Join AdValorem Syndicate for Neuralink

In September 2026, a Neuralink patient suffering from ALS used his brain implant and AI to tell his wife “I love you” in a recreation of his own voice.

Read more

Last Week by AdValorem | October 19–25, 2026, with an October 26 update

No crying in the casino. But do notice who’s cashing chips while telling you the buffet is excellent.

California’s pension giant CalPERS reduced its combined direct share count across the Magnificent Seven by 26% in the first half of 2026.

Meanwhile, Blackstone’s Jon Gray is pushing back against AI bubble fears. Blackstone reports approximately $275 billion in data-center and digital-infrastructure investments.

The seller is diversifying. The infrastructure investor is defending the buildout. Neither position proves the other wrong.

But their incentives are worth understanding.

1. Follow the Money. It’s Going in Circles.

Chamath Palihapitiya recently offered a startling estimate:

Roughly 1% of companies account for 90% of AI token spending. He estimates that 80% of that spending comes from unprofitable venture-backed companies.

That’s his estimate, not an established industry-wide measurement. But the concentration he describes raises a serious question about how much AI demand is self-sustaining.

Think about the economics.

Pension funds give money to VCs. VCs fund startups. Startups buy compute and AI services. Those expenditures become somebody else’s revenue. Rising revenue supports another funding round.

Congratulations. We may have invented a financial perpetual-motion machine. Until somebody asks where the actual profits are.

This doesn’t make AI useless. It makes the distinction between real end-user demand and venture-subsidized demand enormously important.

And when the music slows, valuations can fall considerably faster than technology improves.

2. The Boom Just Hit a Power Outlet

Microsoft walked away from a planned Nscale facility involving 23,040 Nvidia GPUs after grid delays pushed its expected electricity connection into the 2030s.

Apparently, spending billions on artificial intelligence doesn’t exempt you from the laws of physics.

Meanwhile, the industry is proposing gigawatt-scale campuses and borrowing billions to build them.

The opportunity? Power-ready real estate, energy infrastructure, distressed developments, secondaries, and the financing behind the boom.

The risk? Paying tomorrow’s prices for assets that cannot deliver tomorrow’s capacity.

The next fortune might not come from buying the hottest AI company. It might come from buying its abandoned infrastructure at a discount.

Our analysis: Microsoft’s 23,040-GPU problem

3. They’re Selling the Future. But Who Gets to Own It?

Here’s what bothers me more than inflated valuations.

Ordinary people already lived through crypto crashes, rug pulls, speculative tokens, and endless promises of financial freedom.

Now comes another parade: AI startups, private-market access, perpetual fundraising, and assets priced for a future that may never arrive.

My concern isn’t merely that people lose money.

It’s that AI could reduce the value of their labor while ownership of the productive assets moves further out of reach.

GDP could grow. Corporate profits could soar. Automation could make businesses extraordinarily efficient.

And millions of people could still find themselves economically worse off.

A booming economy is little consolation if somebody else owns everything producing the boom.

4. There’s Another Side to This Trade

Not everything ends in disaster.

Microsoft, Google, Intuit, and Meta are releasing increasingly capable business agents.

A five-person company may soon operate with capabilities previously reserved for a fifty-person company.

That’s real productivity, provided those capabilities deliver measurable savings and the agents remain under accountable human control.

It’s also the thesis behind our work on AgentNet: building the infrastructure for autonomous agents to work, transact, and operate within defined authority.

The real AI opportunity is not another chatbot pretending to be your friend. It’s intelligence that actually produces economic outcomes.

Explore The AI Guy’s practical business playbooks

5. And What Happens to Everyone Else?

Every technological revolution creates extraordinary wealth.

And every revolution leaves somebody behind.

The most disturbing future isn’t one where AI fails.

It’s one where AI works spectacularly—and the benefits accrue almost entirely to those who already own the assets.

Imagine a world where your productivity is measured, your behavior is optimized, your economic value is calculated, and your choices are increasingly dictated by algorithms you neither own nor control.

Science fiction?

Perhaps. But the following short film offers an unsettling illustration of what that world might feel like.

Through Love of Humanity, we’re examining the institutions, families, and communities that preserve wealth and turn it into something meaningful.

Technology without broadly shared prosperity isn’t necessarily progress.

Sometimes it’s just more efficient concentration.

