Notes, Conclusions, and Generalizations

• Financial management depends on the marvels of Double Entry Bookkeeping at each stage, but it appears at a different level of abstraction at each stage and the rules and the patterns that emerge are different and change with each stage.

• Often, the most important calculations for company revenue are not done in the general ledger but instead on standalone Excel spreadsheets. The most obvious example is Revenue Recognition, but there are myriad instances where expense calculations are integral to the business model as well. They have to be codified and managed as carefully as the financials.

• Accruals are not the same as reserves, and both become more and more important

• There are automatic checksums in accounting and spreadsheets, but there are no checksums in contracts, so the demands are different as the business becomes more complex.

The chessboard metaphor (q.v.) for spreadsheets and how patterns are recognized is very useful as a guide to understanding the changes as the focus turns from accounting to contracts and legal protections.

It is always helpful to keep the Notes to the Financial Statements or a full-fledged “Finance Manual” up-to- date so the rules of the road are available for all to read, as you go through the stages. True whether you are starting, running, or selling your company.

There are tremendous asymmetries of knowledge in every interaction with customers, suppliers, vendors, and employees. If you do not know where they are, you are on the wrong side. You have to learn about them and exploit them or protect yourself from them as you go through the stages of growth.

The parties’ self Interest will be the starting point for understanding all interactions.

This final job history story is about a set of companies that were spun out of the Physical Optics Corporation (POC), founded in 1985 by the husband-and-wife team of Drs. Joanna and Tomasz (Johann) Janssen, soon after they arrived in the United States from Poland. At least fourteen companies have been spun out from POC in the years since 1985 — prominently IntelliSense, Redondo Optics, InnoSense, and Intelligent Optical Systems (see list in the Appendix). POC got its start with an SBIR (Small Business Innovation Research) grant award from the NSF, and an important characteristic of the POC spinouts is that most relied, to a greater or lessor degree, on SBIRs.

SBIRs and POC

The SBIR/STTR program is huge. Its objective is to use innovative ideas that come from America’s small businesses (up to 500 employees) to provide solutions for problems faced by the large Federal agencies like DoD, NASA, DOE, NIH, etc. It channeled over $7 Billion to small businesses in FY2025 alone. It was begun by Congress in 1982 in its present form, and over its life has made grants totaling over $77 Billion. There is a lot of wringing of hands about the efficacy of the program, and this Congress’s appropriation process was tortuous, but it was finally approved and seems to continue year after year.

POC was not just any random SBIR-related company. It had won over $560 million of SBIRs in its own name through the end of 2024, and the four large local POC spinoffs have won another $410 million. That is almost $1 Billion! However, while POC was sold for $310 million in 2020, none of the spinouts have done nearly as well, and many did badly. Many have been closed and a few are still struggling. In general, it is okay for a company to use an SBIR as an initial source of funds or as an occasional source of non-dilutive speculative project support funds, but not as a steady source of revenue. The fact is that an SBIR is a cash drain in the short run, there are huge uncertainties, and there are always delays in the process. To succeed as an SBIR-dependent company, it has to either stick with a specific technology and get larger and larger grants as their product matures, move it to commercialization on its own, or develop the skill to license the ideas or sell them for commercialization by others. While IntelliSense seemed to be able to do this, the other POC spinoffs did not.

Like many SBIR-based companies, the more a POC spinoff relied on SBIRs, and the less skilled it was at commercialization, the worse the outcome. POC appeared to have succeeded by quickly moving beyond its SBIR roots, and by spinning off its SBIR-dependent losers. The SBIR program has many rules and regulations designed to make it difficult for a company to become an “SBIR mill,” defined as a company depending on getting one Phase I program after another, but there are ways to get around them, at least for a while. The fact is, SBIRs do not generate cash unless the company can move to product into more advanced development phases or move from the SBIR world into licensing or commercialization. The successful POC spinouts developed into commercial businesses. Those that stayed dependent on SBIRs failed.

So, while an SBIRs grant gave POC its early cash infusion, and it continued to rely on them for some revenue, it developed significant other sources of revenue. The spinout companies and their spinouts could not succeed when they depended on SBIRs.

The Fictional Torrance Optics

I am taking the literary license of combining the stories of all the companies and calling it one fictional company, “Torrance Optics.” All the people named are fictional, and any resemblance to any one POC-spinout company or any other company, and any relation to any real person, is accidental.

Over a 25-year period I served on the Board of, invested in, loaned money to, and helped run, one of the POC spinoffs, Intelligent Optical Systems (IOS). I was exposed to POC in great detail and to a few of the others in some detail. For this account, I gathered information about them all so I could write about them. For better or worse, most of my exposure to companies in decline were the POC-spinouts that failed, so they provide me a significant chunk of my learned experience in the case studies I draw from for this post. It is more likely that an SBIR-dependent company will end up in that situation so it is fertile ground.

Most of the lessons learned are not peculiar to SBIRs. They are are just what happens when a small company loses contact with its market, fails to find customers, or simply becomes starved for cash, for whatever reasons. Since I spent more time with SBIR-dependent companies in decline than others, my story documents in particular opportunities for bad behavior that were available to SBIR-dependent companies. There were indeed some features of the SBIR funding process that made bad behavior easier, or even likely, giving SBIR companies added ways to prolong their life as they slipped into decline. Many companies that plateau, decline, and die like Torrance Optics are forced to engage in similar practices.

Companies in decline quickly learn to make cash last as long as possible. The easiest way to do that is to slow or stop paying bills. That is, if you cannot raise money from banks or investors, let your accounts payable fund your company. When faced with insolvency, management learns how much separates practice from theory in business. When the pressures mount, managers — no matter how ethical they are — do things they never expected they would have do, and they get away with more than they ever thought they could get away with. These are the realities any company can and will exploit if survival depends on it. Realistically, the legal system is mostly unavailable to small businesses because lawyers and the use of the courts is too expensive.

The practices of Torrance Optics should be studied by both companies in decline who need survival lessons, and by vendors so they can be prepared for what their customers might do to them. Every transaction has risks for both supplier and customer. When the deal is struck, each side should be sure their risks are covered. While one should not assume all companies in the situations described here will engage in the practices I ascribe to Torrance Optics, it does happen and no one should be surprised by how easy it is for it to happen under the circumstances. Good people will do what they know are bad things to survive.

There are untold numbers of companies and managers who experience the pressures Torrance Optics experienced, and untold numbers of vendors, landlords, and employees who suffer through having customers or clients like Torrance Optics. Financial control professionals deserve to know what is possible so they can put in place processes to deal with it.

What is it about the SBIR process that facilitates or generates bad behavior?

The first problem is that SBIRs do not generate working capital. They are consumptive of cash because there are too many delays in the early part of the SBIR cycle. While SBIRs do generate non-dilutive funding once in operation, the business cannot be started with an SBIR so it has to have more in the bank than the typical startup, or has to have a small-scale operation underway to fund the early stages of the SBIR.

So, the cash that comes in comes in late. The SBIR should be used to get the government to pay for product development, not as a source of ongoing payments for work for hire. There are many situations where a small company was funded to develop an idea and then went on to build a company around it, but many were stopped long before they were close enough to be a developable product. The successes were rare because the typical SBIR had insufficient product development experience. Contractors provided a patina of product development input under TABA payments but it seldom went anywhere.

Torrance Optics was run by engineers and scientists, not professional managers. It showed everywhere in bad management process:

• Meetings were run with little attention to agenda and follow-up

• Financial discipline was lacking. There was no Finance Manual, and the DCAA manual drove all the accounting actions rather than company needs.

• The SBIR process and DCAA rules dominated the reporting, often at the expense of good financial reporting.

• Being an SBIR-based company did not provide an all-encompassing framework for financial control, where the normal accounting principals requiring the revenue and expense to be recognized in the period they were earned, where the Balance Sheet and the P&L talked to each other, et cetera, were not followed.

• The success rate for project proposals was in the 17-20% range and there were interminable, unpredictable delays. As a consequence, financial planning was extremely difficult and forecasts were systematically missed.

• Attention was not being paid to relationship building with government agencies, and focus was on protections not support. When things got tough, the customers did not step in to support Torrance Optics.

• Internal communications were not free and fluid so mid-level managers were not considered contributors to decisions. Good people left.

• Space planning was very difficult. It was difficult to match needs with space available. When Torrance Optics had to renegotiate a lease, there was no channel of communication. Basic legal obligations under real estate lending were often not understood and the sluggish character of the the legal system was used selectively.

• As cash got tight, Accounts Payables began to mount. Each vendor was assigned a priority based on what pressure they can bring on the company. Many vendors have few means to bring any pressure, or are not likely to bring them if they could, so vendors can be induced to accept token payments over time.

What is special about SBIRs in decline?

• While there are good reasons to seek SBIR funding, or to work with SBIR-funded companies, there are reasons to be careful if you do. The pathologies I saw were not caused by the SBIR process, but there are features and subtleties of the SBIR world that breed some of them. It has to be asked, “How many of the pathologies are caused by the nature of SBIRs and the peculiarities of the SBIR process and accounting?” I think there are a few reasons to say many are.

• The efforts are by their nature speculative so there are lots of risks.

• The nature of the cash flows gives too much control to the companies with the SBIR contracts by the nature of the SBIR financial control process. The basic monthly flow of SBIR payments and paperwork provide an opportunity for misbehavior, as follows:

  • In Step 1 an SBIR grant recipient signs a vendor to a supply contract with monthly deliverables.

  • In Step 2, the vendor delivers the work and provides invoices to cover payment of the work done. The invoices become “script” that the grant recipient delivers to the government for reimbursement.

  • In Step 3, the government pays the grant recipient, as long as the reports are in order, typically within a few days of receipt of the documentation.

  • In Step 4, the grant recipient is supposed to pay the vendor. However, if there are delays along the way, or if the vendor is slow to notice it has not been paid for the previous invoice, the vendor may get ahead of itself and provide an invoice ahead of being paid for the prior one. The grant recipient can take advantage of that delay, and even let them accumulate, and begin to use the government as a source of funding, at the expense of the vendor.

  • An observant vender would not send documentation for work done until it received payment for the prior month’s invoices. Otherwise, Torrance Optics would slow pay, leaving it on its books as an accounts payable.

• It is very difficult to get Overhead rates right. It the rates are too high, a contract may be priced too high or allow too ittle scientific or engineering work to be charged. There are many pressures to keep the rates low, but the DCAA regs, as of my reading, were more about being sure that costs charged were actually incurred rather than that they be low.

Behavior when any business begins to fail.

• Shed salaried staff.

• Slow pay, no pay vendors.

• Lose of SBIR agency sponsors

• Failure to perform against grant commitments.

• Failure to move from SBIRs Stage I to Stage II and forward.

• Failure to license or commercialize successful early stage SBIRs.

• Loss of key people, failure to fill important management positions, depending on board member for staff activities.

Appendix: POC-Spinout Companies

• Physical Optical Systems — founded in 1985 and sold to Mercury Systems in 2022

  • POC Spinoffs:

    • Intellisense Systems Inc (ISI) — still in operation, scattering of SBIRs

    • WaveBand Inc. — military radar company, bought by Sierra Systems long ago

    • Broadata Communications Inc. — commercially selling medical data/comm systems.

    • Redondo Optics Inc.

    • Applied Biocode Inc. -- has been selling diagnostic products for more than a decade. Publicly traded on the Taiwan stock market.

    • NanoOptical Materials Inc. – sold quantum dots, and is still listed as being in business. 1

      • Luminit Inc. (formerly LLC) -- commercially selling/licensing holographic display screen products -- no SBIRs

      • Farlight LLC -- Commercial lighting company, pivoted from fiber optics to conventional. Still in business but only locally -- no SBIRs

    • Intelligent Optical Systems was founded in 1998/1999

      • LEDI LLC? -- founded to commercialize LED Incapacitator technology. Effort abandoned.

      • Optical Security Sensing LLC —Vehicle for DICAST* commercialization. Brought in non-SBIR funding, deployed demo systems in several locations. Closed when DoD and DHS went with other chem. sensor tech. Founded Oct 11 2002.

      • OpTech Ventures LLC — created then folded

      • Intellipole -- Joint venture with a polymer telephone pole company that developed AI-enabled autonomous radar/IR/Visible border-security

      • Innosense LLC shut down after 20+ years in business

      • Watchposts. Shut down after 2 years.

        To go back to the beginning of the Memoir posts

        Share

        Henry Lichstein's Of Interest Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

My neighborhood has quiet streets that eventually spill out to the Pacific Ocean. One regular passerby was Jim Birch, who was building a company to distribute legal background music via the Internet, Cloud Cover Media (CCM). The head of his board was Bill Collins, a veteran early-stage investor who leads an investment group. They would eventually hire a top flight tech guy Stu Larking, a new CEO Mark Lehman, a new CMO Mitch Bishop, new VP of Sales Bill Christen, and then a great bookkeeper Debbie Mulvaney. I became an investor, advisor, and then CFO. The Board included John Osborne and Abilash Patel.

CCM’s success was the inspiration for writing about the stages of company growth. It all happened because we were so conscious of there being a process of growth and we watched it unfold day-by-day. To make CCM work as a business was a company-building task and a financial management task. It became a tremendous education for me, different from helping a big company like Citibank work because small things could make a big difference.

This post describes what we did to turn CCM from a disorganized startup with a simple but good idea into a well-organized small company that was getting bigger in a predictable way, month-by-month and thereby become a candidate for highly profitable sale. We were able to paint a believable picture of how the progress could be continued to high levels of revenue and substantial profit. When we turned cash positive on an operating basis, the past and the future were linked and we were able to command a premium price upon sale. When CCM was sold to SiriusXM, a Fortune 50 company at a healthy multiple of future revenue, we returned roughly $7 for every dollar invested.

Moving Beyond Managing By Our Checkbook

When I started investing in the company, CCM was a classic startup where cash was managed as it passed through the checkbook. It had existed as CCM since January 2011. Bills were paid only when they were funded, consultants were preferred over employees for cash management ease, and financial control was somewhat loose. Many basic costs and processes were ignored.

The departure of the original bookkeeper was an important opening to make changes. Debbie Mulvaney became the part-time bookkeeper in June, 2017. She was smart and good and fast. Debbie took over the QuickBooks accounting and I cleaned up the basic financial accounting. I learned about CCM’s Recurly back-office system and ChartMogul Financial Management systems and began to build the links between them, often by creating little models for things like revenue recognition, cost accounting, inventory control, revenue forecasting etc. Bit by bit, things came into focus.

Recurly and ChartMogul are still in use, as of 2026, although I doubt my models have been maintained.

Reporting and accounting were set up to match the eventual SaaS business model from long before the business coalesced, which helped the Start-up begin to look like a real company from very early. By doing things that way, we got a jump on “history” and were able to show growth from the earliest possible date.

And we documented every step of the way. Debbie was demanding and I tried to respond. In the end, the public Financial Notes made us credible to our various stakeholders and the 275-page confidential Finance Manual supported Due Diligence and eventually the sale of the company in January 2022.

What we did:

• With the early rounds of funding in place we could move away from “Managing by Our Checkbook” into a posture of planned growth and having the balance sheet and the P&L talking to each other.

• We structured our P&L as a BAU company showing cash generation and normal cost above-the-line of Operating Profit and the one-time startup costs of creating the company below-the-line of Operating Profit.

• MIS and accounting processes were built consistent with general guides for basic business structure. We made aggressive uses of both the balance sheet and the P&L to allocate costs and revenue to their proper time periods and ledger lines so we could properly report our profits. We wanted to look like a mature company long before we were by using tools typical of later stage companies as early as we could. Financial Earnings were the net of the above- and below-the-line operating numbers but the BAU Operating Earnings were in fact an accurate of portrayal of what the company was capable. We wanted to show that we would generate large earnings if we could turn the crank faster and grow. We were living the SaaS model and it worked.

• Our funds management models depended on our ability to forecast revenue and project costs. We had a law of large numbers working for us on the revenue side: We were earning money in small amounts from large numbers of customers across a number of distribution channels. We were in control of how we spent our money to bring in customers in each channel. Our tech guy Stu Larking and our CEO became masters at buying Google AdWords (now just Google Ads) to drive online sales, and Bill Christen with Mark managed the salesmen and large customer sales. Once we started building reliable delivery hardware boxes, we stopped losing customers and were able to manage revenue growth month after month. That was our revenue projection driver and proved reliable.

• We reported the imputed cost of music rights long before we actually paid them so we could show our profitable revenue growth starting as early as possible. The costs were shown as Accruals of expense and Reserves as sources of funds which were reversed as appropriate as we brought all our rights relationships into compliance prior to the sale to SiriusXM. Each rights contract dealt with history on its own so our lawyers’ respective skills were important. Remaining reserves were moved where needed and reversed upon sale of the company. (Bill Collins affectionately referred to the reserves as “Easter Eggs Henry is distributing around the Balance Sheet.”)

• We got control of Revenue Recognition after a tedious deconstruction of ChartMogul’s database content to see how it worked and how it could become the basis for an accurate source for revenue recognition on a monthly basis. ChartMogul’s Product manager and CCM minder Sebastian Simila helped us get a new report which gave us visibility of EVERY subscription payment by month so we could trace every dollar as it entered A/R and became Revenue. We teased apart every dollar of fixed price revenue to its source. We truly had the Balance Sheet and the P&L talking to each other.

• To both measure true profitability and manage salesmen, we decomposed sales commissions into one-time costs and ongoing costs, and set up balance sheet items to manage their impact on profit and cash flow over time. Again, more Balance Sheet and P&L talking to each other.

• We set up a Sales Commission system to tie each sale to a sales person or organization.

• We realized that our hardware ordering, shipping, sales, and shipping were integral parts of our financial management and created a side system to manage the financial view of inventory management and hardware sales distinctly and in lockstep with the physical side and to link the two through the Balance Sheet (yes, getting the B/S talking to the P&L). CCM’s hardware movement was treated as Items In Transit just as checks are treated as Items In Transit by a bank.

• We plugged a few revenue holes by reviewing back office booking practices, putting the invoice management process into accounting control.

• Early on, we decoded the accounting of the music rights agencies so we could keep precise info as to what we owed each according to their expectations. We understood how they tracked and calculated what we owed them, and reported accordingly. This became critical when COVID hit in early 2020 since we had to pause many customers. We kept our rights payments in synch with our revenues. What we learned was at a much greater level of detail than known to SiriusXM, to what I could see.

• We set up amortization of software costs in such a way that a simple monthly calculation at the developer level rather than a project-level accounting system fed the cost allocation for expense recognition. This allowed us to show reaching financial profitability a year ahead of cash flow profitability, again making us look mature ahead of our time.

• We had not renegotiated the per customer cost of our Recurly contract for back-office services since day one. With our steady growth showing larger and larger numbers of customers, I opened a conversation about a long-term supply contract. I was surprised to find Recurly poorly-prepared and eager eager for the conversation. They offered terms that literally appeared to be too good to be true: they gave us substantially reduced back office costs, at a fixed annual cost despite our dramatic growth over the three years of the contract. (Recurly were bought by KKR soon after we signed the deal. I surmise their interest in getting our three year contract was influenced by their desire for contract depth as they were negotiating their sale. I was a case of ‘too good to be true’ being true!)

• I fought more than a few battles over attempts by some liability holders to convert balance sheet items like loans or deferred salary into “cashless investments.” These fights were harbingers of self-interest issues I was to see in other private equity-managed companies.

• The value of having a Finance Manual was demonstrated at time of sale. We were able to disclose and explain all of our accounting decisions, and digital rights fee calculations, up front. All Reserve calculations were fully disclosed and not challenged.

• Bill had been meticulous in maintaining investor records from the early days and he and I were able to give very good, detailed investor records to Alkali Partners our investment bankers. Shane Hubble and his colleague Chris Skubic, in turn, were great managing the sales process and in delivering SiriusXM as the buyer. (We had aborted an earlier IB contract — it was worth the cost.)

• All of the careful attention to process paid off during the hand over upon sale to SiriusXM. We had a detailed process in place before the sale as to who was responsible for all outstanding expenses. Debbie and SiriusXM monitored payments for four months with no arguments. The outside lawyer who managed the contract preparation was in many of the staff discussions. I became a close contact of the SiriusXM senior accounting and reporting guy Cary Krefetz.

• One of the things I am most proud of in my financial career is that SiriusXM did not take even $1 from the 10%, multi-$million escrow account for accounting issues set up at the close to cover and accounting issues uncovered after the sale. They had 15 months to review what we had done and all the transactions we booked and did not make any claim against the escrow! (Thank you, Mark Lehman!)

• One hole in the documentation of the sale proceeds distribution process became clear after the sale in January of 2022, at least to me. The distribution of sale proceeds to the roughly 100 equity holders was carried out via a huge spreadsheet that literally took months and cost millions of dollars to create. (A book could be written about that alone.) When all was said and done, CCM and SiriusXM booked the numbers and accrued the withholding taxes as appropriate, but all the numbers were aggregated for each recipient, within the model, nor for each investment by each recipient. The model disgorged the totals by recipient, but not the details of the calculations by shareholder holding. That left each shareholder adrift as to how to pay his or her income taxes, which needed that level of detail. In the months after the sale, I spent an ungodly amount of time to generate a spreadsheet to extract the profit by individual shareholder. I was paid a small amount by the company for doing so (thank you, Don Hall), and I charged a small amount to a few of the shareholders for their specific details. I am comfortable that anyone who paid me saved many multiples of what they would have paid to have accountant to do it for them. The ones who did it themselves were guessing.