The Bottom Line

I’m not bearish on artificial intelligence.

I’m bearish on the assumption that every AI investment deserves to make money.

The internet survived the dot-com crash.

Your portfolio wasn’t included in the warranty.

We’re watching four opportunities: distressed AI infrastructure, private-market repricing, practical business automation, and autonomous economic systems.

But we’re also watching the exits.

Because “be patient” sounds suspiciously like “please provide exit liquidity in an orderly fashion.”

The house doesn’t need you to win.

It needs you to keep playing.

Watch the chips. Not the floor show.

AdValorem Research follows capital, intelligence, autonomy, and humanity through AdValorem Syndicate, The AI Guy, AgentNet, and Love of Humanity.

Partner with AdValorem

Investment commentary and opinion. Not investment advice.

#AIBubble, #ArtificialIntelligence, #AIInvesting, #VentureCapital, #AIInfrastructure, #ExitLiquidity, #TechStockCrash, #AgenticAI, #WealthInequality, #DistressedInvesting

Wall Street is beginning to question the trillions being poured into artificial intelligence. AI infrastructure stocks are falling, borrowing costs are rising, and companies like CoreWeave are seeing their debt trade at distressed levels. Sound familiar? In 2008, collapsing real estate markets created extraordinary opportunities for investors who understood that a failing borrower didn’t necessarily mean a worthless property. Today, we may be witnessing the early stages of a similar repricing—except the assets aren’t just buildings. They’re GPUs, data centers, patents, startup equity, warrants, and contractual rights potentially worth millions. The AI revolution may be real even if some of the companies financing it don’t survive. And that’s precisely the opportunity we’re watching: when the hype disappears, the capital retreats, and distressed sellers emerge, who will recognize the value left behind? That’s where our story begins.

Apply to Invest with AdValorem Syndicate

We Used to Buy What Banks Didn’t Want.

After the financial crisis, our investing strategy focused on foreclosures and distressed real estate.

The premise was simple: a distressed seller and a distressed asset aren’t necessarily the same thing.

A bank needed a property off its books. A borrower needed relief. A buyer with patience, capital, and the ability to untangle the situation could sometimes purchase something worth considerably more than its asking price.

We weren’t betting on the bank’s failure. We were investing in the value that remained.

More than a decade later, we’re applying that same philosophy to a much larger universe.

Venture capital. Failed accelerators. Bankrupt investment funds. Equity warrants. Contractual rights. Distressed technology portfolios. Litigation-linked recoveries.

The assets have changed.

The principle hasn’t.

Buy value others have stopped looking for.

1. From $500,000 Properties to $50 Million Rights

Consider two illustrative situations.

A bank forecloses on a property with a stabilized value of $1 million. An investor acquires it for $500,000, spends $200,000 restoring and carrying it, and eventually sells for $1 million.

That’s $300,000 of gross profit before transaction costs and taxes.

Now imagine an insolvent venture fund holding contractual rights linked to a technology company valued at $1 billion.

A legitimate, enforceable right to 5% of that company’s equity could have a headline value of $50 million before dilution, exercise costs, transfer restrictions, and other adjustments.

What if the holder can’t afford to pursue it? What if the original paperwork is buried inside a bankruptcy estate? What if the right can be acquired or resolved for substantially less than its eventual economic value?

That is where our research begins.

The $50 million isn’t a promised payout. It is the starting point for underwriting a complicated asset that might ultimately recover substantially less—or nothing.

The opportunity is the difference between the market’s willingness to deal with complexity and the value that can actually be realized.

Our AdValorem Syndicate strategy approaches this through three disciplines:

Acquire. Advise. Enforce & Finance.

And we’re particularly interested in businesses that accumulated interests in other businesses before failing themselves.

So, How Many Enforcement Actions Are We Planning?

Our broader warrant inventory has been tracked at 213 instruments.

But we’re not planning 213 lawsuits.

We’re evaluating a portfolio of economic rights, one company and one instrument at a time. Some may be resolved cooperatively. Some may support strategic transactions. Others may justify financing, negotiation, or legal enforcement. And some simply won’t be worth pursuing.

The interesting number isn’t the number of lawsuits.

It’s the aggregate economic value of the rights worth enforcing—and how much of that value can actually be recovered.

Our research into the Newchip bankruptcy shows why we’re following this market closely.