• The tax reporting by the funds transfer vendor was not under accounting control: I tried to get their attention but was rebuffed and I am sure they treat all their customers the same and that still bothers me. The “tell” was that the transfer agency accepted the data from SiriusXM by FAX not by spreadsheets with checksum totals. There was confusion at every step of the process. For example, the numbers I got for my own personal proceeds were the correct numbers in each column but the lines were jumbled and the columns did not have totals. I learned what it meant to bring a process under accounting control by these efforts.

• One memory: Anxiety attended my only scheduled interview with the SiriusXM regulatory reporting person Cary Krefetz during Due Diligence. His first question was: “How much of your Accounts Receivables will you collect?” Me: “Every penny!” Cary: “How can you be so sure?!” Me: “Because I knew you that was going to be your first question and I have taken a reserve against what is already a conservative number.” Cary smiled, and we went on to work together very well from that point on.

Continue on Part 6: Torrance Optics” and Companies in Decline

Background & Overview

I decided to set up and run a small company in 2012. Using the name of my street, I formed Dryad Partners, which was a Nevada LLC despite its name. That LLC, which was closed in favor of Dryad Communications, LLC. a So. Dakota entity in 2014. (The name is still active.)

I had kept in touch with the communications engineer I had met in Vancouver a decade before, Karim Lakhani, and his company Advanced Interactive. Advanced Interactive was offering cheap, effective over-the-air Internet service in rural areas and First Nation tribal lands in Canada in with a proprietary software system of his design which used repurposed cable industry hardware based on the DOCSIS cable industry standards. It was operational in a number of First Nation areas of rural Canada using spectrum licenses formerly granted to UHF television broadcasters. The U.S. had set up a better-organized structure under a similar process, called the “Lower 700 MHz Bands”, in the Telecommunications Act of 1996. The FCC took back all UHF TV licenses in the Upper and Lower 700 MHz bands and auctioned the spectrum in bands designed for what was expected to be an explosion of mobile telephone usage by 2009. The 700 MHz bands were most appropriate for use by the mobile operators in rural areas because the 700 MHz bands worked over greater distances than the higher frequencies that had been made available to mobile operators in its early days. The auction raised roughly $20 billion and was a big deal. However, the LTE technology which was expected by 2009 was delayed. so the Lower 700 bands were still not being used. I saw an opportunity in this delay.

In 2012, I decided to try to set up Karim’s system in the US. While the Canadian spectrum was spread all over the spectrum map, the United States had learned from the Canadian experience and was much more well organized. In both cases, the idea was that way too much of the valuable low- and mid-band frequencies use for television coverage was under- or even un-utilized but the of the US Band structure made each license more valuable because they could be combined by an operator into a single network (this fact took us an embarrassingly long time to realize).

From Karim I learned that the US had created something called the Tribal Lands Bidding Credit (TLBC) which offered a few bidders in the $17 billion auction a reduced price for specific licenses if they provided Internet service to at least 75% of the people in designated Tribal Lands in the area covered by the license, as an inducement to offer service in the Tribal Land. Failure to provide the service after a 10-year window would forfeit the discount. The problem was that, some eight years after the auction, the technology for using it for mobile telephony had not matured. LTE was late, and mobile handsets were not available for the Lower 700 bands. The TLBC license owners were at risk of losing their credit if they did not provide coverage and the technology to do so was not available and it was going to be expensive if it were.

I saw an opening to install Karim’s system to protect a licensee’s discount. We could either take over the license at a lower price or be paid to do it by the licensee. It all depended on finding licenses with TLBC encumbrances, offering a workable business proposition to the owner, and being able to install and run the system. Nothing I had ever done before, but I set in motion the creation of a company to install Karim’s system on Tribal Lands in the US.

Creating a Startup

I needed to find partners and then helpers and then contractors.

I knew the local LA VC crowd, and I had been active in the MIT Club of Southern California. I took an MIT grad to lunch who told me “Tamra Johnson got her MBA at UCLA (after an MS in Management Science from Stanford) and she did not go back to work at Northrop Grumman.” Wow, I knew Tamra: she had been a 2001 MIT graduate in Aeronautics and Astronautics, and not long after taking a job in LA with NG she had become the President of the local MIT alumni club. I had met and talked to her at an MIT event and was very impressed. Maybe she would like to try a startup.

I asked Tamra to listen to my proposition over lunch. True to form, she stood up after our one-hour lunch and said, “Let’s Do It.” So, just like that, we were underway. If Tamra had not jumped in at that point, I never would have done it. Starting was ignorance on my part and a big leap of faith for her. We had spent all of two hours together by that point. By the end of the project three years later we had commuted to South Dakota and Nebraska a dozen or more times, in all kinds of weather, hired tribal members to work on our system, negotiated with Tribal chiefs for access to their land, hired many very talented local service suppliers, got invited to the South Dakota Governor’s Pheasant Hunt, and got to know and understand the FCC and its processes.

To get underway, we borrowed space nearby and “hired” interns (we eventually paid them). After a certain point we were working out of my basement. We started by testing what we knew and listing so we could learn what we did not know and collected contacts and information. We got to know the FCC processes in substantial detail. We learned about the needs of the big mobile telephone providers — Verizon and AT&T at first, soon joined by T-Mobile.

Karim put his suppliers and the talented people with whom he worked at our disposal. We got to know his system as much as outsiders could. The English-language version of his technical concept was the maxim “Don’t shout louder than you can hear.” It was the way he matched his transmission power and the set up to the needs of each user base (distance, interference, speed). It was how the system was installed and how we could manage them when they went live. We had to stretch the capabilities as far as they could be stretched to satisfy the people coverage requirements of the TLBC. By the time we were done we were able to work closely with Harry Davis, Karim’s marvelous University of British Columbia engineering advisor, to the point of re-writing with him the technology report by which we were able to demonstrate to the FCC we had provided the coverage required by the TLBC.

At first, we were ignorant of the nuances of the various Lower 700 MHZ Band designations set by the FCC or even why we had to know, nor about the limitations of the various mobile telephone delivery technologies, nor anything about the implementation of the TLBC. We did know that the continued delays of the next generation mobile technology, LTE, was costing everyone money and that the technology itself was very expensive and complicated to operate. We had no knowledge of the business and regulatory structure driving the mobile telephone industry or about how local broadcasting worked. We knew nothing about transmission towers or how to install fiber optic networks underground or microwave systems in the air. We also had to learn about the vendors and the needs of telephony and Wi-Fi and cable television. And about governance and service availability in the Tribal Lands. Over time we learned about all these things and more.

At a critical point, we found that only six TLBC licenses had been issued. When Tamra showed me the list of owners, I said “We are going to make money with this list.” I called the owners during October 2012, and settled on Long Lines Communications, the owner of two A-band licenses which together covered most of South Dakota and the top third of Nebraska. Each license had an obligation to provide service in a Sioux tribal land, one in So Dakota and the other in Nebraska, with no prospect of being able to do it as their deadline approached. They had requested a delay from the FCC which remained unanswered. Brent Olson was CEO, Paul Bergmann COO, of the operating company, which was in Sergeant Bluff, in far north western Iowa. I started calling Paul every few weeks to gauge his interest in talking to us. (Remember my mantra that “There is a thin line between persistence and harassment.") Finally, he gave us an opening by saying, “Henry, we don’t know what we are going to do.”

As this was going on in December, 2012, Tamra told me she had to spend a week in Omaha, NE, working on a project for her brother-in-law. I jumped! Omaha was a natural arrival airport for a trip to meet Brent and Paul in Sergeant Bluff, which was relatively close to Omaha. In turn, that was pretty much on the way to Santee, NE, the home of the Santee Sioux tribe. We could set the whole project in motion with this one trip.

Getting Underway

We got Brent and Paul to agree to a meeting. Then we found nearby a technical and communications service company, Calhoun, Inc in Sioux City, IA, who were candidates to install all of our over-the-air and Internet connectivity hardware, and arranged to meet them. With those meetings set up, we arranged meetings with a Nebraska-based tribal member, Marshall Pritchard, who in turn arranged a meeting for us with a former Chief of the Nebraska Santee Sioux, Butch Denny. Finally, for good measure, we arranged to meet the senior sales team led by Jeff Mason at Great Plains Communications Company in Blair, outside of Omaha, who had connected the whole state of Nebraska with a fiber network, which we needed for backhaul for high-speed Internet access.

Solutions for many of the major pieces necessary for the project to make sense were becoming visible.

Tamra and I met in Omaha, drove north to Blair, continued on and crossed the Missouri River separating Nebraska and Iowa for our meetings in Sergeant Bluff and Sioux City, then crossed back over the Missouri River into Nebraska and met Marshall at an intersection on the road from Omaha north to Santee for a meeting at a coffee shop with Chief Denny, and a current member of the Tribal administration Mike Crosley. We then drove back down to Omaha for the trip home, dropping Marshall at his truck on the way.

It was a very long and action filled day but all the meetings went amazingly well. In many respects that first trip set the stage for the next two year’s work, with things all following in logical order, although not all the steps were predictable.

• Great Plains took us seriously and, based on that meeting, treated us well for the entire project. They would eventually agree to run a fiber connection for us from their fiber network termination point in the town of Santee to the tribe’s tower west of the town (and it is still in use ten year later, supporting the Wi-Fi network we left in place).

• We had a very open and informative meeting with Brent and Paul, but the most important thing was Brent saying “It’s not in our business plan.” Tamra pulled me aside and said, “Did you hear that?” She had caught what I missed: these two guys did not care what happened to the licenses. With that, got them to turn us over to their Board, in the person of Mike Salsbury.

• Our resulting call with Salsbury was really the start of the project. After we described our business proposition, we offered him an almost insultingly small amount of money for their licenses. To our surprise, he countered with the idea that we pay for an option for the eventual purchase of the licenses at their full original discounted cost, conditioned on our having installed our system to satisfy the TLBC obligation. While that meant we had to pay for the system up front ourselves, the low cost of the option was a gift, which we soon after accepted. (Karim was committed and he, too, did not get paid until the end.)

• We really meshed with Calhoun Engineering. They were to provide us good service for the next two plus years. We got help here and there from other service vendors but, from start to finish, the high quality of the work done by Calhoun was critical to our success.

• We worked through the first contacts Marshall Pritchard made for us in Santee at the Tribe and met many more. Marshall represented us at a meeting in Rosebud in late January 2013. His report for that meeting gave us good guidance on whom to talk to and what to ask for so that first meeting in Nebraska got South Dakota rolling as well.

Setting up the system in Rosebud, in the center of South Dakota, thirty miles north of Nebraska, started with the same kind of whirlwind of travel, a process of finding people and service providers, and local operations help.

In Operation

We eventually got the systems up and running, built them up to about 50 customers in our year of operations in each location, and learned how to keep the Advanced Interactive systems up and running with as little help from Karim as possible. By the time we sold the licenses to T-Mobile two and a half years later, we had travelled to the area at least a dozen times.

A number of events, people, and decisions stand out as allowing us to succeed in the whole quixotic effort:

• Our first trip to South Dakota was in February, 2013, about six weeks after our productive first trip to Nebraska. We started in the town of Mission, South Dakota, at a huge coffee shop filled with long-distance truckers because it was the intersection of the major north-south with the major east-west highways in the state. We had been there for 15 minutes when this movie-star handsome guy, Shawn Bordeaux, walked into the coffee shop and came right over to us and greeted us by name. We never learned how he knew who we were, but we instantly knew he was important to us. He gave us a value-free lecture on American Indian spirituality. We learned that his father was head of the local Junior College and an important man in the Rosebud Sioux Tribe, and that Shawn knew not only everyone in the coffee shop but almost everyone on the reservation, whether Tribal member or not. He had nine children, most by adoption, and his family was a voracious user of Internet service. Over time he made introductions for us to anyone we needed to meet and was one of our first Internet service customers. (He later was to become a member of the So Dakota legislature.)

• We worked very well with Long Lines Counsel Mike Salsbury, from the beginning of the project to the final license purchases and sales. Even though we missed a few important deadlines in our contract, he honored the terms of the agreement. He was one of the few heroes we never met in person (but a fellow Board Member from my Teradata experience, Neil Banco, knew Mike professionally and vouched for him, so I was never concerned.) The Option/de facto transfer was signed Feb 4, 2013, and the license sale to Dryad and resale to T-Mobile occurred in February 2015.

• Karim’s software performed as it was supposed to, Karim’s hardware suppliers were reliable, Calhoun installed Karim’s systems as promised. Tamra and I learned how to read the system’s performance metrics and understood the procedure to tune each user, and we were off and running.

• A chance comment by a lawyer in Rapid City, South Dakota, led us to the largest accounting firm in the western half of the state, Ketel-Thorstensen. The second-generation leader of the firm Paul Thorstensen assigned us to a young partner Jennifer Konvalin. Jennifer did all Dryad’s accounting (I did the bookkeeping) and she was a great sounding board. (She was to work with me on three successive gigs: she advised as we closed down RGE, became the accounting firm for Cloud Cover for the sale to SiriusXM, and advised on changes to IOS. She has done my personal taxes for years.)

• We met with Tony Rogers and other members of the Rosebud Tribal Utilities Commission in person in Mission on February 5 2013, at the urging of Marshall Pritchard. Upon their suggestion, we applied for and paid $60 for a Santee Tribe Business License dtd , 2013. It was very useful and they were always helpful. No one ever asked for anything in return in either Santee or Rosebud.

• We were very well served by local, young, talented Tribal Members who served as our installer and local contact, Mike Haukaas in Rosebud and Justin Avery in Santee. They were each recommended by someone locally (Justin had installed the Internet service in the Santee High School and was also highly recommended by the science teacher at the local Junior College, Hank Miller). We kept up with Justin and he eventually took over the Wi-Fi-based system we installed in Santee to use and pay for the last few years of the Golden West link after the sale of the 700 MHz licenses to T-Mobile (of all things, T-Mobile chose to use DSL, not the high-speed fiber link we were paying for).

• Radio towers were important to us wherever we operated. In Rosebud, there was a 500-foot KOYA FM Radio tower in the center of the reservation which was impossible to miss and which we leased access to it on June 21, 2013. Others were harder to find so we became adept at finding them on satellite maps. In Santee we found an antenna’s signature, a long black shadow ending in a little hut. Upon checking, we learned the tribe owned a tower on the only hill that looked down over the town. We got a hearing March 26, 2013 with the long-term Tribal Leader Roger Trudell for the purpose of getting access to the tribe’s tower. I had watched the movie “Incident At Ogala” about the confrontation at the Pine Ridge Reservation Standoff led by Roger’s brother John. I had the image of a very taciturn person. As we arrived at the meeting, Tamra and I faced a panel of Tribal leaders and advisors, with Roger in the middle. I leaned over to Tamra and said. “When we are done, he will not say a word, but he will shrug his shoulders.” We went through our proposal. He looked to his right and to solicit comments, and then looked to his left. When all were heard, he looked straight at Tamra and me. There was a slight lift to his shoulders, and I smiled and stood up. As I was standing, I saw him smile. I leaned over to Tamra and said, “Stand up, we just got our approval.” And, indeed, we had.

• In the end, we, like any Internet Service Provider (ISP),’ were dependent on the high-speed Internet supplier in the state. While we had great support from Golden West in Nebraska, Great Western in South Dakota completely froze us out: despite having good fiber delivery to the county, they wanted to have nothing to do with us because we were threatening them. Their DSL service to Tribal members was terrible and they knew it. We ended up cobbling together an over-the-air link from the KOYA tower in Rosebud down to a local ISP Sandhills Wireless in Valentine, Nebraska, run by Sam Tetherow. Sam’s links provided remarkably steady service in Rosebud despite their precarious nature. Sam also helped us set up the initial system in Santee.

• Todd Tobin, from the South Dakota Gov’s Office, was a great contact for us. Among other things, he got us invited to the Governor’s Pheasant hunt in October 2013.

• We wrote all our own legal contracts from start to finish, with some help from my friend Merv Hecht. We had all the substance needed in each one, and, while they may have been short on protections and boilerplate, we knew our counterparties and got done what we needed to get done. This was the case with our Option with Long Lines, our contracts with Nsight, Two Harbors, and AccessMN in Minnesota, and our sale and operating contracts with T-Mobile.

Dryad took a digression into Minnesota in the Summer of 2013, to try to set up Lower 700 systems with two local ISPs, using Nsight licenses (Rob Riordan, Jim Lienau). Nsight gave us a $500,000 down payment, half of which went to Advanced Interactive and Bob Gullon. We worked closely with Kevin Olson, Cooperative L&P in Two Harbors, MN, who installed the system for a number of his customers; and Jim Nyhus, Lou Pliml, AccessMN, in Evelynth MN. Then, out of the blue, the FCC issued an all-encompassing Report and Order for Lower 700 on October 29, 2013. It rendered our Minnesota contract unnecessary for Nsight (it extended the deadline for TLBC for three years) so we had to shut down the Minnesota project and the partnerships with Cooperative L&P and AccessMN, but it made Dryad’s option with Long Lines in South Dakota and Nebraska very valuable because it forced interoperability of handsets in the Lower 700 bands. Our Option with Long Lines preserved our role in South Dakota and Nebraska. (Dryad later hired Pliml for support in Santee.)

• There was an interesting incident. We experience system-limiting interference in January in southern Rosebud which logically could only come from a station owned by US Cellular in Mullen, 80 miles south. When we called friends at US Cellular, we learned that their whole system had been hijacked by hackers on December 24th 2014 and they had to physically visit their sites, one by one, starting with their most important ones, and power cycle their equipment to reset their stations. . Northern Nebraska was not an active system so they had not visited it, but the hackers started sending random transmissions which interfered with our customers transmissions, starting in January, 2015. Our friends moved up the priority of the retaking of the Mullen station.

Selling the Lower 700 licenses to T-Mobile:

During 2014 it became clear our highest likelihood of making money would be a sale of our licenses to T-Mobile. We knew that no negotiations were allowed while an FCC spectrum auction was underway and they took about a year to run their course. The next one was to start in nine months. I spent a month to find the right person in T-Mobile to talk to, Scott Sundblad. I called to tell him who we were, what we were doing, and that we would likely have control of our licenses within a year and that they would be for sale when we did. He understood my message and confirmed he knew how to get in touch with us. I circled a date on the calendar two weeks before the spectrum auction was to begin nine months hence, and told Tamra, “We will not call them: they will call us by this date.” In the event, two days before the date I had circled, Scott called and asked: “How much would it take to buy your licenses in So. Dakota and Nebraska?” We told him our price and we stuck to it.

We wrote the contract with T-Mobil ourselves, providing for contingencies in the FCC approval processes. We made our case to the FCC on October 17, 2014, and after the FCC figured out how to approve our applications despite some flaws unrelated to the coverage analysis, we received conditional FCC TLBC Waiver Approval and Sale Approval letter allowing us to purchase from Long Lines, dated Feb 26, 2015, and sold the spectrum soon after.

The twist in the process that caused the delay was signaled on the first question from the FCC examiner on October 17, 2014 was, “Did you install LTE?” He knew we did not, so I struggled to figure out what was going on. I paused, and then we had the following dialogue:

Me: “I do not mean to be disrespectful, did you read the analysis report?”

Examiner: “I read every word of it.”

Me: “Then you know we installed the DOCSIS system because that was the first sentence in the Report.”

Examiner: “It was, I just wanted to hear you say that.” And he was looking at his bosses, not me, when he said it. I breathed easier because I realized I have passed the first test they had for me.

We then went through the full report, page by page, detailing how the DOCSIS system was able to cover so much geographical area, showing RF Coverage maps with dots where we were successfully providing service to individual customers, and enumerated where all the people were in the two reservations using census maps to show where we had coverage to demonstrate that we satisfied the 75% population coverage reqwuirement. When I finished, I asked, “Do you have any questions? Have we missed anything? Do you need anything more?” The examiner answered, “You have answered all our questions and we do not need anything more.”

We later found our what had happened: Long Lines requested a delay based on the delivery delay of LTE. We had instead installed DOCSIS, which, the FCC pointed out, was available when the Long Lines asked for the delay. In the end, they decided to give us approval conditioned on our eventually installing LTE, a huge win for us. The complication facing the FCC was that they are averse to approving the sale of licenses with encumbrances, which ours would be. But by getting us to accept the encumbrance, they were willing to overlooked their hesitance. And since we had anticipated that possibility in our contract with T-Mobile, and it was their plan anyway, they had neither the right nor the reason to object to the encumbrance..

Since continuous service had to be available in the coverage area, our T-Mobile contract called for them to pay us a monthly operating fee to continue operating our systems until they installed their replacement systems. We turned off the Lower-700 systems some months after the licenses were sold. We paid back all the money every customer had paid in both areas, and the DOCSIS hardware was packed and shipped back to Advanced Interactive.

All was done by mid-2015. It had been quite a ride. Tamra bought a house in Venice, CA with her part of the eventual profits, and I cleaned up my personal balance sheet with mine. It was fun beyond measure.

In Nebraska, Dryad Communications set up a Wi-Fi based over-the-air system, installed and operated by Justin Avery as a co-investor. That system was later sold at a profit to the Santee Tribe in mid-2020 when COVID hit. It is till operating and we understand Justin Avery is still running it!

Now move to the post in Memoirs re Cloud Cover Media

I passed up a possible job with a first-rate New York City VC firm and went to work for a small startup VC fund in Los Angeles because we enjoyed living in Southern California. It was a mistake and I regret it: I was too old and without the requisite skills to become a successful VC; the guy running the fund and I did not get along; and I did not like the VC mentality when all was said and done — you have to make all your money on the few successes you have so you have to make the deal to good for yourself at the expense of the founders.

The VC fund’s parent company was a well established, small, old-line private equity firm in New York, They were quality people, but I wasted a lot of time and made no money in the VC fund. But I did learn a lot about digital communications, which was very useful, and I was exposed to a lot of the experts in digital communications, especially the cable industry, which helped in a few subsequent gigs; and I met a lot of the VC world in Los Angeles, which was very good.

One of the many painful lessons learned is that it takes a lot of effort to close down the last of your investments and a fund can linger on, incurring costs and frustrating your investors, forevvvvver.

Lucix (2005-6)

My heaviest involvement for my venture firm was in Lucix, a space communications company in Camarillo, CA. I was on the Board as and, later, worked for them as their CFO for a year or more after my VC stint. Mark Shahriary was difficult but smart as can be, and a good manager: I learned a lot from Mark.

As I was leaving Citi, one of my young commercial banking friends, Andy Sriubas, had introduced me to Karim Lakhani, an extraordinary communications engineer who ran a private, little, tax-credit funded, R&D lab in Vancouver, British Columbia. Karim had made a career of using repurposed cable industry gear to deliver internet services over the air to First Nation tribes in the great barren regions of Canada. In time, I was to use his technology in the US so the introduction was of great use and importance to me.

Expert Witnessing (2004)

I was able to do good work in one project I did for Kirkland and Ellis: A patent troll went after Amazon over its OneClick product with a patent that I viewed as having been superseded by 20 years by the Citibank work on ATMs (Kirkland re Gatto Patent 6,149,055). We won, and the judge excoriated the opposing expert (and the lawyer), who had filed against Microsoft.

A few years later, I tried to act as an expert for a small company and it was a disaster. Our lawyer had neglected to brief me that my laptop was open to the opposing side if I brought it with me. (When I looked to him for help, he said, “The rules of evidence for expert witnesses say they can review it.”) So, they combed through my emails and files and used a lot of the material against us. Be forewarned!! Read the rules of evidence if you sell your services as an Expert Witness.

PrivaSys (2005-2006)

PrivaSys was a secure payment solutions and payment card technology company founded in 1999 by Joan Ziegler. I was introduced to the company by Antony Jenkins, who worked at Citibank in the card business at the time (and was soon to move to Barclays Bank as head of cards and eventually CEO). By the time I caught up with Joan she had moved the company to Newberry Park, CA, and was deep into a project with MasterCard. PrivaSys (which became FiTeq) had developed some great IP and produced an active card (battery and all) that set up an encrypted communications link between any POS terminal and the card issuing bank at the time of a transaction. As I liked to say, “You can verify the purchase of a Lamborghini with the PrivaSys card.”

When I stepped in, the books were a disaster and the company was running out of money. The company was watching its patent filing be ripped off by MasterCard and would soon sue MasterCard for extensive damages. The case was taken to arbitration and PrivaSys was represented by a master lawyer of the IP business, Spencer Hosie, of San Francisco. PrivaSys won its case — as the arbitrator said to MasterCard after a day of our testimony, “You got caught with your hand in the cookie jar.” With the settlement, Hosie made a small bundle and PrivaSys paid back its investors and then some (but not excessively). My deal was that I got a low hourly wage to clean things up but a two times bonus if PrivaSys paid off its shareholders. So, I was happy.

One interesting sidelight was that I learned the term imputed interest and dividends, as a way to treat a return on a security in a tax-favored way.

RGE Monitor/Nouriel Roubini and Bill Janeway (2007-2017)

While at Citibank, I had met an extremely smart MIT PhD in Political Science in 1989, Charles Ferguson, as he was preparing to move from academia to the professional world. I stayed in touch with him as he set up shop as an independent adviser and commentator. He became very well known nationally, writing books on technology, venturing, and politics, and creating movies (one won an Academy Award). Charles and a colleague had done a review of IBM for Citibank during the 1992-3 credit review which was very useful for us.

By 2006 he had invested in a macroeconomics commentary and consulting company founded by Nouriel Roubini, Roubini Global Economics (RGE), and he suggested me for its Board. Bill Janeway had become an investor and Larry Summers was already an Advisor. Its CEO was Charles’ former girlfriend in San Francisco, Camille LeBlanc. One thing led to another and Camille left and I became its interim CEO in 2007. I ran the company for a few years before relinquishing the role to a series of CEOs. Fortune did not smile on RGE and it was closed down by 2016. I stepped back in and RGE became my first Stage three to four to five closedown experience.

Things were good in the beginning. RGE was an early information gathering company with an editorial staff of superb writers working out of vaults in a repurposed storage warehouse in the Lower East Side, just north of the Holland Tunnel entrance. Nouriel had been predicting what became the 2008 Financial Crash for a number of years.

RGE rode Nouriel’s fame for a number of years and it grew to about 65 people and a reasonable level of subscription revenue. After a year of commuting, I had handed the company off to Dean Daniels by 2009. Dean eventually left, replaced by Terry Winters, and the last CEO was the former CFO Richard Green, who proved the rule that CFOs seldom make good CEOs. But it was not the only one in the macroeconomic analysis space, and Nouriel had become so tied to the 2008 crash that his story was no longer hot when the economy stabilized. The revenue crashed as Regulation DF cut back Nouriel’s access to seniors and limited his ability to distribute his information to a limited audience.

From a financial reporting perspective, it became a warning lesson on the wonders and dangers of being funded by Unearned Revenue in a subscription business. It is great to have large amounts of Unearned Revenue funding a subscription company when it is growing but a terror when sales slow. The situation has to be managed adroitly and reported and watched very closely. As an aside, most companies distribute Unearned Income to their P&L using spreadsheets that are not an integrated part of their financial control system. This lack of linkage became RGE’s point of failure.

In the end, I was paid an hourly rate to close it down. I learned painful lessons as I backed out of leases in New York City and London; closed down legal entities, tax accounts, and bank accounts in New York, London, Singapore, and Hong Kong; disentangled RGE from partnerships; and closed our, negotiated down, or ignored large and small accounts payable; and closed out the RGE Tri-Net relationship. One enduring lesson was to keep an operating checking account in a company’s name for at least a year after it ceases to exist to deal with transactions that show up long after the company is closed.

Nobody came out as a hero in RGE, which was very sad.

Park Me/Parking in Motion (2009 -2011)

At the behest of a friend, I became the informal finance guy for a bunch of just-out-of college guys who decided there was a need for yet another parking app, in 2009. It became fun to see what could be done with the software stacks in geocoding and mapping, aggressive use of Google Street View and Google Maps, and to be exposed to the world of cash-based small- and large-scale real estate businesses called parking lots. After some success and a few major hiccups, the company and smartphone App came together. It was soon sold off to INRIX, to the benefit of the founders: none of the F&F investors saw any return, and it was absorbed into INRIX in 2015. That INRIX came to the rescue is a very tangible example of my aphorism “There is a thin line between persistence and harassment.” In the case of INRIX and Parking In Motion, I had made a call to INRIX every five weeks for almost a year.

One of the founders went on to become a very successful eCommerce CEO. That is his story, yet to be written.

Now move to Dryad Communications

Moving to Citibank happened by a fluke — a chance meeting with Paul Rudovsky in the streets of Manhattan in 1970 when I was about to take a job with Equitable Life — and was the start of a thirty-year career (1970-2000), with a wide range of challenges and opportunities.

After my two years in the Pentagon, Citibank’s scale was not intimidating. My interview day concluded with Bill Spencer, who summarized me by saying, “Well, he knows all the words.” That was to be the introduction to the first six-years of my 30-year Citibank career was an entry-level senior staff job developing management information systems at the then-new 399 Park Avenue Headquarters.

Following that, I was for two years the Citibank Regional Treasurer in Nairobi, Kenya.

When I returned to New York in early 1978, I was recommended for a job reporting to John Reed, then running the newly-created Consumer Bank. I became his financial controller, soon with all of the Division level financial staffs reporting to me, which made it a pretty big job. I was to spend twenty two years in a wide variety of jobs for John, in finance, treasury, risk management, technology, and representing Citibank to academic and non-profit organizations. In 1984, when John succeeded Walt Wriston as Citibank’s President, Chairman, and CEO, I became his Chief of Staff, carrying out a broad range of assignments.

I never rose above the rank of Vice President at Citibank, which was not a big title, but I was sitting next to or just behind some of the most fascinating executives in American business. Walt Wriston had a commanding position in American banking. John Reed was a person of immense capacity for action. The tasks I was given and the freedom of action I had when I was working for John were extraordinary. John assigned me tasks and never questioned what I did. I tried never to give him any reason to doubt my advice and he supported my work. My time at Citibank was — from start to finish — exhilarating, challenging, excruciatingly hard work. I gave a lot and got a lot. The seniors at Citibank were of the highest quality, and the professional opportunities were limitless.

My time reporting to John came to an end when John was pushed out of Citigroup by Sandy Weill in early 2000, two years after the merger of Citicorp and Travelers in 1998. I had by then moved to Los Angeles, to run some technology projects. I took a retirement package in late 2000 and shifted my focus to the world of Venture Capital and investing in, and managing, SMBs. This post is my view of what I did and what I learned during thirty years in financial control and treasury management in Citibank.

Early Corporate MIS (1970-75)

Walt Wriston was the architect of change who created Citicorp and a new corporate culture in a classic but very internationally-focused NYC-based commercial bank. One manifestation of the new corporate culture was the creation of a myriad of financial systems focused on how the institution was managed rather than what had to be reported to bank or SEC bureaucracies. These systems, called Management Information Systems (“MIS”), sat on top of corporate accounting, providing a profit center and customer profitability overlay to corporate reporting needs. They were in place and helped management decision making as Citibank introduced new technology in the late 1960s and into the 1970s.

I played a role in creating Corporate MIS at Citibank during 1970-75. The process had been started a few years before I arrived in 1970, by John Reed, who was then a recent MIT and US Army Officers Training School graduate, steeped in technology. I worked for for a driven man, Charlie Kelley, and an old timer, Bill Maher. My staff wrote its own software, first on the IBM 1130 process control computer, then much of it on the newly-installed DEC-10 and DEC-20 minicomputers. Citibank was the first user of the ADMINS DEC DBMS package from David Griffle. Later, we became the first US user of the large corporate database management system developed by the German company ADABAS (which remained an independent company until 2025 when it got absorbed into IBM). This was during the early stages of the proliferation of computer-based systems for operations support and the beginning of the introduction of analytical computer terminals throughout the bank. I had a ringside seat to all the computer-based systems in the bank since the central systems drove the decentralized profit center system used for monthly reporting.

Hand in hand with the Profit Center approach to MIS was the introduction of marginal cost pricing of funds flows within the bank. It was “invented” by Charlie Kelley and I became his “bulldog”. It was not popular when rates were high, with the lending groups, and it was not popular when the rates were low, with the consumer groups. Marginal cost pricing forced a recognition of current period cost of funding in all products and for all customers of the bank. The Marginal Cost Pool Rate was the theory driving all of the Profit Center MIS.

In parallel with the MIS work, I was for a time the staff support to Walt Wriston’s Prime Rate Committee, which moved banking from administered lending rates to interest rate market-driven rates with more frequent and smaller changes in the Prime Rate (they were called “hat sized” rates because they were quoted in 1/8’s), and Citibank called its rate a Base Rate rather than a Prime Rate. (I guess I got the job because I had a recent degree in Economics from MIT: the role went hand in hand with marginal cost pricing.) It also put me in weekly and in some cases daily touch with many of the Citibank senior managers, which was later to become a great asset: Paul Collins, Joe Fleiss, Hank Muller then Bob Martinson, Ed Palmer, Jack Heilshorn, Freeman Huntington, Don Howard, Walt himself, once in a while John Reed.

After five years, my independent Corporate MIS unit caught John Reed’s attention (in a bad way) and he got it disbanded. He absorbed some of my staff into his growing Operating Group, and I found jobs for the rest and closed down my operation.

Regional Treasurer in Nairobi Kenya (1976-77)

Taking a chance, I put in for an overseas assignment. Citibank had made its mark in the 19th century as an international bank and most of its leadership had international experience on their resume. I traveled to Paris to meet the person running sub-Saharan Africa and came back as the new Regional Treasurer in Nairobi, Kenya.

Janine and I and our two children packed our West Side Manhattan coop apartment and moved to Nairobi by way of Paris and London. I traded my chops as a corporate MIS guy for a role setting up or closing down treasury functions in newly-created branches in Africa below the Sahara. I had the Regional Treasurer role for two years.

The Treasurers around the world were a savvy, professional, supportive group who were great to work with. I knew the management from my New York history and developed an easy working relationship with headquarters and throughout the world. I would like to think I learned how to adapt to the rhythms of a Citibank as a global organization and to the unique character of its culture. It was very much a “contact sport” because the people were so important and there was scant telecommunications. I was constantly on an airplane (Africa is five thousand miles from East to West, and Nairobi is 4,200 miles from London and 6,800 miles from New York).

My boss in Kenya, Bob Grant, and his support in New York, Ham Meserve, had run the branch system in the Middle East as it exploded for Citibank in the 1960’s. They brought this background to what was hoped to be an explosive period of growth in Africa. I set up an offshore bank in the Seychelles, for which I carried out “funding by teletype.” I closed down operations in Lagos, Nigeria, and negotiated agreements with Central Banks in Senegal, Nigeria, and the Congo. It was a whirlwind and the experience was fantastic. But it was not destined to last.

The rub was that Africa did not explode. In fact, its prospects dimmed. Then, my Treasury boss in New York, Free Huntington, died (put his head on his desk, during a meeting, and “passed away”). Soon after, as 1977 was coming to a close, it was agreed that I would ship back to New York and I ran back to the safety of being in HQ. The peripatetic life as a member of the Overseas Staff was not to be ours. When I asked my human resources person in New York what they could do to help me find a job, she said: “First thing: find yourself a desk.” So, I realized it was all up to me. I indeed found myself a desk, and started to knock on doors.

Consumer Group Financial Management and Treasury (1978-1984)

The Citibank I returned to from Nairobi was a very different Citibank. John Reed had moved out of the Operating Group (the “back office”) and had carved out the consumer branch network around the work as the Consumer Bank. He was installing proprietary ATMs around the world. He had a Finance person who had set up a rudimentary finance system but she wanted to manage a branch, and John needed a new finance system to run financial reporting in the Consumer Bank. When I showed up at his door and asked him if he needed any help, he said he did. After checking me out (thanks, Paul Rudovsky!), he offered me the finance job. After a few months, when I described what I wanted to do as the Financial Controller, John and his Division Heads decided all the division finance staffs should report to me as well while we built the finance system: it had become a big job. (The safety of Citigroup Center, which was completed in 1977, had been called into question and had to be repaired in 1978. By chance, the welding equipment was being hoisted into place as my job interview with John was underway, visible out of John’s office window. I savored the metaphor.)

The software needed for the in-house financial management/MIS system was projected to cost more than $1 million, a lot of money in 1978, and required what Citibank called an “MEP” (a Major Expenditure Proposal), an approval document typically requiring a review and a number of signatures. Notwithstanding, when I handed John Reed the MEP, he just asked, “Did you do a good job?” I said, “Yes,” and he signed it on the spot. I was left standing in his office, holding the signed MEP, with my mouth open. By that act he communicated “It’s all up to you.”

One of my buddies from SSP, Chuck Holland, came to work for us to create the Profit Center system on ADMINS and he first created a nascent version of what we today call Excel, from scratch. It was a marvel. Also, I inherited a talented IT guy named Don Irion and he in turn found a very young Adam Bosworth buried somewhere in Citibank. I brought him in to make Chuck’s hand-crafted system industrial-strength system on ADMINS. Adam showed the brilliance that he later displayed in a series of jobs at Wells Fargo, on his own in a small startup, and then at Microsoft where he overcame Bill Gates late-to-the game Internet Explorer 3.0? and again on his own with BEA Systems. It was fun.

We managed down the errors (delta between financial and managerial books), seeded the system with talent, and got the upper hand in operations to speed the whole process up. The Consumer Bank nearly broke Citibank with a huge surge of credit card losses, but we lived to see the situation stabilized. We had wins all the way around. I once asked John if he knew how much we had accomplished and said, “Yes. I used to get complaints from Corporate about our numbers all the time and now the complaints have stopped.” OK.

I had put my one course in accounting to good use. I understood the stocks and flows, I understood how the P&L and the balance sheet talked to each other. I could hold my own in conversations about reserve calculations and capital accounts. However, it would be twenty five years before I really learned how to manage vendors and keep cash under control, when I became the CFO or CEO of small companies.

Working at the top early taught me one thing that served me well: “Never have to use your boss’s name.” In an early interaction, when I delivered a message to a senior manager: “John wants you to do such-and-such.” The next time I brought a request, he said, “Is John asking me or are you?” I realized that I had made a mistake and henceforth tried always to deal with needs, not who is expecting what. The value of being able to speak in my own name only was rewarded time after time.

Just as we got the financial control process under control, the Savings and Loan Crisis of the late 1970’s settled upon us. It was decided I would use my Treasury knowledge and contacts around the world to create the role of Consumer Bank Treasurer, so I went to work. The S&Ls were encouraged by the Reagan Administration to amortize their losses rather than recognize them. We realized that having the S&Ls relieved of having to recognize their losses meant the New York Consumer business would be under severe pricing pressure. We had to cut way back on mortgage lending so I was not popular, but I protected the bank from making what would prove to be a big batch of bad loans, and Citibank did better than the S&L industry, by far. I used my office to spread truth. (The S&L’s were to become zombies because their numbers were hiding the situation. It multiplied the eventual S&L losses by a lot.)

During that time, the Consumer Bank lived by the precepts of John Reed’s famous six-page “Memo from the Beach” (Summer, 1976). It has become one of the most quoted statements of business intent ever written by a business leader. It defined consumer banking in the coming technology driven world. It popularized the concept of Customer Net Revenue, and made it clear that the Consumer Bank had to keep the concept of customer need central to its business offerings and that profit as a business was to be made up of profit from each of its customers.

Consumer Banking System (CBS) (1984-85)

When John became CEO of Citibank in September of 1984, replacing Walt Wriston, I became a free agent for a while. The first project came along quickly, when Rick Braddock, who was put in charge of the Consumer Business when John became CEO, came into my office, his first visit ever!

Rick asked me to undertake a review of a major effort called the Consumer Banking System (CBS), being built at Citibank’s tech subsidiary TTI in Los Angeles as a branch banking terminal network and integrated back office. John had created TTI in the early ‘70s to accelerate the introduction of digital communications into Citibank. TTI was built on a tech group called Quotron, who had created a stock sales terminal for Merrill Lynch, but by then all the major Quotron managers had left. CBS was the largest system TTI had ever tried to build, with well over $100 million spent on the project by 1984 (a lot in those days!). The New York Branch business had no interest in installing the system, and this had just become Rick’s problem. He was asking me to take it on as my problem.

CBS had John’s fingerprints all over it and Rick was afraid to kill it himself, but if I did, he would not be blamed. John agreed to the plan, knowing that I would do the review without clearing the outcome with him. I moved out of my 399 Park office into an office in the New York Branch network.

I began to commute to Los Angeles as 1984 wound down. It was clear from the beginning that things were very much awry. Work on CBS was proceeding without any commitment from the user. No software was up and running, and the custom hardware’s design was in flux. Nothing was considered finished. A major IBM mainframe system suitable for back-office processing was set up and running in TTI’s basement but was used for software development for an hour a day, if that. Documentation of the software was in a stack five feet high but no one ever referred to it.

I knew the senior tech guys on the project well and we set out to review where things stood. After only a few meetings it was clear that the guys had not come to terms with answers to some significant questions about the system. It was no wonder there was pushback from the intended users of the system.

Our first task was to decide where we were on the project. We agreed on six questions or open issues that had to be answered before we could proceed. Failure to resolve any of the six would doom the Project. They were:

1. Can the chosen technology at its current state support the customer needs of the system?

2. Can the New York branch bank make the cutover from where they were to the new system?

3. Can the hardware supplier for the chosen specialized communications hardware make all the changes needed as part of its standard offering so Citibank was not left modifying successive hardware generations.

4. Same for the database management system software on which CBS back office was to be built.

5. Can the IBM mainframe back-office systems necessary to support CBS be made to work.

We plowed through all six questions during the summer of 1985. To everyone’s dismay, the answer to every one was “no”. It turned out that functionality was insufficient to the task; the hardware and software vendors declined to commit to add our needs to their products; it would have been impossible to run the two systems in parallel and a single switch cutover was considered too risky; and, the final straw, was that when a team of testers came from New York to test run the back-office system, which was supposed to be finished but had never been tested, it was shown that nothing was ready for production. The people in Los Angeles were amazed to watch the New York team work harder than they had ever seen computer operators work and yet they could not get the CBS back-office systems to work.

I never had to make any announcements. The LA people realized they had to cancel the project without any prompting from me. The IBM back-office mainframe computer system was packed up and moved from LA to New York (as a backup to the vulnerable standalone system on which the New York Branch System depended). The whole COBOL unit was disbanded. The hardware and comm guys went over to other projects (Paul Glaser, Shel Klee, George Festa). New York implemented their system (Mike Cappi, Howie Schectman). I spent some time on the Long Island City real estate project with Hal Duford.