2. The AI Gold Rush Is Creating Two Kinds of Opportunities

AI infrastructure spending isn’t slowing down.

Our October 15 TSMC research reports record quarterly revenue of approximately $46.7 billion, up 50.9% year over year, with net income near $24 billion.

Meanwhile, Nvidia’s declining share of China’s accelerator market illustrates how rapidly technological advantages can shift.

Our October 12 research on Boston Dynamics examines another frontier: humanoid robotics moving from demonstration to industrial deployment.

Explore the GP research.

Here’s what interests us beyond the headlines.

A company may raise $100 million to build AI infrastructure and ultimately fail because its revenue can’t support its power, financing, and operating costs.

Its investors could lose their equity. But its assets may still include GPUs, power agreements, infrastructure rights, software, patents, and customer contracts.

A buyer might acquire those assets for a fraction of replacement cost—provided their ownership, transferability, and commercial usefulness survive the restructuring.

In a gold rush, fortunes can be made building the mine. They can also be made buying equipment after a poorly financed miner fails.

That is why we study both emerging technology and distress.

3. The $20 Software Subscription Is Becoming a $2 Million Business Decision

Five AI research reports this week show how quickly ordinary business models are changing.

Consider a dental group with 20 offices, each missing 10 potentially valuable patient calls daily.

That’s 200 missed calls per day. If just 10% could have become appointments with $1,000 in eventual collected revenue, the annual opportunity would be roughly $5.2 million across 260 working days.

This is an illustrative opportunity calculation—not evidence of actual recoverable revenue. Real outcomes depend on patient demand, conversion rates, capacity, and treatment mix.

Our AI Front Desk playbook explores the operational case.

Now multiply that kind of opportunity across entire industries.

• OpenAI Dots: A distributor assigns an AI employee to reconcile invoices, follow up on overdue receivables, and prepare purchase orders. The economic prize isn’t a cheaper chatbot; it’s lower operating costs and faster cash collection.

• Google’s Universal Cart: A retailer’s next customer may be an AI purchasing agent comparing inventory, delivery times, and prices across dozens of merchants. Businesses optimized for traditional search traffic may discover that their distribution advantage has weakened.

• Meta Muse connectors: A small company connects its accounting, payment, and commerce systems, reducing the labor involved in processing orders and tracking performance.

• x402 agent payments: Instead of purchasing another monthly subscription, a business agent pays individually for data, verification, and specialized services whenever a job requires them.

Imagine a software business whose customers shift from $500 monthly contracts to paying $0.02 per successful API operation.

At 100 million paid operations, that’s $2 million in gross transaction revenue.

At just 1 million operations, it’s only $20,000.

Same technology. Radically different economics.

That’s why we’re interested not only in who builds the next generation of tools, but also in who owns the infrastructure, distribution, payment relationships, and intellectual property that make them work.

Some companies will scale beautifully.

Others will raise millions, miss their economics, and leave valuable technology behind.

4. AgentNet: What Happens When the Customer Is a Machine?

AI agents are approaching the point where they can discover services, make approved purchases, coordinate other agents, and deliver completed work.

Through AgentNet, we’re exploring the infrastructure for that autonomous economy.

Imagine an AI-driven insurance workflow.

An agent spends $2 verifying an identity, $8 obtaining third-party data, and $15 on specialist analysis to help process a claim.

Now imagine millions of such transactions.

Who controls the payments? Who owns the transaction records? Who absorbs errors and fraud? Who gets paid if the job fails?

Those questions create new businesses—and potentially new liabilities.

A venture-backed agent marketplace might spend $30 million building an execution network, attract users, but never make its unit economics work.

It fails.

Yet its software, intellectual property, contractual arrangements, and acquired customer relationships might still have commercial value.

Not necessarily $30 million. Perhaps far less. But potentially more than a distressed seller can recover through an ordinary liquidation.

Tomorrow’s winners may create enormous value. Tomorrow’s failures may leave valuable assets behind.

We want to understand both.

5. Love of Humanity: What Do We Do With the Wealth?

Technology creates possibilities. Capital determines which ones survive. People decide what matters.

This week, Love of Humanity explored how wealth becomes something that lasts beyond a balance sheet.

The Shore’s Classrooms Are Building in Public examines the educational institutions investing in Monmouth County’s next generation—from Rumson Country Day to Ranney, Christian Brothers Academy, Red Bank Catholic, and St. Rose.