Corporate Treasury (1985-86)

Nancy Newcomb was the Citicorp’s Treasurer and I joined her office in mid1985 where I worked on a number of projects: helped implement a large-scale international Risk Management System and wrote a Treasury Manual (which showed the value of the written word – Peter Gallant was my champion) with Betsy Iger.

Chief of Staff to the Chairman

Financial Reporting at Citicorp

By early 1986, two years after he became CEO and Chairman, John asked me to rejoin him to create a staff support unit in the Chairman’s office. He was remodeling part of headquarters at 399 Park so he could manage the way he liked to, and I was to take on a variety of financial management and other analytical and coordination roles for him. I took over a small staff he had assembled and added a number of my new people.

One of my first “assignments” was to make the MIS for the whole corporation tell the same story that was being told by the financial reports. That is, make the whole be the sum of the parts. When he arrived as CEO, the sum of all the business unit profit was roughly twice what Citigroup was reporting in total. He was asking me to do for Citigroup what I had done for the Consumer Business — find a way to have the profit center reports tell the same story as the financials.

Reminding myself that I should “never use my boss’s name,” I had to manage down the errors in institutional MIS, again, with as light a hand as possible. For a whole year, I published summaries with the old way of aggregating profit on the left-hand side of reports and the error-corrected way of aggregating profit (no double counts, for example) on the right-hand side. After a year, I dropped the old numbers on the left-hand side.

Risk-Based Capital and Capital Adequacy (1986-1988)

I had done some work for Walt in the early days dealing with a review of Loan Loss Reserves and their relationship to bank capital requirements and I was familiar with the vocabulary of bank capital, so it was easy to get involved when the Federal Reserve came out with a proposed Capital Adequacy Policy proposal in early 1986. I had worked with Hans Angermueller who was tasked with answering the Fed so he and I went to work. We were to collect input from around the institution, write a response and get John to bless it.

When the response began to take shape, we looked at the calendar and realized that John was soon to leave on his four-week summer vacation, not to return until a few days before the response was due. If we could get his agreement on the approach we were taking based on an early draft, we could finish the response early and get it in early, before he returned. That way there would not be a mad rush in the few days he would have when he returned.

I gave John the draft of the response and a cover note a week or so before he was to leave. To my dismay, when he came in the next morning, he said “That was terrible.” Since I really thought we had done a really good job, I had to figure out what was going on. Saying to myself, okay, at least he read the material, I asked “Did you like the response to the issues?” “Yes, they were fine. But I hated your cover memo. You started by saying “we like the approach” and have a few changes. Instead, you should have said we hated your approach and here’s why.” So, I put a tape recorder in front of him and said, “Great, say what we should have said.” John then dictated what became the cover memo and we soon after sent the response to the Fed, early.

Since the Fed made public each response as it came in, ours was among the very first to arrive and the first of the big New York banks, and it caused a stir. Our position was taken up by most of the major banks, and Citibank’s position paper became the standard. The apparent unanimity among the banks was caused in part by John’s vacation schedule. (To be fair, history has not treated the position taken by Citibank and other large banks in 1986 as a good one. q.v.)

In the months before the 1987 Crash, John had travelled to Latin America. He returned with an urgent message to the staff that Citibank should issue as large a capital infusion as the market would allow. Citicorp raised $1.2 Billion? in what was then the largest single bank capital issue in history(?). Relatively soon after, Citi also took a preemptive $3.3 billion charge against LDC loans in the second half of 1987. Not long after the October, 1987, stock market crash occurred. Citibank looked prescient. Interesting to watch from the office on “the second floor”.

Group of Thirty Report on Clearance and Settlement (1987-89)

The October, 1987, crash exposed a large number of gaps in what is called “clearance and settlement” among the national securities clearance systems of the world. This is the arcane process by which buys and sells of market securities are first “cleared” — buys and sells offsetting each other for a net position — and then “settled” — where the net positions are paid. It is a precise dance among the hundreds of counterparties in each country and between counterparties in different countries, involving hundreds of billions of dollars each day.

The problem is that the dance is carried out to different music and following different conductors in each country. The market swings in the 1987 crash were so extreme that counterparties could not clear and settle because they did not know if they or their counterparties were solvent. The system froze at major choke points.

John called me in and described a number of things going on in the area in the month after the crash, and I was off on a new, fascinating two-year project dealing with the international clearance and settlement problem. While I had passing knowledge of the process and of the people at Citi who ran it, I had absolutely no expertise. That was about to change.

Morgan Stanley had a major stake in the process. Its ability to operate in each center and around the world depended on clearance and settlement. Its flaws threatened Morgan Stanley’s very existence. Morgan Stanley had been the impetus behind a non-governmental organization of people representing major central banks and investment and commercial banks formed in 1978 at the behest of the Rockefeller Foundation to be able to address these kinds of issues, called the Group of Thirty. The head of the Group of Thirty in 1987 was a former Governor of the Bank of England Lord Gordon Richardson. Their offices were in a Morgan Stanley building in London.

John was visited by Morgan Stanley’s Lewis Bernard in November of 1987. At the time, Lewis was a member of Morgan-Stanley’s four-man Management Committee and widely regarded as the one who guided the firm's strategic direction. He was accompanied by Jerry Lynch who ran Morgan Stanley’s back office. Lewis asked John to take responsibility for an effort to be run out of the Group of Thirty office to address the clearance and settlement issues. John agreed to announce the effort at a set of meetings driven by the annual cycle of Central Bank conferences around the world. Jerry and I were tasked with assembling the working group of operatives around the world and writing the report.

Jerry and I settled in for a lunch after Lewis left. Jerry told me confidently that he could write the report by himself if necessary but we needed to get the rest of the industry behind whatever came out. While that sounded pretty brave, it turned out to be true, in all respects. Jerry knew more than probably any one person anywhere in the world about the process, but that we did need to get the rest of the world involved and behind the proposals, whatever they were to be, for the effort to succeed. He was a delight to work with and the two years working with him and his myriad friends and professional associates around the world were marvelous.

I ran down to 111 Wall to enlist the help of Carlos Salvatori and Norberto Spangaro, Citibank’s counterparts to Jerry Lynch. Carlos very generously seconded to me Richard Stanley, his financial staff person, only 26 years old at the time, for the duration of the project. Richard became a friend for the rest of his life, and a tremendous help for the entire project. He assembled in one week a briefing book for John’s first speech about the project that was a marvel. While John ignored our recommended conclusions (his were much better), he had internalized the main points the project had to address and set the tone for our work. Richard went on to tremendous success in Citi, then in positions beyond Citi, and then tragically succumbed to leukemia at the age of 48 in 2009, when he was head of the Development Bank of Singapore Group.

Writing the report took over a year. It was to become the basis for efforts at the country level to synchronize the clearance and settlement systems around the world. Jerry Lynch assembled a working group of about twenty of his peers from New York, London, Paris, Tokyo, Hong Kong, and Singapore. Lord Richardson’s former Chief of Staff at the Bank of England, David Holland, showed up at our working group meetings and he and I became a team, acting together as scribe and writers for the process of collecting ideas and writing the report. By the time we issued the report, Lord Richardson had retired and unceremoniously pushed David into retirement. I passed the hat for operating funds for him to continue to work on the project and found him an office in Citibank’s offices in London. He spent about three years going to conferences to promote the report at the country level, and then he died. (Lord Richardson lived until 2010.)

The final issue before we finished the report was to get a definition of the phrase ‘The lender of last resort’ from Jerry Corrigan. When he blessed our formulation for the final report, the G-30 Report on Clearance and Settlement was finished (1989).

As an outgrowth of my work on the international team, I joined the US implementation team, working with David Kelley, Doug Engmann, many others, and for a few fun years was a member of the Board of the NSCC (National Securities Clearing Corporation) where I learned a lot about the settlement process and about risk management in clearing organizations.

Teradata Investment, Management, Sale (1986-1991)

Some of the early TTI management had broken away and formed what was to become a very successful hardware-oriented large scale database computer. With a projected scale measured in Terabytes (when the largest software systems were then much, much smaller), the firm and its products were called Teradata. At a certain point I got John to approve a significant investment in Teradata, which was a tremendous success. It was managed by Peter Geary, who was credited with the investment. Oh, well.

As a side note, as a Board Member I read the R&D partnership contract Teradata was about to sign with its marketing partner the NCR corporation to fund Teradata’s large product development costs as the contract was being drafted. To my dismay I found that NCR was asking for the contract to be abrogated if Teradata were sold to another company, with nothing for Teradata in return. I said, “At least, make it symmetrical — what if NCR is bought out?” In the event, two years later NCR was, indeed, itself bought by AT&T, who happened to be 40% of Teradata’s sales! AT&T was forced to spend over $500 million more to buy Teradata. Otherwise NCR would have lost access to its most important asset. Had the contract term been one way in NCR’s favor, it would have ruined Teradata’s prospects. (It turned out NCR never told AT&T about the buyout clause in the R&D partnership agreement and I was told by an AT&T senior manager they had only budgeted $250 million to buy Teradata, not the $525 million finally paid.)

Quotron

Citibank bought Quotron for $680 million as what was to be its entry into the computerized delivery of financial products to the investment banks. Merrill-Lynch was 40% of Quotron’s business and Merrill took its business to Bloomberg, not wanting to be beholden to a competitor Citibank for such services. And, Quotron’s system was proprietary. When Quotron was finally sold off in 1994, it was for a reported $100 million. In 1991, John had asked me to justify what Citibank would have to take as a write-down for Quotron. To my surprise he actually took the write-off I calculated, in the third quarter of 1991.

Credit Process, Policy, and Management (1990-1993)

Following a series of losses in loans to large retailer buyout and real estate company loans totaling in the range of $2 billion, John initiated a corporate-wide review of Credit Management that ran from 1990-1 through to 1993. It was managed by Bob Martinson as the chief credit guy and Bill Rhodes who was vice chairman at the time. The loans at issue were issued by the Corporate Bank then run by Larry Small. Larry Small left Citi in 1991, by which time he had been moved out of that position.

My role was to be John’s staff representative to the process and I ended up intermediating a number of specific analyses and ran a best practices peer review process among large New York City financial institutions (Chase, other commercial banks, Morgan Stanley, etc.). My conclusion from that Best Practices Review was remarkably simple: You have to know what you are and design your control processes appropriate to your mode of operation. A credit process for a commercial loan business is different from that of a trading room operation. I prepared the slides to describe the process to the Board, which was given in early 1992 and changes were carried out into 1994.

Early in the process, as the corporate bankers were trying to assess how John was going to insert himself into what they thought was their role, I was asked to find out if John Reed wanted to sign off on a $900 million credit line to IBM. When he said to me “Why would I want to be involved with that?”, I concluded we were on the right path. The changes were to create a new credit process that separated the decision-making process for individual instruments from the process by which overall credit-worthiness was determined. I followed my precept of never using your boss’s name and since I had helped him look into the credit worthiness of IBM as a company (note 2) and knew he did not want to make individual instrument decisions, I could push the efforts in ways he would be comfortable with.

Note 2: Charles H Ferguson Charles R Morris: “Computer Wars: The Fall of IBM and the Future of Western Technology” 1993.

Board Presentation as focus for changes, early 1992. Review of credit process 1992-1994.

Windows on Risk started with a single sheet of paper sketched by John in 1993 which was taken on by David Martin, who made it into a very successful professional career, first in Citibank and then in Alliance Bernstein and outside. The work Bill Rhodes did became part of the Wolfensohn Panel 1995??

Interacting with Technology from New York (1992-1996)

The loan losses Citi faced in the late ‘80s and early ‘90s caused it to pull back from a lot of projects in technology and elsewhere. By 1992, things were starting to look up and I asked to assess Citi’s position in technology around the world. I spent the better part of 1993 developing what I came to call Strategic Circles, putting all the major projects like smart cards, new ATMs, communications capabilities, etc. on a single pictorial with their technical interactions “around the circle” and time shown by the concentric circles. The present was at the center of the circles and five years out was on the outer edge. It was useful to look at the pictorial three ways, from the center out (how to get to the future from where we were now), and from the edge in (where we were trying to go and what has to be done to get there), and around the circle to show how the various things interacted with each other at any point in time.

It was clear from this view that there were many things going to happen all around the world. Supporting this was the newly emerging juggernaut called the Internet. Started in the DoD in the late 60’s, it had escaped and was about to revolutionize commercial communications around the world. I set about getting Internet-based projects underway around the world 1994/5, by hook or crook (money, encouragement, nudging). In the early days, there was surprising pushback from some at the top. John gave me freedom and I got things underway.

I had the chance to investigate the Mondex scheme in the early ‘90s, before it became public and I became a detractor (John Reed used my quip about MI Rising when the Mondex chip was broken in Stefik’s book The Internet Edge). My concern was that an unaccounted card that depended on absolute security of the chip was bound to fail, and there were other flaws. The early supporters NatWest and Midlands sold out to a 17-bank group and then 51% to Mastercard in 1996, with Mastercard picking up full ownership in 2001. Mastercard sank more than $6 billion in Mondex by my estimation. Citibank supported the VisaCash smartcard cash product and Wells with Mondex and Citi set up a trial in the Upper West Side of Manhattan in 1996 but despite a lot of good technical and marketing work, the consumer did not share Citibank’s enthusiasm.

One very small personal interaction was instructive on a number of levels. I had direct access to people at all levels of many Consumer Bank organizations, often in technology. From that, I once found myself in an awkward position vis-a-vis a very senior person: I solicited for the Chairman a memo about a situation in an organization reporting to the senior, for reasons I thought valid. The senior then showed up in my office to ask why I thought it was appropriate to reach into his organization to generate a memo to the Chairman, going around him. He pointedly ignored my invitation to have a seat at my desk so there we were, standing facing each other.

I realized there was a copy of the offending memo on the top of my inbox. I wanted to defuse the situation but to preserve three things: the confidence of the person I was working with inside the organization; the working relationship I had with the senior; and my responsibility as a confidential source for the Chairman.

I put my hand on the offending memo and said: “The memo is right here. I will give it to you if you insist. But I am telling you it will not affect you and nothing will change whether you read it or not.” I was asking him to trust me, and to not diminish me by forcing me to give the memo to him. We continued standing, facing each other, for nearly a full minute. He then slowly turned away and the subject never again came up.

Creating Openings in China 1994-2000

Bill Rhodes came to my office in late-1995 with a sheepish look on his face. He told me a story of a long relationship with an American ambassador’s daughter, Lady Diana Dougan, who became the Director of the State Department’s Bureau of International Communications and Information Policy. She had created a Global Information Infrastructure Commission (GIIC) designed to increase the opportunity for opening telecommunications with China.1 He spoke of conferences being planned for the next month in Shanghai and Tokyo at which he had committed to get John to speak. He went quiet for a moment. I made it easy for him by interjecting “..but you never had the courage to ask John to go and now you are in a bind: Yes, I can go. I was last in China in 1981 and I would love to go back. I can fill in for John (assuming I can get a visa).“

That lead to an exhilarating series of meetings and repeat visits where I spoke often about Electronic Commerce using credit cards as a basis for payment. At the time, China had only a primitive telecommunications network and well less than 1 million credit cards outstanding. But all of that was starting to change, explosively. It is likely that at least some of the founders of the eventual giants of e-Commerce in China attended one of the five or six presentations I made. At one point, Kenneth Liu and I met for a series of meetings in Shanghai. His family had had a long history in pre-WWII China. Jack Pan and I brought a number of smart card company representatives to conferences we set up in Beijing and Shanghai in March and April of 1998.

With Kenneth Liu’s help, I was able to establish a wide range of contacts for Citibank in China in telecommunications and bank regulation, which eventually led to Citibank gaining access to China’s local currency ATM network in 1994 (I was on Chinese TV using my ATM card). In a fun twist, Rich Stanley became head of Citibank’s China Country Manager in China in 1999 (and later China CEO in 2005). He was there when China joined the WTO in 2001. The Citibank ATM in the Peace Hotel in Shanghai was accepting nearly a $1 million a day in local currency on some days.

In one almost bizarre event, I went by taxi cab to a meeting of the foreign speakers at a conference with the Chinese Premier Zhu Rongji in the Blue House, the Chinese version of the U.S. White House. I was stopped by a guard with a printed list of attendees: I pointed to the name Bill Rhodes and and said: “I am standing in for Mr. Rhodes.” So he let me in the front gate and I let into the Blue House, and was ushered into the meeting of about 20 speakers and the Chinese Premier. All the while I was carrying a large briefcase which no one had ever inspected.

Operating Citi Advanced Development in California (1996-2000)

In late 1995, Pei Chia asked me to move to California and support Jerry Rao, who was to become the new head of TTI, which I did. Jerry was did not need me to sit outside his office. So, since I had already begun what was to become the first Internet-based Banking Application (DA 6 - 1996), and had set in motion a substantial amount of work in Human Computer Interaction, with Charles Mauro, I asked to be relieved of the support job with Jerry.

So, when, soon after Janine and I moved to Los Angeles, TTI began planning its move from the building on Ocean Park Blvd near Santa Monica Airport to a newly refurbished building just outside the new Marina del Rey development, I set up an expanded Advanced Development unit, operating out of a separate skunk works unit carved out of a building adjacent to the main TTI facility.

The new 12 or so person unit included Judy Darr, Miguel Abramowitz, and was able to tap into some other very skilled TTI people including Jack Pan and Warren Tan. At the same time, a flawed customized browser version of Citibank online being developed by a bespoke unit set up in Chicago was shut down, to my relief but after another $100 million was spent, and the powers that be in New York shifted Citi’s effort to the native Internet Direct Access. It was to be called DA 6.1 since my toy system had gone live with real customers as DA 6.0.

At a certain point I moved back to the main TTI building and began to commute to China and to mine Silicon Valley for ideas and Investments. My path was to include an investment in WebTV, which Microsoft bought for $425 million and a big profit for Citibank in April, 1997, a license for WAP, Others.

I was winding up my Los Angeles effort and planning to become bi-coastal with an apartment in Manhattan when John was pushed out by Sandy Weill in late 1999. I cleaned up a few things and negotiated my departure from Citi by late 2000.

Next post in Memoirs is my time in venture capital and doing financial control gigs in small companies. That is described here

“Memoirs: Simplify, Organize, and Connect the Dots” is the record of my job and activity history, broken into six Substack posts. It describes my professional activities over six decades and an acknowledgement of my many intellectual debts and influences. I learned new things and developed new skills at each stage of my career, trying to make sense of what I was doing and what was going on around me. I was an active contributor, participant, and often a manager in large and small organizations, in government, profit-seeking, non-profits, and philanthropic organizations, and I tended to manage my own wealth. I had a lot of fun doing what I did.

I want to convince readers that my conclusions and the lessons I learned came from something lived, something real, so they believe what I have to say about them. For that reason, I describe the jobs and activities in some detail, and I name many real names, the many people from whom I learned a lot. (I created one fictionalization, for reasons that will be obvious.) Full books could be written about many of the topics I describe in only a few dozen pages (such as Citibank or Cloud Cover Media).

This Introduction post sets the stage for all six posts of the Memoir entitled “Simplify, Organize, and Connect the Dots.” This Introduction includes a brief run through of my first two jobs after university, two years each in Texas Instruments in Dallas, Texas, the Office of the Secretary of Defense (Comptroller) in the Pentagon. In successive Memoir posts I will describe what I did at Citibank, in venture capital, Cloud Cover Media, Dryad Communications, companies in trouble, and in philanthropy.

While many of the Substack posts may just remain as posts, I plan to knit together a few sets of my Substack posts to create a long-form books. The first such book-style post would be all six posts comprising the complete Memoir (and including the post on Publishing, for example.) or both the Five Stages book and the Mantras I have written six posts which together constitute a Memoir

The second book-style post will be “Stop Managing By Your Checkbook: the Five Stages of Company Growth©.” This describes what I learned about financial control in companies at various stages of growth during my six-decade professional career. I may put a pay wall blocking off the book compilations.

The post “Mantras” is short form post of aphorisms that encapsulate some of the things I came away believing. It might be expanded into a long-form book.

I focused mostly on financial control, not necessarily on the whole picture. That was where I knew what was happening.

Acknowledgements

I had the benefit of great teachers in school, tremendous managers and colleagues in the big and small organizations I worked in, and marvelous like-minded friends and collaborators in my philanthropic work. I also gained access to professional practitioners and source material in many, many fields as I made my way through the experiences recorded here: I owe a tremendous debt to each and every one of them.

Prologue:

I often use a Chess Board metaphor to explain why depth and knowledge and having lost of experiences is necessary to see patterns. I like this metaphor because it helps to explain why people who have lived in a world can connect the dots in that world, while others who have not lived in it will miss the connections. Here is the metaphor:

Experienced chess players, when exposed briefly to a chess game in progress, can then recall the position of all the pieces on the board. They can do so because they know how the game was played to the point of reviewing it and therefore why the pieces came to be where they are. By contrast, novices, given the same exposure to the board, do not know where the pieces should be because they do not know how the game was played to the point of seeing it.

That is also why an experienced manipulator of spreadsheets knows how to find quickly the important numbers in a new spreadsheet or see the errors in the numbers, or why an experienced reader/writer of contracts knows how to read a new contract: he or she knows what should have been done to construct the spreadsheet or write the contract and knows where they should look and how they should work to tease out answers. Similarly, it is why an experienced writer of business plans knows what to look for in reviewing a new one because he or she knows the process that should be followed to write it.

Further, I subscribe to the thought that self-conscious observers are constantly moving back and forth among many levels of abstraction, seeking the parallels and differences by which the world can be explained. Such observers are self-conscious about their processes and know how to get at truths. It is this self-consciousness that has to be exploited and why I write about what I see so I can try to understand it better. This was best expressed by the iconoclastic architect and mathematician Christopher Alexander, in his book “Notes on the Synthesis of Form1”. You will see this book on the bookshelves of computer architects as often as building architects.

Taken together, the chessboard metaphor and the requirement that we all be self conscious, command command that we “Simplify, Organize. Connect the Dots.” This approach is what there are organizing principles like grid systems, relational data bases, checksums, matrices, well organized vocabularies, clean work spaces, etc.

Education

I grew up in the San Fernando Valley of Los Angeles, North Hollywood, and went to good public schools, graduating from North Hollywood High School in 1961. In the Summer of 1960, I was fortunate beyond measure to go to the second summer of the Summer Science Program (SSP), a six-week immersive residential program then at the Thacher School in Ojai, California, founded in the wake of the launch of Sputnik in 1957. Sputnik served as a “wakeup call” for the teaching of science in the United States. SSP was transformative. It was my first exposure to a group of science wonks much smarter than I, and good advance warning of what college would be like.

I went east to MIT for the five years 1961-1966, receiving two bachelor’s degrees — one in Electrical Engineering and Computer Science, and the other in Economics — and a Masters in Management. The rigor, hard work, and exposure to spectacular classmates and teachers in every class, at SSP and then MIT, was incredible.

The seeds of a life-long interest in philanthropy and the opportunity to leave a legacy defined by deeds and money were planted by SSP and MIT, and would be defining for me later in life.

At MIT, I established friendships with fellow students for life: Bob Blumberg, Joe Perkell, Ed Linde, Paul Rudovsky, Dave Mundell, John Beckman, Paul Richman, and many others. I also was able to work with many professors and administrators over time, including Howard Johnson, Rudy Dornbusch, J. Dan Nyhart, Jeff Newton, Angela Belcher, Susan Hockfield, John Reed, Bob Millard, Jon Gorenberg, Whitney Espich, Rob Scott, Eric Grimson, Rory Harrington, and many, many others. I also became lifelong friends with people connected to the Summer Science Program, including students from my class, Steve Cotler, Chuck Holland and Jerry Nelson, and other alumni who stayed in touch with the program over time like Mitch Kapor, Michael Weiss, Ron Irving, Dick Bowden, Amy Barr Mlinar, Doug Duncan, Franklin Antonio, Susan Jerian, and Janice Bishop. When the alumni took over the Program in 1999, I got heavily involved and worked with Dick, Steve, and Chuck, and Dick as Ex Director and Amy Barr Mliner, as CAO to create first an administrative structure and later a Board of Directors. The Board had alumni members like Mitch Kapor, Mike McKay, Ron Irving , Michael Weiss, Doug Duncan, Susan Jerian, Janice Bishop and notable non-alumni members like Susan Taylor, and Jenna Rikkas. Creating and then serving on the Board of Governors until 2023.

First Jobs (1966-1970)

TI in Dallas and OSD in Washington D.C.

My first job out of school was with the Corporate Staff of Texas Instruments (TI) in Dallas (1966-1968). I went from TI to a job in the Office of the Sec of Defense (OSD) just as President Lyndon John was leaving office (1968-1970). Janine and I got married in Dallas in 1968, and we moved to Washington DC, spending our first two weeks as guests of my old office mate from TI.

The four years at TI and the OSD were my exposure to senior staff jobs surrounded by very capable people dealing with central administration, policy making, and information technology and data manipulation. I learned to deal with the intricacies of organizational structure, coordination, staff work, preparation, the importance of artifacts, policies, manuals. There was constantly a move from foreground to background activities, and the reverse. These two jobs were on my resume when I literally ran into Paul Rudovsky in New York City in 1970. Paul gave me an introduction to Citibank, just then emerging from the old line First National City Bank of New York, and the extraordinary management team of “Walt and Bill” —Walter Wriston and Bill Spencer.

There is always a reason each of us supports one philanthropy or another, or many. The reasons might be that we are an alum of a school or once went to a transformational program, that we grew up in a particular church or mosque or synagogue, that we received services from a hospital or social service we want to support, we answered an ad, or we lost a loved one whose memory we want to honor by helping others who had the same experience or loss. That is, we “became philanthropic” because something captured our attention. What each of us does when that happens depends on our interests, means, capabilities and personality, and how the charity or philanthropic organization presents itself to us.

What the philanthropy does with that interest is determined by what it does to develop donors. This post is about that intersection of what happens when people become philanthropic and a philanthropy develops an ask and responds to their interest.

Over time, Janine and I have devoted substantial amounts of time and resources to a number of philanthropic organizations. We did each because we attended schools, participated in programs, signed up as members, accompanied friends in their activities — any number of ways. Working to support each of them, and all of them, has always been rewarding to us in many ways. Because we believe in what we do, we try aggressively to involve others in the same activities.

We have learned a tremendous amount about the process of philanthropic support and development from professional fund raisers and from other like minded, in some cases gifted, volunteers. This post is an attempt to convey what we have learned working with five unique organizations: my alma matter MIT; the Summer Science Program (SSP); the Didi Hirsch Social Services Agency; the Santa Fe Institute (SFI); and the New York Academy of Science (NYAS). The lessons we have learned might apply to any similar situation others may face.

For us, like most such relationships, each case started with exposure through some personal interest or participation or chance encounter. In all, the long-term relationships we established were because the organization was led by high quality, dedicated, knowledgeable people; the organization did good things; and each offered us an opportunity to provide financial or organizational or intellectual assistance based on our capabilities. (My “day jobs” usually afforded me time to spend on these activities, and the people I worked for and with understood that these activities were important to me: not everyone is so lucky.)

For an organization to warrant our support and that of your friends or other contributors, it has to be of high quality. Without that, none of our work would have mattered. We learned from the relationships, and the organizations benefitted from our participation. No amount of fundraising effort, stewardship, or arm twisting of any kind will have any effect unless the philanthropy is run well and benefits the community it intends to support.

We can point to our lasting impact — our “legacy” — on each with pride. Here is what we did in each of the five organizations noted:

• At MIT our legacy includes: starting a funded mental health program1; my active roles in fundraising for twenty years; and my long-standing organizational activities including a manual by which a Corporate-level fundraising organization was managed for a decade.

• At SSP,2 my legacy initially was the establishment of the early governance structure including the first formal Board. I guess now, it is the creation of the endowment that now exceeds $200 million. I had the privilege of stewarding the relationship with the SSP alum Franklin Antonio for two decades that led to the huge gift to SSP from his estate. He gave because the program was great for him and because he believed it was run well. I kept reminding him how important his support was for the Program.

• Janine and I supported Didi Hirsch3 financially and Janine served with distinction on their Board for thirteen years.

• I served as an active member of the boards of the NYAS4 (1985-1995) and the SFI5 (1994-2014) .

I write this post in the hope that readers will follow their own reasons and do some of the same things we did for the organizations they support, perhaps motivated by the stories in this post or facilitated by the lessons we learned.

There are many other stories written by others involved in philanthropic enterprises worth reading, for sure: Bill Brody about his time running Johns Hopkins University6, and Howard Johnson about MIT’s response to the student unrest of the 1960’s.7

Last edited October 18, 2026

Background

I was a STEM geek in high school. Thus, in the summer of 1960, I attended the then recently-organized Summer Science Program, between my junior and senior years of high school. The next year I went east to attend MIT. These two academic experiences drove my subsequent professional life, they provided many of my social contacts over my lifetime, and after I was graduated I stayed involved in development or operations for both. They allowed me to continue working with high school and college friends for decades, and they gave me a chance to leave legacies, of which I am proud.

My job working in the senior staff at Citibank put me in touch with the leaders of the Santa Fe Institute early in its life, and with the New York Academy of Sciences (by then nearing the end of its second century of its life!). The connections in both led to other board and social memberships and to many extremely rewarding professional and long-term personal friendships. We also had the good fortune to go the Aspen Institute summer program on the Great Books in 1977. A resource member of our group later assembled a study group for which two of us (Gus Tyler and I) wrote a document published and widely circulated by the Association of Governing Boards, Trustees in Troubled Times, AGB,1992.8

The Didi Hirsch Social Services Agency provides support for a variety of mental health needs in Los Angeles. We lost our son Daniel to suicide in 1991 and we became involved with Didi Hirsch through a chance social contact after we moved to Los Angeles in 1996. Janine joined the Didi Hirsch Board, and over time we provided significant financial support. The Agency has been run by extraordinary people, and its Board is superb.

To varying degrees, I got involved with the fundraising and governance activities of all of the organizations and participated in their management. This rest of this post is devoted to what we learned over the decades from our work with these five and others.

Characteristics of Good Philanthropies

To be successful at raising funds, all philanthropic organizations have to be run well. The five organizations we supported all met all of the criteria a philanthropy must meet to warrant support, which are:9

1. Satisfies a legitimate need.

2. Organized and governed professionally and managed well.

3. Staffed by competent people who know what they are doing.

4. Maintains an open interactive environment and communicates well with its stakeholders.

5. Truly needs financial, organizational, and material help or resources.

6. Has a good history of performance and communicates that it will continue operating long into the future.

This list is much like a the list of minimum criteria for a startup to attract investors.10

Pillars of Any Development Program11

• The first priority is to create a compelling case statement for support that is 100% backed by management and volunteers. It must be stated in a way that makes sense for the appropriate audience as an opportunity to participate rather than as a charitable need, or as an opportunity offered to the potential donor to leave a legacy.

• The second is the collection of names of people and organizations that might offer support, based on

  • Affinity: graduates, parents, like minded/believers, similar interests

  • Location

  • Capacity and Inclination

• The intentional path: Identification » Cultivation » Solicitation » Stewardship. Thus, development is a continuous, strategic process, not a series of events on an annual. Each campaign has its structure and implementation. There is literally a donor life-cycle loop.

• Setting up an organization and collect the data for prospect management and tracking.

There is no substitute for hard, detailed, persistent, day-by-day attention to detail work in development. The set of steps is well know, but the messages have to be honed, the channels of communication have to be established, the relationships built and nurtured over time. Every development organization and every development campaign is the same and yet different, in ways unique to the time and to the organization. Opportunities appear at their own pace: they have to be created and then seized.

How to Run a Development Organization

It is easy to list pillars or to recite mantras, or to set up checklists. In the end success comes only after a slog of doing the right things over and over, for a long time, until they start to work. If your story is good, and it is presented in the right settings, by the right people, to the right donor candidates, after the sufficient preparation, the work will pay off.

Things you must do to succeed in development include:

1. follow the mantra “If you don’ ask you don’t get.” You must have people in the organization who can ask for money. It is a talent that can be learned or, especially in the case of volunteers, those who have given are often good at communicating how they came to decide to give so their help should be solicited for help. They should steward your key potential supporters.

2. Have a robust story about the origin, management, objective, and the reasons the organization will enjoy success long into the future. This is the “Case Statement” pillar.

3. Collect lists of all possible groups of supporters, segregate the groups by capacity. Work the lists, work the lists. Knowing that “There is a thin line between persistence and harassment,” you have to learn what that line is for each person on your lists, by group. Then you apply pressure using personal visits, telephone calls, emails, group lunches, etc. You use all the ways candidate supporters can be contacted.

4. Have a complete, detailed, list of asks, at every level of giving from small to very large. You never know what will strike the fancy or interest of any particular supporter, and you do not always know their capacity. Never limit yourself to the easy ask: I once noted to a potential donor that the cost of renovating an MIT main-group building was on the order of a $billion. When he responded with a straight face, “Well, I don’t think I could do a whole building,” we raised our expectations for him!12

5. Be organized and thorough. Work the lists, work the lists. Constantly seek support. People who want to help let you know.

   • People who attend events, who write letters, who give advice, or who give money are all candidate supporters

   • So are all alumni, graduates, people who received help in their student days.

   • If they gave $25,000 once, they could give $50,000 or even $500,000 the now.

   • If they support another organization, they may well be willing to support yours.

6. You have to let them become comfortable with you and with your organization. They have to become convinced of the need and of the continuity of existence of what they are being asked to support.

7. Use volunteers to augment professional staff. Since people give to people, not organizations, the professional staff should be connected to volunteers in an orchestrated fashion. The professional staff makes sure the volunteers are supported. Board members have peer-to-peer relationships via other board contacts and local organizations that can open doors staff cannot.

8. Some volunteers are good “scouts” who find potential donors, some are good as “marksmen”, the ones who make the asks.13 The volunteers can develop local personal contacts that can be mobilized if visitors come to town. See Case Study below re CDC Handbook.

9. You can never celebrate and credit all who help too often.

10. Develop all the tools needed for the job:

    • Annual funds

    • Financial Assets: CRUTS, Flip CRUTS, DAFs, etc.

    • Inside the organization: President’s Councils, Alumni Clubs, Legacy Societies.

    • Outside the organization: Foundations, local and national civic organizations, governments, NGOs

    • Major Gift Opportunities

    • Professionally prepared paid lists

    • Create a Handbook if the numbers of volunteers warrants the effort.

Case Studies

MIT CDC Handbook

When John Reed became Chairman of the MIT Corporation he activated its Corporation Development Committee (CDC) and I ended up writing a 35-page Handbook for that Committee, in 2013. The CDC Handbook codified the role of the volunteer Membership, roughly 130 people organized into roughly 14 regions around the world.

The qualification for CDC membership was a commitment to MIT development, demonstrated and transferable skill at grant getting and giving, and willingness to work. (Not all people capable of giving large sums of money are willing to ask, or willing to work, at the process of development.) My work in the Summer Science Program was put to good use and, with John Reed’s support as MIT Chairman, at the top and in Cambridge, the CDC became an active, supportive organization.

The example of Los Angeles was instructive. The geographic area covered is immense. There are something like 5,000 MIT alumni in the area, with concentrations in the South Bay, Pasadena, the West Valley, and DTLA. An evening event might draw people from 60 miles away. The CDC members like me made a point of participating in local Alumni Club activities to meet a variety of age and interest groups. Many of us were active in other philanthropic organizations, another local institutions. We slowly amassed knowledge of local eating and convening sites, etc., so we could help plan social events. And, importantly, in the event of senior’s coming to town, a call from us doubled or better the likelihood that someone would sign up for a dinner or a lunch. We were also on call to answer questions about local relationships, where we knew something.

The person-to-person character of the activity came to an end for the duration of the COVID-19 starting in early 2020, and the CDC is just now being rejuvenated.

Portraits of Resilience Book Distribution

MIT is blessed with some extraordinary people. One is Professor of Computer Science Daniel Jackson. Prof. Jackson happens also to be a master photographer and a great friend of social services at MIT. His has been a dorm family, he is very active in MIT Hillel and its program “Jewish Minds at MIT.”

Over an extended period of time, Prof. Jackson assembled a magnificent portfolio of portraits of students, faculty, and staff at MIT whose lives were affected by mental health trauma of one kind or another, being published as the book Portraits of Resilience.14 Janine and I were asked to support the distribution of a copy of the book to all incoming Freshman, in 2018, as an experiment to see if these stories might motivate others with problems to seek support. Daniel kept me up to date as he worked with the MIT Press on the book’s production. (It was my early introduction to the book publication process and I loved it. 15

The distribution served its purpose of bringing attention to mental health needs. While it was not repeated, the distribution was considered a success and Janine and I were happy.

MIT Mental Health PostDoc

The Portraits of Resilience book project led to a much more ambitious project mental health-based project: The creation of a post-doctorate position as an addition to the 30+ person professional staff of the MIT Mental Health group. The need for the position was identified as part of the MindHandHeart initiative of Chancellor Cynthia Barnhart in 2015. The proposal to us came about through the initiative of Jeri Thys and Rory Harrington, in MIT Development. Janine and I had challenged him to generate an opportunity for us to support something new in Mental Health. As a result of the collaboration, our family funded a three year program. At the end of our three-year grant, MIT chose to continue it as a permanent position. We were happy that our willingness to entertain a new idea generated a creative project and proud of what was done as a result.

Conclusion

I challenge others to be as creative as they pursue their interests in philanthropy and to be as supportive as possible in every philanthropy they support. It is rewarding as can be.

Share

Return to “Of Interest” Website

Your most important financial asset is your ability to earn a living, year after year, typically from a job with a wage or salary, or the income from a profession, or some form of work for hire. That asset should determine your lifestyle. Savings is an additional asset you should have to assure your ability to maintain that lifestyle during short periods when your ability to earn might get interrupted. Your “Savings” should be in an account that earns a steady income, risk-free. That account should be large enough that you will be able to sustain your lifestyle if you get sick and cannot work for whatever reason: illness, when you move from one job to the next, or you decide to take an extended time off.

By contrast, investing is what you do with the rest of your money. It is how you mobilize the money you accumulate beyond what you need to support your lifestyle to make more steady income, from your wealth rather than from your work. With that new source of income you might, for example, improve your lifestyle, or give more to charity, or use it to offset your need to earn a living (salary or other) and fulfill some other lifelong dream.1

Thinking of wealth as a way to change your lifestyle extends the time scale for your investing decisions and forces the discussion into the realms of duration, rates of return, the management of risk, the focus of intellectual energy, and how to make decisions when incentives are complex or information is unevenly distributed or outcomes can be “all over the place.”

In the arena of money earned from a salary or for work for hire, reward is highly correlated to level of professional development and hours expended. When you move to earning money from investing wealth, there are different dollar and time dimensions. and decision-making is episodic and success depends on different kinds of activity. This activity relies more on historical paths, on embedded knowledge, on instinct, on contacts, on learned behavior. The rewards of intelligent investing – money, power, trappings – are different than the rewards from earning a steady salary in a respectable job or profession. It demands time, a different kind of attention, specific new skills, and exposure to different suppliers, vendors, advisors, and sources of information.2

Some of my friends were active investors all their lives, either because of inherited wealth, early investing success, or high current income relative to lifestyle needs. They tended not talk about their investments, and, because it was part-time, their investing was relatively passive. I became a serious investor beginning in 1998 when Citibank and Travelers merged and my stock and bond portfolio became significant in terms of potential earning power and further wealth creation.

After retiring from Citibank in 2000, I began to devote time to managing our wealth. came to treat wealth differently from salary and savings because it was money I was using not to spend but rather to earn more money, to change our lifestyle and to have more impact on the world around us.

The challenge for someone who did not grown up as an investor is to find the time and sources of information and the people to learn enough from to make valid decisions managing investments over time, at reasonable cost. Done well, you can leave a legacy of assets for your family and of good works. I made a lot of mistakes and, fortunately, a number of good decisions. So, I have some things to write about.

The formal word for the aggregation of all that you have accumulated as you invest your wealth is portfolio. The description of a portfolio, and the process by which the investor maximizes the utility of the wealth to his or her needs, is called Portfolion Management. There is no better place to start to understand the process than Andrew Lo’s In Pursuit of the Perfect Portfolio.3

This post describes some things I learned as I accumulated this wealth and how I intend to manage it, focusing on how I made my decisions, and what I did. I will also describe some of my assets. My wealth is 25% in marketable securities, 30% in multifamily real estate, 20% in a long/short hedge fund, 10% is in cash and a short-term Note, 8% in venture investments and 6% in raw land. I am comfortable with that structure and that level of diversification.

Overview of Wealth Management

Managing wealth typically means investing in listed securities, or in Alternative Investments, or real assets. If you want to invest in listed securities there is no better way to do it than to open a Vanguard account and choose among the wide variety Vanguard offers. Use their tools and their funds. Vanguard is a mutual organization and lives by the credos of low cost to investors and plowing profits back into the business. I will not discuss what to invest in beyond observing that most people will have most of their investment wealth in these kinds of assets. Only at levels above $1-3 million should investors broaden out into Alternative Investments, because it is hard to become skilled at investing in that class and it takes a lot of time and costs money to learn what to do. Many of us have an investment in a home, a real asset.

There are many, many ways to invest in Alternative Investments. Many vendors and advisors will tout their expertise in specific areas such as stock picking, private equity, oil and gas investing, highly specialized asset classes, etc. There are 16,000 or more Registered Investment Advisor firms (RIAs), and just under 400,000 individual Investment Adviser Representatives (IARs) ready to give advice for a fee . You should cultivate relationships with a number of them accessible to you. They do a lot of work up front, but expect to earn a continuing fee in the 0.5-1.0 % range. It is always worth learning by talking to them, and some earn their fee and are worth engaging on an on ongoing basis. Most will have a limited list of Alternative Investments to tout. It is impossible for all of them to know all that exist. Most choose a small number and try to put all their clients in them.

If you decide to investigate Alternative Investments on your own, some of your early losses will be the cost of your education in that kind of investing.

Investment Opportunities

The scope of investing possibilities is vast. Retirement fund assets alone are $50 trillion, directly held securities another $50 trillion, and investment real estate another $20-25 trillion. That is an infinite variety.

These assets in turn are offered in a variety of forms by intermediaries ranging from the $ trillion scale of a Vanguard, a Fidelity, BlackRock (the world's largest public asset manager, for index funds and ETFs with over $12 trillion in assets) and Blackstone (the world's largest Alternative Assets manager, for private equity and real estate), and Morgan Stanley, all the way to local investments offered by or touted by independent advisors, newsletters, and websites/Substack writers. And the cost of getting access to these assets ranges from 1/10’s of a % to 3-10%, depending on type and scale.