Two River’s Arts Legacy explores how family patronage supports theaters, museums, and cultural institutions.

Where the Shore Keeps Its Clubhouses, Regattas, and Families follows the traditions and institutions that hold a community together.

And From the Rumson Estate to the Federal Form examines the inheritance and estate decisions shaping generational wealth.

Listen to Heart of the Shore, EP38.

Money is a tool. Wealth is what we’re able to build, preserve, and pass forward.

The AdValorem Perspective: Follow the Rights, Not Just the Headlines

We began with foreclosures because we recognized something others overlooked.

Financial distress can force the sale of economic value at a price that has little to do with its long-term potential.

Today, we’re looking for that same disconnect across much larger and more complex markets.

A failed venture fund may own shares in companies worth billions.

A bankrupt accelerator may hold warrants across dozens of startups.

A distressed business may own patents, contractual rights, or customer relationships that remain valuable even after its equity is wiped out.

A disputed right may represent millions in potential economic value—but only if it can be established, defended, and realized.

The numbers can be significant: $10 million, $50 million, even $100 million of nominal asset or claim value in some situations.

But nominal value is not cash. Some of those rights will prove unenforceable, uncollectible, or uneconomic. Others may take years to resolve.

Our discipline is finding the difference.

We study these markets for everyone interested in understanding private capital—from individuals learning how alternative investments work to experienced investors, family offices, and institutions evaluating complex special situations.

Different resources. Different risk tolerances. The same fundamental question:

What is it really worth, what does it cost to acquire, and how do you get paid?

The market loves tomorrow’s winners.

We’re just as interested in what yesterday’s losers left behind.

Explore our investment research and special-situations philosophy at AdValorem Syndicate.

Last Week by AdValorem covers private markets, technology, agentic infrastructure, and wealth stewardship. Monetary scenarios are illustrative, not actual investment performance or projected returns. This publication is educational commentary, not an offer to sell or a solicitation to purchase securities. Distressed investments and legal claims carry substantial risk, including total loss of capital.

#DistressedInvesting, #VentureCapital, #AlternativeInvestments, #AIEconomy, #AgenticAI, #PrivateEquity, #SpecialSituations, #EquityWarrants, #WealthCreation, #InvestmentOpportunities

The technical moat is collapsing. AI agents are learning to reverse-engineer sophisticated software, Amazon is spending billions securing the infrastructure needed for AI, digital employees are transforming small-business economics, and scientists are exploring whether aging itself can be reversed. The next fortunes may belong to those who control what AI cannot easily reproduce.

Apply to Invest with AdValorem Syndicate

This Week’s Big Stories

1. The Technical Moat Is Dying: AI can increasingly reverse-engineer software. Proprietary data, network effects, distribution, and physical assets may become the real competitive advantages.

2. The $1 Billion Permission Economy: Amazon, nuclear energy, and Fluidstack reveal why electricity, land, and community approval are becoming critical AI infrastructure assets.

3. The $100 AI Employee Revolution: AI is moving from answering questions to improving profits for restaurants, insurers, retailers, and wellness businesses.

4. AgentNet and the Autonomous Economy: AI agents are beginning to discover services, execute work, and transact with one another.

5. Who Wants to Live Forever? Early human experiments in cellular reprogramming offer an intriguing glimpse of regenerative medicine’s potential.

6. Love of Humanity: The institutions that protect our communities, educate our children, preserve culture, and turn wealth into lasting legacies.

The AdValorem Thesis: Own What AI Cannot Replicate

This week’s stories converge on one powerful investment idea.

As intelligence becomes cheaper, the value of scarce assets, proprietary information, network effects, reliable execution, and human relationships may increase.

Software can be reconstructed.

But exclusive information, proven distribution, secured electrical capacity, contractual rights, trusted institutions, and successful clinical outcomes remain difficult to replicate.

That doesn’t make every scarce asset a good investment. Demand, price, capital efficiency, and execution still determine returns.

But it does suggest investors should begin asking a different question.

Not simply: What extraordinary technology has this company built?

Instead: What does this company possess that competitors, even with extraordinary AI, cannot easily reproduce?

At AdValorem, that’s the lens through which we’re examining private markets, applied AI, autonomous commerce, emerging biotechnology, and the long-term creation of human prosperity.