It would be fair to say that there is an explosion of good (and bad) information becoming available. The slow retreat of the big commercial banks and the rise of the RIA profession and the tools available on websites and from Substack has meant that the individual investor can get access to a tremendous range of information. The difficulty is to wade through the clutter.

How to Gain Investment Experience

I now invest aggressively only after getting involved on a small scale to learn what I can about the people and the ecosystem involved in the business opportunity. This is possible especially in the case of venture and partnership investing, specifically in real estate and oil and gas. If the organization did what it said it would do, and the projects proceeded like the promoters or management promised, I was willing to consider adding to my small early investment and go in on a bigger scale. Only after I watched management deal with success and with problems could I judge how the company was doing and scrutinize their numbers. In some cases it took years to watch projects mature and management to learn their lessons. It is a contact sport and there are many game stages. My biggest failures were when I jumped in too aggressively before I knew how the game was played, or how it moved from stage to stage, or how skilled and how knowledgeable, and how honest or dishonest the players were.

Most mantras on what drives success in investing are a variant on “The Five M’s: Management (people), Market (product), Model (plan, idea, ecosystem), Money, and Momentum.4 Wherever you start. you will have to deal with all five “M’s” before the company is on its way. And the environment constantly changes, on multiple dimensions and timescales, so no plan can be carried out without constant reassessment. And learning has to be constant. Know the people, know the story, know the ecosystem, know when you are hearing something that is new to you and maybe not known to many.5

There is a tremendous problem of finding reliable, actionable information. While the AI tools are making the process easier and easier, it is hard to have reliable information formatted in a way that allows you to compare assets of comparable risk, over finite lifetimes. Harshly, it is a hard to know what you are getting into. Starting small forces diversification and limits up front losses.

I became more disciplined when I began investing the money we made from Dryad Communications and Cloud Cover Media. I had learned what I could about the Registered Investment Advisor world (RIA) world by interviewing a number of them, but pretty much invested on my own. There is a lot of valuable talent available and likely I should have used them more.

The farther away you are from the specifics, the less you have to know but the more you are dependent on others, and the more return you will have to give away to those between you and the investment’s source of income. To paraphrase Aristotle, “The journey is not really interesting, but the conclusions and lessons learned are.”

Participating: Investing As A Contact Sport

You will have to decide how much you want to take responsibility for your investments, and you have to decide how far from the basic business the investment has to be. The closer the investment is to the specifics of the business, the more you have to know about the business rather than only the investment. The farther away you are from the specifics, the less you have to know but the more you are dependent on others, and the more return you will have to give away to those between you and the investment’s source of income.

Getting good advisors is not necessarily expensive and it can save a lot of heartaches. While it is extremely hard to find disinterested, competent, honest advisors, when you do, they are likely to be worth their cost, for at least the portion of your wealth that is outside your area of expertise.

Over time, you will have to get involved in the details of almost every alternative investment. You will learn how to read contracts, how product development is carried out, what an ecosystem is, how models are constructed, who the experts are and how good they are, what conferences to go to, what business lifecycles are and how they are managed. (See also my Mantras.)

“Investing is a Contact Sport” absolutely applies to venture capital/early-stage companies and partnership investing. On one extreme is the approach taken by the manager of the MIT endowment, where he has been cultivating access to fund managers for decades and has said often, “We try to be the first institutional investor served by the fund.” That is, he is heavily dependent on those funds for their expertise. On the other extreme is to get involved in the company, to a greater or lesser degree. I have done better when I am hands on in the company, and you can only do that one at a time.

Visibility and Accounting

If even one of your investments throws off a K-1, you have to be familiar with the accounting for and taxation of Partnerships. If any of your investments are reported as multiple K-1s then you are in trouble because at that point you are likely to being taken advantage of: there are too many decisions being made at the line item detail in the balance sheet of each deal. At scale, your investment manager should be operating as a fund, not a set of K-1 investments — see below.

When you cede agency to deal managers there are not enough accounting conventions in VC investing or partnership investing to protect the passive investor. It is too easy for the self-interested managers disadvantaging the investor. Controlling investors and management makes a great many big and small decisions which affect the outcome of the business and the distribution of the spoils, at every step of the process. The only way an investor can protect him- or herself it to be hands on (“close to the investment”). In many cases I became a part-time manager of early stage companies in which I invested where I could add value, to be sure that I was not being taken advantage of, and to know what was happening in the business. That is why I say you should not be investing with a manager who reports a lot of K-1s

Investing Stories and Lessons Learned

Venture Capital

When I left Citibank in 2000 I immediately joined an LA-based Venture Capital company in formation, Centre Palisades. I spent a little more than three years with Centre. In retrospect, I was very poorly prepared for the venture capital world. I learned a lot but essentially broke even on the time spent. (I eventually invested in another Centre Partners fund and it turned a handsome profit so the whole experience ended up okay.)

Managing Small Companies

I then helped clean up a troubled PE funded company PrivaSys which paid off its investors through an IP violation by Mastercard, for which I was paid a handsome hourly rate, and for a short time became the CFO of Lucix, a Centre investment in Camarillo, CA.

In 2006 I had the hands-on experience of running a company Roubini Global Economics/RGE, I learned a lot and the chance to work with Nouriel Roubini and Bill Janeway were well worth the time but we all lost money in the process.

Passive Investing

My mother-in-law in Dallas did some successful investing in raw land in the growth corridor north from Dallas to Oklahoma in the 1980’s. I followed up with her promoter, to slow, steady success in the years after 2005. Raw land is uncomplicated and does not generate non-operating losses, but the promoter takes a big cut at the end (the “carried interest”). It is purely passive investing.

My best passive investment ROI was from a small equity investment and a follow-on option-incented loan to a small private company Triple Canopy, co-founded by a former Citibank colleague Tom Katis. Tom started the company in 2003 after leaving a post-Citi re-up in the US Army Special Forces. Triple Canopy grew dramatically from the start, navigated a number of shoals successfully, and was essentially self funded the whole time. Along the way, Tom eased out a dishonest CEO, put the company back together, and then grew it tremendously. The business model was very profitable, and Triple Canopy was sold after eleven years, in 2014. Since I kept most of my money in the whole time, my return was well above 10X. The lessons here are that people matter, and you can give them time and agency and they can do well for themselves and you.

(Tom followed the sale of Triple Canopy with Voxer, a communications company built on ideas from his Triple Canopy experience. Voxer made a lot of money — ca $175 million — from a patent infringement suit in 2022, after a long legal slog, but that is another story.6)

Raw Land

It is all about the access to land source and access to sales channels. It is where the iron law that you make money when you buy not when you sell is validated. In the case of my investments in Dallas, the company, eventually named the GC (Golden Corridor) Land Management, was buying tracts of land along the Pecos Trail between Oklahoma City and Fort Worth. That storied historical artifact funneled cattle from the grazing lands in Oklahoma south to the railroad yards in Ft. Worth, Texas. The path (the “Golden Corridor”) was hemmed in to the west and east by lakes, rivers, and mountains, making it easy to run the cattle. It also channeled the growth of housing, roads, and businesses as Texas grew. The population growth north from Dallas went along the roads that had followed the Pecos Trail, hemmed in by the same geographical features that channeled the cattle.

There are many fewer tax complications from investing in raw land, and “buy and hold until sell at a profit” is a straight-forward strategy. The promoters took too high a percentage for their carried interest and I have held off investing more. The ROI was in the 12-17% range, which was fine.

Start-ups

My best return on time and invested capital/risk were from two startups I participated in directly that were very successful: First, I co-founded Dryad Communications with Tamra Johnson and its main assets were sold to T-Mobile in March 2015 for a handsome profit.

Soon after, I began investing in, and eventually became part-time CFO of, Cloud Cover Media (CCM), which provided legal background music over the Internet to businesses throughout the country. CCM was sold to Sirius XM in January 2022. CCM returned 7 times the amount of money invested in the company.

In both Dryad and CCM, we learned more and more about what we were doing as the businesses developed. I started small and piled on my investments as the efforts were succeeding. I was very much playing it as a contact sport, getting very close to the investments. The two investments provided me some wealth I have been able to invest. These two startups are also the subject of other of my posts about how to use financial control in successful small companies.

The SaaS model is at risk with AI, and there is a tremendous amount or PE money tied up in unsold IPOs at the moment, so there is substantial risk in the business now as well as somewhat more limited upside but is still the best.

Real Estate: Deals, Funds, REITS

Lately I have spent all my time organizing my investments in real estate. I began investing in multifamily real estate in about 2005 with a few small-scale brokers and promoters who invested in specific buildings in Los Angeles and Arizona. Those early deals worked out but the promoters seemed not to find deals on a regular basis. Beginning in about 2018 I started to build a portfolio of individual partnerships that were local, one-off deals, assembled by an accounting firm in Chicago, SMB. I stopped adding to that portfolio after about six years because the firm was not learning and was not getting better over time, and the deals seemed never to come to completion.

I came to mistrust the process by an investment firm creates on an ever-growing number of K-1 reporting partnerships. In this situation, the GP is trading off his interests against yours every minute of every day. He or she cannot be honest every time. And your capital is returned in the early years so your earning capacity is steadily eroded. What looks like cash income is, rather, invested capital being returned, and your earning capacity declines. More importantly, if you are building up non-operating losses you are not getting the tax advantages the asset is generating. The returns have not been what I expected. If you are not tracking your tax returns in great detail, you are probably building up non-operating losses like I was.

I am currently investing in RE through a set of well defined funds that comprise multifamily housing projects in the Sunbelt, managed by Origin Investments. While the underlying assets in the Origin funds were superficially similar to those of SMB, the investment strategy and the fund management philosophies of Origin are much more disciplined and better for the individual investor than SMB has been.

Origin’s ability to blend investments to manage tax efficiency is a key feature of their funds and is what makes them attractive. But more to the point, their disciplined process got better and better as they honed it over time, and they have done better and better for themselves and for their investors. In the process, they have moved the investor further and further away from the assets in the funds. Their process is detailed and allows them to assess assets in a much more careful fashion. The assets have proved to offer steady, low risk investment returns.

From these experiences, I concluded that unless the entity and the management is getting smarter at every turn, the returns of successive investments or funds offered begin to suffer. In my real estate investing experience, someone who keeps doing the same thing over and over is likely to lose out to someone who learns how to aggregate and build on successes.

The problem is that the smart people move on, or the high return investments attract more investors than the manager can handle, or the manager gets greedy. It seems that when the organization is not doing bigger and smarter things, or not finding ways to earn more from the prior successes, they start to fail. My investing with operators of real estate projects where each is operated as a standalone project have not run to completion on a regular basis so the jury remains out, after many years.

In the case of Origin Investments, they were partnered from early on with Kovitz and then Focus Financial Partners rolled the combined partnership into their huge network of independent RIAs and wealth management firms. In turn, Focus is privately owned by investment vehicles affiliated with the private equity firms Clayton, Dubilier & Rice (CD&R) and Stone Point Capital.

REITs were created to institutionalize this process. While the process may be honest or not, hopefully that comes out in the results. Since self-interest drives all decisions, even the best of people cross the ethical line now and then. If you are dependent on individuals making the right decisions on their own every time, you will lose some of the time. It is NOT good to invest in an operation where there are lots of individual deals (think real estate projects, oil wells), so REITs are a good alternative if you want to invest in RE.

Indexes and Funds

There is little to distinguish the logic behind constructing an Index from the logic behind constructing a Mutual Fund, and they co-evolve. After all, an Index is a construct that attempts to give a metric to a group of companies at a point in time and then track that metric over time. The construction involves choosing companies to be included and how to weigh each and how to normalize for something like stock price or earnings potential. It also involves rolling out old companies and in new ones. The same decisions that would be involved if you were setting up a fund. Thus my Mantra, “Designing a fund is the same as designing an index.”7

If you invest in ETFs or Index Funds, you have to read Rob Arnott’s book, The Fundamental Index.8 In it he describes how he moved away from indexes based on stock price to indexes determined by fundamental measures of value like revenue, earnings, dividends, or some combination. He has very strong arguments supporting his thesis that using price to drive an index exacerbates swings between value and growth stock periods. At this writing, we may be seeing the turnaround based on questions about the AI boom.

You should need no more incentive to at least consider his approach than to know he sold the rights to the index and his back office for a half a billion dollars. Good return on writing a good book.

Oil and Gas

I have so far lost a lot of money investing in oil and gas. I do not recommend it to the faint of heart. T. Boone Pickens made his money on the floor of stock exchanges not in the oil field. And he had degrees in petroleum engineering.

The general rule that “Investing Is a Contact Sport” is certainly true in oil. In oil, the mantra becomes “Get As Close to the Well as Possible.” The big returns come from knowing what is going on at the drilling site. There is 3,000 to 11,000 feet of rock and sand separating the operator from the oil and gas he is going after. No amount of technology can tell him EVERYTHING that is happening at the bottom of the well.

I have read four or five books on the geology of oil fields so I can understand the discussions of lithology, fracking, drilling, etc. etc. But nothing prepares you for the vagaries of the actual oil field environment. And there is no better popular book on the subject of the vagaries of the legal underpinning of oil revenue flows than The Last Trial of T. Boone Pickens.9

The risks and rewards are very big. The most important thing to know is that the initial investment in any given site is only the beginning so do not start big in any oil and gas investment. Start very small. “What the mud log gives you, the e-log takes away.”10

Fortunately, our last well was successful so the story is still unfolding.

General Lessons

Domain: Pick an area you know something about.

You are never smart enough to control any outcome but you should never invest in an Alternate Investment you know so little about that you cannot judge the value or accuracy or completeness of what you are told. You may be fooling yourself.

“Don’t take hostages you are not willing to shoot”

The protections you build into deals with friends, or the penalties for failure to perform in loans to companies you want to help, never work. Where you are committed to continuing the activity, protections tend to fall apart when crossed. Specifically, any simple loan or other short-term obligation that might not get paid must have equity conversion coverage, arranged up front.

A counter-factual example of this issue is a discussion of the medical practice of Senator Marshall of Kansas, where the NYTimes detailed the results of over 700 patient suits that sent 81 of them to prison. That is what you have to sign up for if you are willing to shoot the hostage you have.11

What to do when a Borrower Defaults or a Customer Fails to Pay

There is a lot to learn about how to deal with an asset that goes bad. This includes a debt that has not been repaid, a vendor obligation that is unfulfilled, or when any obligation by a counterparty to an agreement with you fails to perform. In short, when you have an asset in “default.”12

The steps are as follows:

• Take steps to preserve your rights. Do not waive your rights or establish a pattern of by forbearance by inaction.

• Document your position as to the nature of the default as fully as possible, referencing signed documents as available. If there was a monetary penalty, like 10% of the missed payment, it has to be assessed.

• Explicitly state any rights you are not waiving, and include the right to identify further defaults should they occur or be found.

• Set penalties for any rights you forebear. Payment can be in the form of added interest or fee payment or other economic cost or risk reduction. Deal with the value and cost of any such forbearance explicitly as to time, extent, or whatever you feel is reasonable. Reserve any rights you are not willing to forbear. Write a forbearance agreement and get signatures by both parties. If appropriate, revise the document controlling the relationship to define the new one.

• If there is a risk that the asset will lose value because the company may never repay, an equity component should be added to any forbearance agreement.

• Consider the tax implications of steps being taken for both parties. Be sure neither party receives a tax break of no value to them nor incurs a tax exposure they did not have prior.

• Do not renew any agreement with any default outstanding before dealing with the issue.

• Add new information requirements if monitoring the situation requires new information, like financial statements in a specific level of detail. You must have enough data to assess payment likelihood, and to support tax treatment.

• Always look at any new obligation or agreement through the lens of this list.

Bibliography and Notes

There are a great many books about starting and running a business as well as a great many about managing wealth.13

I have spent my professional life designing and operating financial reporting and control systems, managing organizations, investing, and supporting philanthropies. I have often written about what I do as a way to organize my thoughts and then get things done. Now I want to write to share what I have learned with others.

I am creating and publishing a body of written material, first as a series of articles (or “posts”) on Substack and then perhaps as one or more books. This post describes what I learned in the processes. It looks at the tools available, the technology involved, and the sociology involved in publishing.

Publishing

I became interested in publishing as a discipline while I helped Janine get her family memoirs book Dear Oliver out the door.1 (It was a fourteen-year project for her, by the way). Then, I read Ben Smith’s book Traffic (2023),2 an eye-opening account of the early history of social media and social networks and their news offshoots, from an insider, newsman’s, perspective; and Graydon Carter’s 2025 memoir, When the Going Was Good,3 which described Carter’s career in magazine publishing and his very successful 25-year run as the Managing Editor of Vanity Fair. The Art of Making Magazines4 in The Columbia Graduate School of Journalism series about publishing magazines specifically filled in many of the details Carter only hinted at —and that, too, was a kick to read.

To write and make available a body of work, I needed a writing, editing, and publishing platform. I wanted to assemble the material as documents; then make the documents available for review and modification; and, ultimately, organize the documents as published pieces, perhaps as one or more books.

As I have been actually doing it on Substack, I find that having all the individual posts available and “up on my screen” continuously is like having a book in draft form, available at any time. Substack’s ability to render my posts as a website allows me to have a permanent but continuously changing substitute for the books that I might otherwise published. Selling an annual subscription to my website is like offering my readers a continuous window into my thought process. And, if Substack one day offers to render my posts as a book, so much the better.

Because the process of writing, editing, and publishing is complex and can be built on a flexible technology base, there is a wide variety of online writing platforms and other applications designed for would-be writers like me. There is also a wide variety of social networks for communicating and sharing what we write in real-time, each taking distinctly forms and having different personalities.

I reviewed the platforms and tools available to understand what they do, how they came about, and how they are evolving, and made some choices. This post describes what I learned, what I chose, and why. I go into detail about Substack, my choice at least for now, below. I explain what I want it to do, how I will use it, and how I make it work for me because I suspect others may be facing the same questions and may find this discussion of how to publish helpful.

Last updated on October 19, 2026

Capabilities and Affordances

This post ends up looking at the forces shaping the technology and the social and financial needs to which social media and social networking organizations are responding.

In the process I came to understand how and why the tools available to write and publish are where they are now by pursuing the issues of capabilities and affordances, and cognition. I came to see that Substack has the capability to render my posts into a website (as do a number of other tools), and I can create links between my posts that may be a substitute for my publishing them as a book, and from these posts I can create a newsletter and a website. These are affordances that Substack offers me. I go into this in detail below and likely in a longer post on the subject.

Process: Why I write “Iteratively, interactively, continuously, and in parallel”

I write to organize my thoughts and what emerges over time becomes my Substack posts. The posts are rendered by Substack as my website (and I can vary the form at whim).

I try to connect dots of what is happening on a big landscape. My writing process is to knead the words until they communicate the thoughts I want to share.5 The process brings to the surface what I have to learn, or have to develop. Juna Kuhlmeier describes the process this way: “It is generally helpful to write as a way of thinking, so this essay is part of that intellectual workflow.”6

In the past, that meant writing a Word document and sending it around for review and use. In yesterday’s world I might have tried using Twitter, Facebook, or LinkedIn to make the interactions happen, but archiving is an afterthought for most social networking tools and I wanted something that would leave the posts in place as a legacy. I could have set up my own personal website and fed it by an email list, but creating an email list and keeping it up to date is hard work, and using it to get the work out can exhaust one’s friends.

Today’s information-rich on-line world makes it easy to gather material together while one writes because AI gives you access to a tremendous variety of sources that you can fact check, do research, and organize material on the fly while you write, 7and the publishing tools are so effective, that writing has become more of a parallel, continuous, multi-path, process. Documents are easy to structure by working iteratively, continuously, openly, and in real-time. Finished posts, notes, or even books, settle out from the process quite naturally.

It turns out that many of the publishing platforms support this approach to authoring if that is what the writer wants.

Platforms: the Internet Publishing Infrastructure

The technology, the psychology, and the social science underlying social networks and social media and publishing platforms and communication tools and online databases is broad, deep, and rich. Each discipline has its academic institutions, culture, suites of vendors, and its own roomful of writers. Taken together, they provide the publishing infrastructure with a constant flow of nutrients for change and growth.

I had long been a low-level user of social media (content) like YouTube and Google, and social networks (sharing) like LinkedIn and Twitter, and the hybrids like Facebook. I found each wanting from one viewpoint or another as a place for my writing, mostly because they offered little support for authoring tools and the process of writing and the needs of archiving, and their business models were changing, on the way to becoming huge machines to profit from advertising and other applications. They were becoming bad repositories for any work, and not a place to leave a legacy.

I considered writing platforms like Medium or Project Syndicate but they were set up for professional writers with a reputation who could generate traffic based on their name, which I am not. I could have set up a website but I did not need the breadth of publishing tools websites need to support a broad range of users nor the e-commerce platform support an e-commerce business model most websites need. Finally, I considered WordPress, but that is a big ecosystem by itself and, as a standalone writer, it would have been a big commitment to make it work.

Most social networks are set up to allow their members to get access to other like-minded or similarly inclined people to share ideas. This was the primary initial targets for Facebook, for sure, where Mark Zuckerberg put the picture book of his classmates at Harvard online and got them, famously, to declare whether they were or were not currently ‘in a relationship.” Over time, the real “customers’‘ of social networks often became not the members but rather the stakeholders who provided the social network with revenue, the ones who had something so sell or a job to fill. This “mission creep” has allowed the emergence of behemoths like Facebook and Google.

If, in the beginning, a member of LinkedIn was a professional seeking to interact with other professionals in his or her profession, their aggregation became attractive to other stakeholders, largely advertisers who wanted the members’ attention, vendors who wanted to sell members something, or potential employers or their agents looking to hire them. Specifically, once Google had millions of people looking for things in its search engine, or LinkedIn had a critical mass of members describing their job credential, they were able to sell access to their members to advertisers or people with jobs to fill seeking candidates to fill them. Their value became immense.

So, the focus of social networks and the forces shaping them shifted over time. The social network or platform’s evolution over time was dictated by its architectural design and how well its capabilities supported its users and satisfied its other stakeholders’ needs.

Each social network has its stakeholders, which can include a wide and disparate variety of groups: funders, owners, and creators; and writers, readers, and programmers; and the people or businesses or organizations who want to get access to the users to do something else or learn something else: and to earn money. They include anyone who signs up to read what others write, and writers.

Oher stakeholders are technology vendors seeking a place in the social networks’ technology “stack”, and regulators looking at the activities of the social networks’ members and the network itself. And of course there are advertisers, attention seekers, and so forth.

The most important social network stakeholders are its “users” or “members” (college students, job seekers, writers, etc.), or “writers” and “readers.”

Finally, each social network has its own business model, and its unique way it to get its tools and capabilities into the hands of its users.

Platform Capabilities…

Interest and success for most stakeholders is determined by what functionality the social network or platform offers —that is, what its capabilities are.

Many of the technological building blocks for publishing are shared among the platforms. Every social network and writing platform has chosen a particular collection of technology, organized in a “stack,” and presented to the stakeholders. This unique collection comprise the platform’s capabilities. This collection of tools and capabilities is a major part of the social network’s DNA, its basic building blocks. The technology DNA of each social network dictates how it evolves, sets its limitations, and, finally, allows it to adapt and thrive or causes it to wither and eventually disappear. 8

Note that capabilities are what the social networking or media application offers, from its perspective. It includes text manipulation needs, rendering needs, ecommerce needs, possibly footnoting or bibliographical needs, mapping, search, etc. Many of the needs are visible to the user, many are not. The way the application makes money may be explicit, like taking a percentage of subscription fees; or it may not, like learning interests and preferences of readers for sale to others. In the former, the interests of the user and the provider are aligned, in the latter they are likely not.

…and Affordances

If capabilities are what the platform offers, then affordances are what the stakeholders want. The word affordance communicates the sum total of what a user seeks from a social network or writing platform. An affordance is how a tool can be used by a user and what actions are possible for the user if he or she has access to the tool.

The idea of affordances was used by Don Norman in his books about physical things from 1975 onward.9 David Robinson, when he quit OpenAI recently, mis-used the word affordance in his Ezra Klein interview, equating it to capabilities.10

At about the time of cute dog pictures and other early artifacts, Clay Shirky started writing about the rise of the Internet as a confluence of changing social activity, lowered transaction costs, and ways of organization. He used the word affordance correctly as it applied to the way social media and social networks evolved and how they in turn affected organization development and other things of interest to social scientists in his ground breaking 2008 book Here Comes Everybody.11 (Before that, he was part of an early incubator, called the Accelerator Group.12)

Social scientist Zeynep Turfekci built on Shirky’s work (and credited him) as she looked at the tools used for organization and communication in the revolutions that shook the Middle East in the Arab Spring of 2010-2012 in her 2017 book Twitter and Tear Gas. She defines an affordance as “the action a given technology facilitates or makes possible.”13

An example of how capabilities and affordances interact is the ambiguity of using a system where you are both a user and an administrator. App developers have settled on the standard of using tabs on the right hand side (RHS) of the screen for users and tabs on the left hand side (LHS) for administrators. Substack offers authors the same screen for editing as that seen by readers EXCEPT the author screen has the three-line hamburger in the upper right hand corner of the editor screen that is absent in the reader screen

I will expand at length on the story of capabilities and affordances in a separate post.

Publishing Platform History

Writing and other platforms followed the same pattern as those in publishing. Capabilities are offered and affordances are intertwined by stakeholders. Since the stakeholders could have contradictory needs and and the affordances were not necessarily consciously designed in at the start, it took a while for the social networking and platform worlds to sort out.

The variety of platforms became very broad. Philanthropic and equity crowd sourcing platforms appeared, cap table back offices, monthly invoicing services back offices, on-line document authentication services, etc etc etc. Some things that appeared worked just well enough that the site would gain traction early but then lose out to a copycat with more capabilities, or a better sales story, or better distribution. Some things that appeared found no traction, some were needed but emerged slowly. And some of the trends shifted and caused the site or platform to decline. Many once-popular social networks and platforms are now gone.

It was a complex dance between the capabilities inherent in the application or system and the affordances the users needed and wanted and learned how to use. If the dance was not synchronized, the site, platform, or product failed.

The range of participants is constantly in flux, with new entrants emerge regularly. Well-entrenched players often fall into decline. Business models change.

Examples of this process of continuous change include:

• GeoCities and MySpace were very early small, simple websites for non-techies, and each saw its life cycle play out and then disappeared. Now Facebook (Meta) may be going through the same cycle, albeit at a much larger scale.

• Huffington, Drudge, Breitbart, cute dog pictures, the gold/green dress, Gawker, Buzzfeed were all chronicled by Ben Smith, or any of a slew of books about the transformation of the NY Times. It is amazing to read how simple that world was in comparison to what has emerged now. Smith does a good job introducing the issue of how the NY Time shifted its direction in order to survive, but dozens of books and hundreds of doctoral theses have been written about it.

• Twitter started as shared, short messages, with its growth driven by algorithmic propagation. It grew into a huge, overburdened behemoth until it was taken private in a misbegotten purchase by Elon Musk. Musk stripped it of much of its excess staff, both technical and content, and it is now about 60% smaller, with a faster-reacting tech staff putting our more new things but at much lower revenue, advertising, and usage levels.

• Google began as a search engine, with Alta Vista as its competition. Its algorithm was superior and its sparse style was unusual and successful. Once it achieved dominance in search (and now AI) it became a hugely profitable machine using an advertising revenue model appended to its search engine once it reached scale. The advertising model was managed carefully enough that it has not overwhelmed the site’s usefulness, and its dominance of advertising and ad publishing. It is a perennial target of antitrust investigations but it shows no sign of slowing down.

• LinkedIn gave access to professionals via job histories. Now, it is being overrun with advertising which may overwhelm the interactions with friends so the site’s usefulness seems to be declining.

• Reddit has individuals flitting in and out, but there is a corpus of subject mater curators and posts that persist. It evolves constantly and without intervention and self-regulates and can be mined so it has value and seems to thrive.

• Amazon upended first book selling and then the entire retail sales environment, forcing all retailers from bricks and mortar to hybrid or fully to on-line sales.

• First the outlaw music providers and then Apple totally transformed the music business, and legal background music appeared, delivered by the Internet instead of iPods.

• The whole environment is moving to AI-based systems, as fast as data center construction allows.

Substack

Substack and newsletters seemed to co-evolve in an accident of timing — Substack came along just as the ad-based revenue stream of the majority of print-based vehicles which writers shared with advertising was moving en masse to web-based information distribution apps like Craig’sList, or Google, or social networks like Facebook and LinkedIn, which had turned themselves into advertising platforms. This migration put writers out of work. It was a dramatic restructuring of how people got news and found suppliers in the United States and around the world.

This following chart tells that story by looking at newspaper advertising revenue in the 20th century as it rose and fell — the data stops in 2013 because it is no longer being collected!:

As print media faltered, writers were laid off and eagerly flocked to Substack because it focused on the kind of publications writers who might otherwise write for newspapers or magazines. It has thrived while independent, magazines and newspapers have withered.14

Today, Substack is rapidly coalescing around an ever expanding set of newsletter-format publications that are bringing like-minded writers together and are looking more and more like the magazines they are replacing. Some of the Substack-based newsletters have an organizing ideology which needs a voice, with a collection of high profile writers and attention-grabbing names like The Free Press, Bulwark, or The Contrarian. And newspapers like the Guardian and magazines like the New Yorker are starting to create Substack-based offshoots as they move more and more of their energy to the web, transporting their voice and style. (You can see the Substack influence in the design of the Wall Street Journal’s newish online version.)

While that is neither good nor bad, it is not a guarantee that each will live and thrive nor that their writers will find sufficient income. The writers are all competing for the same limited attention span of the readers, and all-too-often occupy the same echo chamber as their peer publications. It is hard to get and hold an audience. See below for details. The loss of curation that is a hallmark of published periodicals causes originality to suffer in this kind of environment.15

As a side note, there have been many other standalone aggregation sites, who have just not offered a broad enough menu of affordances. There is a site and organization called The Conversation which uses WordPress and a creative commons license to allow millions of academic authors to republish articles and achieve broad circulation.16

This is not to say that alternatives to Substack like Medium and Project Syndicate will not survive. The issue is whether and how a publication provides curation. Medium and Project Syndicate are essentially invited writers with an expertise recognized by people willing to pay for access to them. The writers generally have an independent standing and some name recognition. The publishing platform provides value from curation and their ability to concentrate interest in a subject, which is their “network effect.” Medium is apparently surviving by generating enough revenue for its good writers that is will survive. Project Syndication’s hybrid may be the solution.

Over time, readers may find the Substack world filled with too many individual writers vying for the same audience, so readership will remain too diluted. Medium and Policy Syndicate seems to have enough high quality writers to attract attention on their own, based on the quality of the writers. The Conversation may move to Substack as well.

I mentioned the New Yorker has an active Substack site, and other magazines may as well. The New York Times, by moving very early may have recreated itself early, achieving critical mass independently early enough and for a wide range of affordances that it will survive on its own, at scale. So may a few others, but it may be that utilities like Substack will provide the infrastructure for the few publishing survivors. If the rules of AI dangers apply here, it may be that outsourcing the Social Media publishing function to Substack may cripple a magazine. The jury is still out on that issue.

Keep in mind that blogs, posts, social media, social networks, wikis, author platforms, publishing platforms, etc., are all infrastructure forms satisfying a need that were previously provided by a more physical work analog like flysheets, newspapers, newsletters, emails, pamphlets, books, encyclopedias. There are reasons that magazines took the form they did and attracted the writers they did. The same with newspapers, and books. Each was an art form all to itself, and the needs of readers drive these forms so the online world usually conforms to the same needs as the analog world, so their evolution can be predicted.

The Allure of Substack

I settled on Substack, at least for now, for a number of reasons which I outline here, and I am pleased for the time being. Using Substack, I can create posts of about 3000 or so words (about a chapter if I were writing a book), using Substack’s MarkDown-based authoring tool. Substack seems to allow authoring to happen in pieces produced in parallel. Once written down, an idea may begin to grow and evolve. In Juna

Substack has been designed to give a broad range of writers a very flexible means to create posts and newsletters while also getting them visibility and a chance to earn money. It has four or five basic templates by which we authors can arrange our posts, with each template basically mirroring what a reader expects from a website, a newsletter, a magazine, a blog, or a document repository.

The template most Substack users want focuses on articles or posts and renders on a browser or iPhone as a newsletter. The template is very rich and is what most people think of when they think of Substack. Bari Weiss/The Free Press has a rich newsletter format, which Substack uses to render the Free Press’ content as a periodic newsletter.

Another Template focuses on notes, which is rendered to look much like a blog.

The newspaper template I use renders my subdomain henrylichstein1 very much like a simple website. All that I write is on my subdomain (henrylichstein1) of the domain substack.com. If my readers access my subdomain, I have told Substack to render it as a website that looks like a newspaper, which has a iconic format.

I am over-simplifying perhaps but that is what I see.

Substack affords me an ecommerce platform by which I extract some money from my more generous friends, so I have no need for the affordances of Shopify because I am selling only a small variety of things. If I get more syndicated writers, that might change.

I use Substack as a place to house my standalone posts, at least for now. I am not yet to the point where I am worried that a best-selling book is languishing because it is hidden away from the Google SEO engines. One day I will revisit that issue. There are indeed tools outside of Substack that allow an author to take Substack posts and notes and assemble a long-form document like a book.17 They depend on the features of of Extensible Markup Languages (XML). (Readers old enough to have used WordPerfect remember editing the commands in parentheses, which was its XML, even if you did not know that’s what you were doing.)

Whereto for Substack?

Founded in 2017, Substack had 1 million users by November of 2021, 2 million by February of 2023, 4 million by August of 2024, and 5 million in March 2025. It received a $100 million investment in July 2025, at a valuation of $1.1 billion. 18Those kinds of investments presage change. There is evidence of advertisements arriving. “Notes” are evolving into Twitter-like instruments of forced interaction. The discovery algorithms are being tweaked. Hamish McKenzie’s book will be published on November 17th and it will be interesting to see what he says about its evolution.

It is easy to see why Substack has grown so fast. Its ease of use is extraordinary, and its revenue model in the beginning depended completely on writers generating revenue so Substack’s interests are aligned with its writers, up to a point. It does everything possible to help the writer get words on posts, and posts out to subscribers.

Furthermore, its algorithmic distribution is built on links to readers for autonomous expansion of writer subscriptions, at least in theory. Substack should get a wider exposure for me than I could ever could on my own, at lower cost. If Subscribers share what I write to other Substack readers, or to readers outside Substack, they will bring me new Readers. In theory, I should have a much bigger audience by publishing on Substack. In sum, Substack should afford me what I want as a solo writer. And as Substack readers share Substack writers posts to readers outside Substack, the total of readers will grow.

It turns out that many of the features of Substack involving selling any one author to all the readers on Substack make all the writers on Substack competitors of each other. It makes all the members compete with each other, with the Substack platform the only winner of the game. It is a bit like Amazon showing competing sellers to the customers of each seller when the customer selects a product of any seller.

So, in practice the Substack algorithm may not work out as well as writers hope. Another reality is there is an echo chamber with many writers saying the same thing. Finally, there a growth of writers as well as readers, and one senses that a lot of writers are publishing and waiting for readers, which is lonely.

There is no shortage of Substack sites offering advice on how to use Substack and how to expand your readership on Substack. Most writers’ rate of paid subscribers top out with only 3-4% of total subscribers paying (the rest are free), so the platform does not provide a living wage. A favorite topic for writers on Substack is messages to other Substack writers about how to expand readership and to drive conversions to paid subscriber status. They tend toward exhortations to comment and like other authors and make meaningful comments, etc. Very much like how participants establish status on Reddit.

The issue of Substack vs. WordPress is explored in a post by Martin Dubovic.19 He notes that WordPress powers 40% of websites. I see that Clay Shirky leaves his material on a WordPress-based standalone Website.20 The problem is that WordPress is an ecosystem with each added function accreted by installing a WordPress extension.

And the internal formatting undertaken by Substack is antithetical to Google in native mode, so each author has to learn tricks that makes content visible to SEOs and to Google search. If a Substack author wants to appear on “organic” searches that originate outside Substack, he or she have to become a techie for a week, or hire someone who knows what I just said means and how to implement the changes.

And Whereto Lichstein’s “Of Interest” Newsletter?

So, for now, I will stay with Substack as my writing platform and using its authoring tools, simple website, and archiving capabilities are good enough for now. My posts are available to all my Substack readers, whether subscribers or people who find me organically or to whom I send to the site by a link. I need no more e-commerce capabilities than I have now.

I am like many independent Substack writers for whom the promise of Substack as a large scale platform is at risk at the moment. I will not escape the issues of writer saturation, talking in a echo chamber, changing discovery algorithms, and reader exhaustion.

I started all of this to grow my readership to interact with more people and that is still attractive. So, while the promise of autonomous subscriber growth via the Substack algorithm was attractive, its failure will not bother me but I will have to work to open my posts to Google more aggressively and build up my owned email lists.

I am not about to go through the process of setting up a WordPress website.. So, for My latest post drafts and soon other documents will be available to all my Substack readers, whether they are subscribers or people who find me organically or who I send to the site by a link.

A Typology of Social Networking

Posts do not sell themselves: Substack generalizes this by alerting authors that only 1% of posts generate traffic if they are not accompanied by an email announcing their publication, and swarm investing requires the funds seeker to find his or her own “marks”.

But neither do books sell themselves. Even the most popular authors have to hustle to sell their books.21 So, I may bother readers with occasional emails. A way to think about this is a typology of the different ways for a writer to establish a relationship with his or her readers,

Eric Schwartzman22 described the social networking activity by which one earns the trust and custom of other people on the Internet in his 2021 book Digital Pivot, after years as a successful SEO consultant. He describes the typology of link classifications as follows:

• Owned: An email list you can use to communicate with your contacts directly.

• Shared: The links you have as a member of a social network like LinkedIn or Twitter or Facebook, which you share with the other stakeholders associated with the social network.

• Earned: Where another author or website refers to you so you earn the support implicitly or explicitly of another actor on the online state, and

• Paid: Where you pay another site for exposure of some kind.

It would appear that all successful affordance solutions will end up in some variation of that classification scheme. Any one user will become a hybrid user of all four.

Alternative Book Formats

My meanderings caused me to focus on book formats. There are the classic form of novel and non-fiction textbook, of trade book. There is a FettScript, 23which is an unabashed first person person appreciation of another person, usually later in the subjects life. I was struck by the fact that Daniel Jackson split his book on software design to a first half where he described his view of well-structured code and the second half to his Notes, and he said his readers could read them independently: I mused that he really should have given his readers the value of integrating the two so we learned more.24 Daron Acemoglu and Jackson’s Power and Progress had a 50-page “Bibliographic Essay” (in addition to a 41-page References) instead of just a bibliography and it was fascinating. And Eric So’s book Collision had no index, which made his book just a little less usable (was it because he used an authoring platform like Substack, which is not structured as pages? — what are the implications of that?)

Other Topics Of Interest

I write about financial management and control, technology management, philanthropy, investing, and publishing. My level of expertise is roughly in that order. I will organize my posts by those headings. Along the way I will write posts that are “Of Interest” to me and see if they have any traction.

Examples of posts I intent to write about include:

• Off-loading bad investments or assets.

• The impact of Christopher Alexander and self-consciousness

• More on affordances

• How investment funds should evolve and the design of indices such as Rob Arnott’s Fundamental Index

• How to choose valid Alternative Investments

• How to get support for a candidate, or a philanthropy

Why does a Substack reader have to Subscribe to read posts?

On Substack “Liking” an author is like patting then on the shoulder, lightly — it is not good enough. Having subscribers is in Substack’s DNA, so it relies on readers actually subscribing — even if it is a free subscription — to communicate the continuing relationship between an author and a reader. Subscribe as a Free, Monthly, Annual, or Retainer member. An author can add value to the paid levels over time.

This is my Substack Website, with all my Substack-based posts, notes, and book drafts. Each is the product of an iterative, interactive, continuous publishing process using Substack as my publishing platform. Posts are first circulated as notes to everyone and then updated continuously. Substack renders my notes as a Website for henrylichstein1.substack.com so you can always find the most up to date version of any post there. Website

We all have instant access to so many interesting things going on in the world that it is difficult to keep focused. Content arrives from near and far at increasing speed. We have available to us an extraordinary range of tool that are constantly changing. And the reality of our daily lives is changing at a faster and faster pace. I use this newsletter to try to keep my attention on what is “Of Interest” to me, now. I share it with you with the hope you will indulge me by reading it. ‘Publishing on Substack’.

Please take a look at my posts and make comments as you can. I add links to sources as often as possible so you will know what influences me. I have tried to keep my anecdotage to my memoir posts, each a subset of “Connecting the Dots”. First chapter in ‘Memoirs’

I hope to earn an audience for what I write and to get all kinds of reader feedback. Eventually, a linked set of posts may be published as a book or two.

Share this or any post, and please recommend my Substack to whomever you can.

Unsubscribe any time.

Links to other posts :

Mantras

Philanthropic Organization and Development

Investing is a Contact Sport

Draft of Book: “Stop “Managing By Your Checkbook”©

Mantras are compact summaries or pithy sayings that convey general lessons or encapsulated insights, reduced to a few words . They are meant to invoke ideas much larger than their few words. Some are pretty clear and specific while some are abstract and appeal to a general concept. In some cases, the Mantra summarizes what could be a whole book, and in many cases, book titles are themselves Mantras.

The Mantras listed here are ones I have depended on over the years to get a point across. Some come in multiple versions to make the same point in different contexts. Some are original, many I think are original but I likely read or heard from others. I try to give credit where credit is due and I try to link the mantra to its source in the listings below. I link each Mantra to where it is used in my posts and give their history as background.

I chose to call these “Mantras” but others have called their lists Laws, Lessons, Rules, Proverbs, Precepts, Chakras, Maxims, etc. I include similar lists which may be “Of Interest” from others like Donald Rumsfeld, Elon Musk, et al., as appendices.

Last edited October 28, 2026

Lichstein’s Mantras Summarized

• One: “Be on your own Agenda, not Someone Else’s.”

• Two: “There is a Thin Line between Persistence and Harassment.”

• Three: “Once you simplify and organize, become knowledgeable, and get prepared, connecting the dots is easy!”

• Four: “Stop Managing By Your Checkbook!”

• Five: “Establish ‘Accounting Control’ wherever you can.”

• Six: “Standards Follow Practice, They Do Not Lead.”

• Seven: “If you don’t know the rules, don’t play the game.”

• Eight: “Only new questions create new knowledge.”