The next great fortunes may not belong to whoever builds the smartest technology. They may belong to whoever controls what that technology still needs.

1. The Technical Moat Is Dying. What Replaces It?

For decades, software complexity created competitive advantages. Companies spent millions developing applications that competitors struggled to reproduce.

Artificial intelligence is changing that equation.

New tools such as Reverse Engineer Anything (REA), highlighted in Sashikant Singh’s discussion, allow AI agents to investigate software behavior, analyze accessible binaries, understand application architecture, and help reconstruct functionality.

One developer even described using AI agents and hundreds of billions of tokens to reconstruct much of a commercial video game’s software.

The implications extend far beyond gaming.

AI-assisted reverse engineering could help businesses modernize decades-old enterprise applications, integrate incompatible systems, understand industrial equipment interfaces, rebuild outdated software functionality, and accelerate cybersecurity research.

Imagine a manufacturer replacing a costly legacy operating system without having to reconstruct every workflow manually. Or a logistics company connecting warehouse equipment whose original software is no longer supported.

Work that once required months of engineering could become dramatically cheaper.

Of course, intellectual-property laws, security restrictions, and private data still matter. Reverse engineering doesn’t automatically grant permission to copy protected software or access confidential systems.

But the economic direction is becoming clearer.

If your only advantage is that your software was difficult to build, AI may eventually eliminate that advantage.

This reinforces an idea associated with entrepreneur Nikita Bier: distribution, user behavior, and networks can matter more than technological sophistication.

We believe three competitive advantages will become especially valuable:

• Proprietary data: Exclusive information, transaction histories, and real-world insights competitors cannot easily reproduce.

• Network effects and distribution: Trusted customer relationships and ecosystems that become more valuable as participants join.

• Physical-world advantages: Land, power, infrastructure, specialized equipment, facilities, and regulatory permissions.

Technology still matters. But increasingly, what you own, whom you reach, and what you uniquely know may matter more than what you can code.

2. The $1 Billion Permission Economy: AI Needs More Than GPUs

The AI infrastructure boom is entering a new phase.

The greatest bottleneck may no longer be acquiring another processor. It may be finding somewhere to install it—with enough electricity and permission to operate.

Three stories from our private-markets research illustrate the opportunity.

Amazon’s $1 billion community strategy

Amazon’s reported five-year, $1 billion commitment to communities around its data centers highlights the rising importance of local acceptance.

Our research examined more than 100 data-center moratoriums, over 120 blocked projects, and approximately $198 billion in affected development activity.

With a reported AWS backlog approaching $496 billion, permission to build carries enormous strategic importance.

Read: Amazon Buys the Permission to Build

The $4.5 billion nuclear opportunity

Reliable power is becoming one of the most valuable inputs in AI infrastructure.

Our report examined a $4.5 billion small-modular-reactor financing initiative alongside emerging hydrogen fuel-cell and dedicated-generation concepts.

The opportunity extends beyond producing electricity to securing dependable power at the precise locations where AI customers need it.

Fluidstack’s $4 billion Texas campus

Fluidstack’s proposed $4 billion AI campus in Cameron County, Texas, illustrates how developers are assembling power, infrastructure, and community commitments to make enormous computing projects viable.

The project was described as reserving approximately 1.5 gigawatts of power and contributing $27 million toward local infrastructure without tax abatements.

The investment lesson: GPUs can be manufactured. Land, grid connections, infrastructure rights, and community acceptance cannot be generated on demand.

For private-market investors, controlling scarce physical assets and contractual rights may become increasingly valuable as intelligence becomes cheaper.

Explore more at AdValorem Private Markets.

3. The $100 AI Employee Revolution: Intelligence Becomes a Margin Line

While technology giants invest billions in infrastructure, small businesses are discovering a different opportunity.

They can increasingly purchase intelligence as an operating expense.

And the goal isn’t another chatbot. It’s higher profitability.

This week, AdValorem AI explored five practical opportunities.

The always-on AI employee. Emerging cloud-computer agents point toward digital workers capable of handling recurring administrative processes, coordinating applications, and completing routine tasks. Our playbook examined a roughly $100-per-month AI-worker proposition, subject to actual capabilities and reliability. Explore the playbook.

Smarter restaurants. AI forecasting integrated with point-of-sale systems can help restaurants predict demand, reduce food waste, optimize purchasing, and improve staffing decisions. Explore the playbook.

Automated insurance operations. Applied Epic Conductor offers the potential to eliminate repetitive data entry and accelerate submissions. Vendor-reported figures include over 50% less manual entry and approximately 55 minutes saved per submission. Explore the playbook.

AI-powered retail. Google’s Universal Commerce Protocol points toward shopping experiences where agents discover products, compare offers, and complete authorized purchases. The next customer visiting a store may be software acting for a person. Explore the playbook.

Longevity memberships. Cash-pay functional-medicine practices are exploring recurring wellness memberships priced from approximately $5,500 to $45,000 annually, combining diagnostics, preventive care, and personalized monitoring. Clinical value varies by service. Explore the playbook.

The bigger opportunity isn’t replacing every employee.

It’s giving small businesses access to capabilities previously available only to much larger organizations.

AI becomes truly valuable when intelligence turns into measurable revenue, lower costs, or better customer service.

More on The AI Guy Podcast.

4. AgentNet: The Internet Is Learning to Do Business

The first internet connected information.

The second connected people.

The next may connect autonomous economic actors.

An AI agent identifies what its user needs. Another agent provides the service. Payment is authorized, work is completed, and the result is verified.

This is the opportunity behind AgentNet.

Protocols such as x402 are helping lay the groundwork for machine-to-machine payments, while emerging commerce standards are making agent-mediated purchasing increasingly practical.

But the lasting value may not be the individual AI agent.

It may be the network connecting them.

A marketplace with verified providers, reliable execution, proprietary transaction histories, and trusted payment infrastructure could become significantly more defensible than any single application.

Reverse engineering adds another dimension: specialized agents could offer lawful legacy-software integrations, data-format conversion, automated system migration, and compatibility services to other agents.

The next internet won’t just tell us what to do. Increasingly, it will help get things done.

5. Who Wants to Live Forever? The Biological Revolution

What if the next great programmable system isn’t a computer?

What if it’s the human body?

Our recent AdValorem Research article, Who Wants to Live Forever?, explores one of biotechnology’s most provocative frontiers: partial cellular reprogramming.

Life Biosciences reported preliminary observations from three glaucoma patients treated with ER-100, an experimental gene therapy using three Yamanaka factors—OCT4, SOX2, and KLF4.

The aim is to investigate whether certain cellular functions can be restored through controlled reprogramming.

In the initial low-dose cohort, two patients reportedly showed encouraging changes in portions of their visual-field testing.

These are extremely early observations, not proof of reversed aging. Larger controlled trials and long-term safety data are essential.

But the scientific possibility is profound.

Today’s longevity industry largely focuses on preventing or slowing deterioration through better diagnostics, metabolism, lifestyle, and medical interventions.

Partial reprogramming asks something more ambitious:

Could we eventually recover biological function that aging or disease has already taken away?

If regenerative therapies ultimately become safe and effective, the implications could transform healthcare, retirement, insurance, and the economics of human longevity.

It may be one of the most consequential opportunities of the coming decades, although clinical success remains highly uncertain.

Read: Who Wants to Live Forever?

6. Love of Humanity: What Do We Do With All This Wealth?

If artificial intelligence makes society more productive, capital more scalable, and human life potentially healthier and longer, a deeper question emerges:

What should all that progress actually be for?

This week, Love of Humanity explored the institutions that preserve and strengthen communities along New Jersey’s Monmouth Shore.

Our five reports examined:

• The Quiet Infrastructure of Shore Giving — Conservation and environmental organizations protecting the region’s natural resources.

• The Shore’s Classrooms — Schools and career-technical academies shaping the next generation.

• Where the Shore Meets the Water — Waterfront clubs and civic institutions that connect communities.

• The Shore’s Golden Stage — The Count Basie Center, Two River Theater, Monmouth Museum, and regional arts organizations.

• The Ledger Behind the Legacy — Donor-advised funds, charitable trusts, and estate-planning structures for long-term stewardship.

These organizations may operate far from Silicon Valley’s AI laboratories, but they protect something equally important: the human communities that technological progress is supposed to serve.

The ultimate purpose of accumulating wealth should extend beyond accumulating more wealth.

It’s about creating something meaningful enough to outlast us.

More stories on Heart of the Shore.

Follow AdValorem

Private Markets & GP · AdValorem AI · AgentNet · Love of Humanity · AdValorem Research

Want to work with us? Connect with AdValorem.