• Nine: “If something works for you, keep doing it.”

• Ten: “Tell suppliers ‘I do my own plumbing’ (or words to that effect).”

• Eleven: “Contracts have to balance asymmetries in knowledge, value, cost, risk, and fear.”

• Twelve: “‘Going to Court’ is seldom used to resolve issues.”

• Thirteen: “You can always break a commercial lease.”

• Fourteen: “Get equity rewards if you are taking equity risks.”

• Fifteen: “Know your counterparty’s accounting rules and their needs so you can anticipate their responses to change.”

• Sixteen: “If you have someone who is indispensable, fire her!”

• Seventeen: “Never ‘Split the Difference’”

• Eighteen: “The first rule of fundraising is, ‘If you don’t ask, you don’t get’. The second is ‘If you want advice, ask for money. If you want money, ask for advice’.”

• Nineteen: “The mongoose always catches the cobra.”

• Twenty: “Designing a fund is the same as designing an index.”

• Twenty-One: “I can call my kid ugly, you can’t!”

• Twenty-Two: “Never Solve in Theory what you can solve in Practice.”

• Twenty-Three: “All a patent does is to give you the right to sue.”

• Twenty-Four: “He’s an acquired taste.”

Lichstein’s Mantras One-by-

One: “Be on your own Agenda, not Someone Else’s.”

Get done what you want to get done, not what others want. Own your own time.

Other versions:

• See also Rumsfeld’s inbox principle — likely where I got the idea for my mantra!

• ‘You make money when you buy not when you sell.’ This is often true because you are on your own timetable when you buy but you are on the buyer’s timetable when you sell. 1

• ‘Write the first draft of all your contracts.’ There are subtle biases in every choice of topic and word. Why let your counterparty control the contract by making those choices. If you do not write the first draft, read whatever you receive and comment as soon as possible. Move the conversations to your narrative and your needs.

  • In Dryad Communications, we wrote our own contracts:

    • Earned the right to two FCC licenses controlled by Long Lines by providing Internet on Tribal Lands.

    • 50% deposit from InSight for a cancelled contract in Minnesota.

    • T-Mobile spectrum purchase contract anticipated FCC conditions and obligated T-Mobile’s purchase.

    • FCC spectrum-buys claim, “I never lose money buying spectrum,” 2 was accomplished by never selling until he had a profit!

Two: “There is a Thin Line between Persistence and Harassment.”

The line is different in every situation. Find the line by being as persistent as possible using judgement as to how to move things along as fast as possible.

• Examples in Memoirs:

• Dryad called Long Lines every three weeks for three months and eventually bought their spectrum licenses.

• I called INRIX every six weeks for six months to get attention and INRIX eventually bought ParkMe.

• Steve Cotler’s self-description, “I am a ‘scout’ (persistent), not a ‘marksman’ (harassment)?

Are you “Connecting the dots” or “Jumping to conclusions”?

Three: “Once you simplify and organize, become knowledgeable, and get prepared, connecting the dots is easy!”

Logic emerges from the self-conscious collection and organization of knowledge.3 Focus on the basics, describe the flows, capture the ecosystem, make a flowchart, do whatever is necessary to see how the pieces all fit together.

• All spreadsheets should flow and tell a story:

  • Group lines and columns by topics, types, subsets.

  • The numbers should add up no matter you look at them; top-to-bottom, side-to-side, and page-to-page. Formats should match. There should be no questions demanding to be answered.

  • Create checksums, by row, column, and page. Aggregate page checksums to the Summary and workbook.

• All contracts should flow and tell a story4:

  • Group by sections & by topic & by paragraph. Keep like things together.

  • Follow a template.

• Reduce complex problems to the underlying mathematics involved, as “It’s in the math.”

  • Fast Fourier Transforms (LTE, etc.)

  • Quantum Mechanics

  • Wave Equations

  • Signal Theory

  • Bayesian Statistics5

  • LLMs

• Examples from things from my memoirs:

  • Knowing that “Great Plains Communications just up and drove a tractor from one end of the Tribal land to the other,” told Dryad that the local telephone company had quietly installed fiber to the town of Santee — why else would they drive a tractor through a Tribal Land other than to lay a fiber conduit!?

  • Knowing there could be no negotiating during an auction, Dryad waited out T-Mobile until just before a spectrum auction. When T-Mobile called Dryad just days before the next year-long auction began, they were in a corner. They asked, “What would it take to get you to sell your licenses?” Dryad gave them a them a take-it-or-leave-it offer and they and they had no choice but to accept.

  • Since the first step to secure music rights is to get a license from the Library of Congress to transmit copyrighted material, CCM concluded that fees were due only when copywritten bits were actually being transmitted. CCM paid no fees if no copywritten bits were flowing.

  • The Rosebud Sioux Tribe told Dryad, “First, get a Business Permit.” They wanted us to accept their authority by paying the $60 fee, which did, and they noticed.

  • Karim Lakhani’s maxim “Never Shout Louder Than You Can Hear” was the secret to his proprietary OTA Internet system: He set the transmit power to reach the furthest user, then attenuated the signal for each closer user .

  • By arraying Citibank technology strategy with projects as a ray of segments on concentric Strategic Circles, where each circle was the effort for a year of time, with current efforts in the segments near the center and future effort in segments in concentric circles spreading outward, all relationships among the projects — going out the rays from center to edge for time and around the circle in each year for interrelationships — we could see progress from year to year and could explain technology strategy plans on one sheet of paper.

  • The initial CCM financial review revealed that the product was inherently profitable and showed a steady revenue growth pattern. This was the basis for our funding strategy and led directly to an eventual successful liquidity event. The dots lined up, over a three-year success story.

  • CCM decisions to accrue real expenses and capitalize development costs from the earliest days made it possible to show a steady march toward profitability and exponential growth of profits from the start. By getting the profit story right in the financial presentation of the company from the very beginning. the maturity of the product and of the company was moved forward by as much as three years.

Four: “Stop Managing By Your Checkbook!”

Managing by your checkbook is abdication of your responsibility to manage your assets.

• When the checkbook is the financial control system, the bank is the controller and forward planning and therefore company growth is impossible. Your bank controls how you deal with your vendors and limits your cash planning. Move beyond that as soon as possible.

• You cannot grow if your actions are limited by how much money you have in the bank at the moment.

• You must know in advance what is going to cause you trouble, so you can plan to prevent problems before they cause you harm.

Five: “Establish ‘Accounting Control’ wherever you can.”

Only by establishing accounting control over process, using the tools of double-entry bookkeeping, can a company grow. Double-entry bookkeeping causes the balance sheet and the income statement constantly to talk to each other. Setting up this conversation is what financial management and control is all about. Only when you and your counterparties are under accounting control will your accounts balance, your transactions settle, and your relationships stabilize.

When you establish accounting control, good things start to happen:

• Cash Management processes can be created.

• Reserves can be set up to control timing of recognizing income and expenses.

• Sales expenses can be tied to the revenues they bring in so revenue recognition makes sense and profits are correctly reported.

• Inventory can be managed, especially if there is accounting control done in both the financial and physical domains.

• Product Costing systems can be set up.

• Capitalization of expenses can be initiated.

The successful sale of Cloud Cover happened because we put each and every process under accounting control. Cloud Cover Story

Six: “Standards Follow Practice, They Do Not Lead.”

Standards can be codified only after there has been enough practice to support them.

• Look to the world of shareware — XML, Apache, MarkDown, Linux, SQL, JAVA — or hardware — the RJ11, the Tesla power plug, railroad track spacing. They became “standards” only after they emerged from “the practice” as adopted by users. Economists call it “path dependence.”6

• QWERTY was the practice of the Remington typewriter for a decade when Mergenthaler faced its keyboard decision. Staying with ETAOIN SHRDLU condemned linotype operators to learn a different way to type.7

• The concepts of Freeware and Paywalls are not inventions of the Internet, they have been the engine of growth of many industries over a long period of time:

  • Free or 2 cent newspapers generated advertising revenue by which they grew(Tim Wu8)

  • “Walled gardens” failed but advertising worked as the source of revenue on the Internet.

  • KDP has recently set long-form writing publishing standards.

• Affordances are what the user can do with a product, what the product provides its users. A product’s “DNA” — how it came together and its capabilities — drives its affordances. Look at how the model for social networks evolved — Twitter, Substack, LinkedIn, or WordPress — or products like Shopify. Successful networks drive huge ecosystems which preserve the social network’s command of a market. Successful products and their customers/users co-evolve. Capability bundles are driven by customer affordances (form follows function).

• MONDEX failed because it was never really used.

• I deal with affordances at length in Writing, Publishing and I may write more.

Seven: “If you don’t know the rules, don’t play the game.”

• Said other ways:

  • If you’ve been in a poker game for more than 30 minutes and do not know who is the mark, you are the mark.

  • If the service is “free”, you are the product.

  • When you start calculating your profit it is time to take some money off the table.

  • If you’re thinking about quitting/getting divorced/leaving a job, the process is likely already underway. It will happen whether you like it or not.

• This, too, is dealt with in Writing, Publishing

Eight: “Only new questions create new knowledge.”

• Said other ways:

  • My earlier version was “Truth is a Question” but that was too obscure.

  • You start to learn only when you push the boundary of knowledge until you reach what is NOT known.9

  • The strongest argument against Artificial General Intelligence taking over, or against AI being smarter than humans, is that the boundary of knowledge is pushed outward only by asking questions no one has asked before. 10

  • See also Eric So11 and my post on Coming to Terms with AI

Nine: “If something works for you, keep doing it.”

• I often say I am “Solving the problems of the world, one spreadsheet at a time.” Mike McKay (RIP) described himself as “Solving the problems of the world, one Power Point presentation at a time.”

• Malcomb Gladwell, Michael Lewis, and Walter Isaacson keep writing the same book over and over, to great acclaim. The risk is that the formula that works limits what you write about.

Ten: “Tell suppliers ‘I do my own plumbing’ (or words to that effect).”

Put them on notice you know how to do their work. You are in control if you can credibly question what does not sound plausible.

• Corollary: “Do your own taxes.” It it the only way you learn the importance for your investment decisions of, for example:

  • Tracking NOLs from year to year.

  • What makes your investment portfolio truly diversified and balanced.

  • See discussions in Investing is a Contact Sport regarding offsetting taxes and dealing with multiple K-1s from one asset manager and especially among multiple asset managers if you find yourself in that situation.

Eleven: “Contracts have to balance asymmetries in knowledge, value, cost, risk, and fear.”

Write contracts so failure to achieve goals or to perform is costly and the effects are cumulative. Anticipate the unexpected but extreme outcomes (edge conditions). The contract has to be self-enforcing since using lawyers and going to court is is expensive.

• Contracts should be written to create a balance among risk and return for all parties.

• Recognize things could go very right or very wrong.

  • Consider total failure and dramatic success

  • Consider failure to perform, unusual environmental conditions.

• No party should give up anything without getting something in return.

• Details matter:

  • Changing one clause, from, “If Teradata is sold” to “if either party is sold,” doubled the small company Teradata’s value when its huge R&D partner NCR was eventually sold to AT&T, Teradata’s largest customer at the time .12

Twelve: “‘Going to Court’ is seldom used to resolve issues.”

Said another way, “Never take a hostage you are not willing to shoot.”

• If you are the creditor, deal with past due debtors, missed deadlines, and defaults as soon as they occur because you are seldom really willing to shoot the debtor.

  • Confirm missed obligations, reset terms if the deal was bad from the start.

  • Force your debtors to live up to their new obligations.

  • Implementation of laws in court is arbitrary.

• If you are a debtor with more accounts payable than you can pay and are being pressed by creditors:

  • Tell them your story in detail

  • Immediately negotiate for a better deal.

  • Failing that, tell what you can afford to pay over time.

    • Some will settle at a discount.

    • Some will accept a promise of a little at a time.

    • Commit only what you can pay.

  • Scrutinize any new invoice from old creditors; they try to improve their position, thinking you are ignoring them.

• See story of Torrance Optics for examples of spinoffs that lasted long after they were unable to pay their bills.

• See investing stories of managing by K-1s.

Thirteen: “You can always break a commercial lease.”

• Leave the premises as soon as you know you cannot afford them.

• Vacate because occupying property without paying rent is an Unlawful Detainer, which is theft, a crime.

• Once you vacate, you are violating a contract (the lease), which is a commercial dispute, subject to different rules, procedures, penalties than a crime, and is an issue that can be negotiated.

• If the landlord abandons the issue it can be removed from the books after the statute of limitations.

NB: This is not legal advice, it is business advice.

Fourteen: “Get equity rewards if you are taking equity risks.”

• If your employer defers income, or your customer defers payment, or your borrower unilaterally extends a loan, you are being forced to take equity risks.

• Demand equity returns like increased payments, and ones that grow with time (compounding), or equity in the company, or some onerous combination.

Fifteen: “Know your counterparty’s accounting rules and their needs so you can anticipate their responses to change.”

Examples from Cloud Cover Media

• CCM cut back music rights payments payments when COVID-19 hit.

• Rightsholders payment calculations drove their expectations and left CCM in control of the calculations.

Sixteen: “If you have someone who is indispensable, fire her!”

You cannot long survive with an irreplaceable key person, so act before you lose them! Once you have backups, no one is indispensable.

Seventeen: “Never ‘Split the Difference’” 13

• The real goal is to get all or give nothing. You just have to be patient.

• See Dryad Communications. By waiting until just before an FCC auction was to start, T-Mobile offered to split the difference but Dryad said, “No, this is our ‘take it or leave it’ number.”

Eighteen: “The first rule of fundraising is, ‘If you don’t ask, you don’t get’. The second is, ‘If you want advice, ask for money. If you want money, ask for advice’.”

There is much more to fundraising but start there!! Philanthropy Post

Nineteen: “The mongoose always catches the cobra.”

A mongoose would die if bitten by a cobra but they almost never are. A cobra will always strike when a mongoose lunges, hissing, with its mouth wide open, fangs and venom ready. Both retreat, and the cycle repeats. The mongoose knows the cobra’s range and stops short, out of the cobra’s range.

But there is a subtle changes each time. The mongoose slowly shortens its timing each cycle just a little, changing the “phase relationship” with each cycle. Finally, the cobra is fully extended and helpless after the mongoose has regained his position: At that point, the the mongoose moves within range and pounces on the fully extended cobra and kills it. The mongoose had learned that it could win by changing the timing each time!

• You can see this what an asset manager keeps making changes that are fair, until he makes one that isn’t, and you lose something.

Twenty: “Designing a fund is the same as designing an index.”

The technical details and the design issues are very very similar whether you are creating a fund of a mix of assets for a purpose or trying to track the value of the same assets for the same need. The decisions that underlay the construction of the index are the same as would be needed to create a fund of the same assets.14

Twenty-One: “I can call my kid ugly, you can’t!”

I think this is original. It is pretty much self-explanatory.

Twenty-Two: “Never Solve in Theory what you can solve in Practice.”

I heard this first from Abe Peled ca 1995, when he worked for IBM. We went on to become a force in the world of digital communications.

Twenty-Three: “All a patent does is give you the right to sue.”

There are no rights inherent with a patent. You have to actively assert your ownership to gain the benefit of a patent. No one is prohibited from violating a patent.15

• SBIRS seldom have the capital to sue to protect their patents.

• The stories of Paul Richman and Tom Katis

Twenty-Four: “He’s an acquired taste.”

Some of the most interesting people I have known, for good or bad, could be described as “acquired tastes”, which is how one of my friends described single malt scotches. I am thinking of people like the departed Merv Hecht and Reuben Sandler. Like single malt scotches, their character is strong, often pungent, very distinctive, and not always pleasant at first exposure. But something about them keeps you coming back, for good or bad.

Books have been written by and about many of the people who might be called “acquired tastes.” In some cases, the ones that are written miss the point, and that is frustrating. The book about John Van Neumann, with each chapter a view of the subject from the very different perspectives of that of a lover, an offspring, a competitor, a colleague is nearly perfect.16

______________________________________________________

Other Useful Things (mini-Mantras)

• Keep a valid checking account open for at least a year after you close a company. There is any number of reasons you will need it.

• Fill shelves to no more than 85% full.

• Never sign up for a free month subscription for something unless you intend to use it immediately and you expect to continue to use it. The vendor knows you will forget to cancel. It will be hit your credit card for months before you notice it.

• Making contacts is not enough. The contacts have to produce something. Don’t confuse making connections with having actions underway.

• Tape Out: I worked at Texas Instruments in the late 1960’s when integrated circuits (ICs) were first designed like they were pieces of art, and then manufactured at scale. The design for the lithographic masks by which ICs were made were sketched by engineers, which were given to draftsmen who put pencil lines on vellum. When the design was finished, drafting tape was put down on the pencil lines to create the negative for the multiple lithographic masks. Thus, “Tape Out” really meant the design of the IC was finished. It was the final step before an IC was manufactured!!

Appendices

• Donald Rumsfeld’s Rules

• Elon Musk’s “Production Algorithm” and “Idiot Index”

• Stan Spiegel’s Rules for investing in Malls

• Dolly Parton’s Terms

• Peter Thiel’s Rules

Donald Rumsfeld’s Rules

• The Inbox Principle: “If you are working from your inbox, you are working on other people’s priorities.”

• The “Washington Post” Test: “Don’t do or say things you would not like to see on the front page of the Washington Post.”

• The Exit Strategy: “It is easier to get into something than to get out of it.”

• Accountability & Speed: “If you foul up, tell the president and correct it fast. Delay only compounds mistakes.”

• Managing Up: “Be able to resign. It will improve your value to the president and do wonders for your performance.”

• The Measurement Maxim: “You get what you inspect, not what you expect.”

• Emergency Protocol: Adapted from Navy pilots: “If you’re lost, climb, conserve, and confess.”

Elon Musk’s “Production Algorithm'‘

As adapted by from Walter Isaacson’s Musk biography17

1. Question every requirement. Go as far as getting the person who made it justify it. Make them “less dumb” wherever you cannot eliminate them.

2. Delete any part or process you can. If you do not eventually add back at least 10% of what you deleted, you did not delete enough.

3. Simplify and optimize after you eliminate all the requirements and processes you can eliminate.

4. Accelerate cycle times.

5. Automate where it makes sense, and only after you have simplified everything.

There are also his corollaries:

1. Technical managers have to have hands-on experience

2. Comradery is dangerous

3. It is okay to be wrong. Just don’t be confident and wrong.

4. Never ask your troops to do something you’re not willing to do.

5. When there is a problem, meet with the level below the manager

6. Skills can be taught, but attitude is built in.

7. Operate at a maniacal sense of urgency

8. The only rules are the ones dictated by the laws of physics. The rest are recommendations.

And the Musk Idiot Index: “How much more costly a finished product was than the cost of its basic materials.” If a product’s Idiot Index was high, its cost could be reduced significantly.18

Stan Spiegel's Rules 19

1. Buy low. Pay 10-15% of original cost, typically from a bank that is losing money on a property. Argument to present: “Listen, you are losing money year-after-year on that property. You can’t manage it. I can.”

2. My gauge of how good a property is is how well a name store in the same mall does. If it does as well as its counterparts elsewhere, the property is good. If the rental roll is low, that is a function of management, not the mall. Buy the property and fill it up. Do anything to get the space occupied.

3. Move fast. My typical deal takes 30 days. No financing. If you finance, do it after the purchase. Buy low enough that it does not matter. If you use the rules above to set price, there is no need for extensive due diligence.

4. Demand “estoppels statements” from every renter. That is the only way to assure yourself of revenue, and it establishes a legal claim against the renter, not the seller.

5. Do not buy malls, their maintenance costs are too high.

6. Take over management of properties. Property management firms make outsized profits, has the wrong incentives, and typically takes advantage of naïve owners. Taking over management typically reduce management cost by 30-50%, which goes right to the bottom line.

7. Do not buy to sell. Buy to hold.

8. Banks are fickle. I was Wells’ biggest Texas customer at one point — on their Texas Advisory Council. Despite never missing a single payment, they called my loans when they were tight. Never again.

9. Hire women whose husbands get benefits where they work. I pay high salaries but no benefits.

10. Keep overhead low. I have 4 million square feet of retail space and a staff of 12, total. Half are bookkeepers.

11. Have impeccable bookkeeping on each property. No buyer will touch the property if you do not. Not worth saving money on accounting. (Story of great deal that failed when bookkeeping reviewed.)

Dolly Parton’s Terms20:

1. “Find out who you are and do it on purpose. And do it with purpose.”

2. “Don’t ever sacrifice who you are, your principles and your morals and your values to just try to get ahead in business. It ain’t worth it.”

3. “We are who we are, and we should be allowed to be who we are.”

4. “Dream big dreams; learn everything you can learn; and care for all those who care for you.”

Peter Thiel’s Rules from Zero to One (as adapted by HAL) 21

1. If you are copying Bill Gates, Larry Page, or Mark Zuckerberg you won’t learn from them.

2. Successful people find value in unexpected places, doing new things.

3. Competitive markets destroy profits.

4. All happy companies are different.

5. Monopolists lie to protect themselves. So do non-monopolists. But Monopolies drive progress.

6. A great business is defined by its abilities to generate cash flows in the future. It must grow and endure.

Eric C. So’s Rules from The Collision22

1. Value the Struggle

2. Value Who You Are Without AI

3. Reinvest your Surplus

4. Make AI your Cognitive Trainer

Lots of Copies Keep Stuff Safe (LOCKSS)

to come23

Notes