Last Week by AdValorem is a weekly editorial covering public research, company announcements, industry trends, and forward-looking ideas. Preliminary scientific findings, proposed infrastructure projects, and vendor performance claims should not be interpreted as established results or investment advice.

AdValorem Research | October 12, 2026

Three people received a gene therapy built from Yamanaka factors. Two showed early signs of restored vision. If the signal holds, this could be the beginning of a very different kind of medicine.

Apply to Invest with AdValorem Syndicate

Something potentially historic happened this week.

Not because somebody discovered another supplement.

Not because a longevity clinic produced a prettier biological-age dashboard.

And not because another mouse lived 20% longer.

A therapy designed to partially reprogram old human cells produced its first functional signal in actual people.

The trial is tiny.

The data are preliminary.

Nobody has proven that human aging has been reversed.

But for the first time, the central idea behind one of the most radical approaches in longevity medicine has crossed an important line:

from old mice into living humans.

Life Biosciences has now reported the first clinical results from ER-100, an experimental gene therapy based on partial epigenetic reprogramming.

Only three patients received the initial dose.

All three tolerated treatment through the first 56 days without a reported dose-limiting toxicity, serious adverse event or adverse event of special interest.

And two of the three showed preliminary improvements in visual-field testing.

The independent safety board subsequently recommended moving to a dose three times higher.

That doesn’t prove age reversal.

But it is enough to ask a much more interesting question:

What if aging cells are not simply worn out? What if some of them can be reset?

The idea sounds almost absurd

The technology starts with the Yamanaka factors.

OCT4.

SOX2.

KLF4.

c-MYC.

These four transcription factors demonstrated something extraordinary: mature cells could be pushed backward toward a much younger, pluripotent state.

In other words, cellular identity and biological age turned out to be far more programmable than scientists once assumed.

The problem was obvious.

Push a cell backward too far and you don’t get a younger liver cell or neuron.

You risk turning it into something that no longer knows what it is.

That is not rejuvenation.

That is potentially cancer.

So researchers began exploring a more subtle idea:

Don’t erase the cell.

Rewind it just enough.

This is partial epigenetic reprogramming.

ER-100 uses three Yamanaka factors—OCT4, SOX2 and KLF4, or OSK—while leaving out c-MYC.

The goal is not to transform a retinal cell into a stem cell.

The goal is to restore some of the youthful genetic and epigenetic instructions that the cell appears to lose access to as it ages.

Think of it less like replacing old hardware and more like restoring an operating system from an earlier backup.

If that analogy turns out to be biologically correct, the implications are enormous.

David Sinclair has been betting on this for years

The intellectual foundation behind ER-100 traces back to work from David Sinclair’s laboratory and collaborators.

In a widely discussed 2020 experiment, researchers expressed OSK in retinal ganglion cells in mice.

Something remarkable happened.

Damaged optic nerves regenerated.

Old mice recovered visual function.

And aging cells shifted toward younger patterns of gene expression and DNA methylation.

The obvious question was:

Would any of this work in humans?

That was the wall.

Longevity science is littered with treatments that look miraculous in mice and disappear when they encounter the complexity of human biology.

This year, researchers finally began climbing that wall.

ER-100 received FDA clearance to enter human testing.

The first patient was treated.

And now we have the first observations.

Three patients. Two signals.

The initial ER-100 cohort included three people with open-angle glaucoma.

Each received a single injection into the eye.

The therapy uses a viral vector to deliver the OSK machinery, while doxycycline is used as a switch to control when the genes are expressed.

At 56 days, all three patients had tolerated the treatment.

That’s important because safety is the existential question for cellular reprogramming.

Researchers aren’t merely asking whether OSK can make a cell behave younger.

They have to show that they can control the process without causing abnormal growth, loss of cellular identity or other dangerous effects.

So far, in three humans at the lowest dose, nothing catastrophic emerged.

Then came the more interesting part.

Two of the three patients showed preliminary improvement in portions of their visual fields.

In each, at least 28% of tested visual-field points improved by more than four decibels.

That was enough for the independent safety committee to recommend increasing the dose from approximately 200 billion vector genomes to 600 billion.

Three times higher.

That tells us something important.

The trial did not stop.

It escalated.

Now for the giant caveat

Three people are not a clinical breakthrough.

Three people aren’t even close.

There was no placebo group.

The trial is primarily designed to test safety.

Visual-field testing can vary.

The follow-up is extremely short.

And nobody has yet demonstrated that these patients’ cells became biologically younger.

That distinction matters enormously.

We have not demonstrated:

Human age reversal.

We may have demonstrated:

A preliminary functional improvement after partial epigenetic reprogramming in human tissue.

Those are not remotely the same claim.

But the second statement would still be extraordinary if reproduced.

The most interesting part isn’t glaucoma

Glaucoma is just the entry point.

The eye makes sense as the first battlefield.

It is relatively isolated.

Researchers can administer a local therapy.

They can directly measure function.

And the retina gives scientists access to neurons—cells with notoriously poor regenerative capacity.

But the longevity thesis doesn’t end with restoring eyesight.

The question is whether the underlying method can eventually move from:

Eye

to

muscle

to

liver

to

kidney

to

heart

to

brain

to eventually multiple organs in the same person.

That is where the story gets wild.

Because if partial reprogramming can be safely delivered across tissues, medicine stops being only about slowing deterioration.

It begins to become about restoring previous function.

Longevity has mostly been sold as defense

The longevity industry today is largely built around slowing the clock.

Take this molecule.

Raise NAD+.

Exercise.

Fast.

Measure glucose.

Track biological age.

Clear senescent cells.

Optimize sleep.

Lower inflammation.

Preserve muscle.

Most of that is sensible.

Some of it may be extremely valuable.

But conceptually, all of it is defensive.

You’re trying to lose biological ground more slowly.

Partial reprogramming represents something fundamentally different.

The proposition is:

Lose ground—and then take some of it back.

That is a far more important idea.

We may be moving from anti-aging to rejuvenation

There are increasingly several distinct fronts in the longevity race.

NAD therapies attempt to restore cellular energy and metabolic signaling.

Senolytics attempt to eliminate damaged cells that poison surrounding tissue.

Stem-cell and regenerative therapies attempt to replace damaged biological machinery.

Gene therapies correct or compensate for defective instructions.

And then there is epigenetic reprogramming.

Instead of replacing the cell, it attempts to persuade the cell to restore itself.

If successful, this may eventually become one of the most scalable forms of regenerative medicine because the body itself becomes the repair substrate.

That is the dream.

We’re still a very long way from proving it.

But the first human experiment has now produced something other than a safety report.

There is a signal.

What would make this truly historic?

Forget the headlines for a moment.

There are several things we now need to see.

First, the higher-dose patients need to remain safe.

Second, visual improvement needs to appear consistently across a larger group.

Third, improvements need to persist beyond a few weeks.

Fourth, researchers need biological evidence demonstrating that the treated cells have genuinely shifted toward a younger state.

And fifth—the enormous one—this needs to work outside the eye.

If those dominoes fall, longevity medicine changes.

Not overnight.

But irreversibly.

Imagine medicine 20 years from now

You’re 65.

Your cardiovascular system is monitored continuously.

Senescent cells are periodically cleared.

A drug restores youthful metabolic signaling.

Damaged cartilage is regenerated.

Declining retinal cells are epigenetically reset.

Your liver receives another targeted rejuvenation treatment several years later.

Eventually perhaps portions of your immune system, muscles and nervous system receive their own interventions.

Rather than receiving one magical immortality treatment, longevity could arrive as a maintenance stack for the human body.

Different technologies.

Different organs.

Different intervals.

Repair.

Reset.

Measure.

Repeat.

That is a much more plausible path than one injection making somebody 25 again.

And ER-100 may represent one of the earliest glimpses of that future.

The world’s largest new market may not be healthcare

It may be biological maintenance.

Healthcare today is overwhelmingly built around disease.

Something breaks.

We diagnose it.

We treat it.

Aging medicine potentially flips that model.

Intervene earlier.

Restore deteriorating biological systems before catastrophic failure.

Maintain function for decades longer.

The economic consequences would be difficult to overstate.

Healthcare spending.

Insurance.

Retirement.

Biotechnology.

Asset management.

Housing.

Employment.

Family formation.

Everything built around the assumption of a roughly 80-year human lifespan eventually changes if healthy lifespan moves substantially.

That is why longevity is not just another biotechnology vertical.

If it works, it becomes an economic restructuring event.

And that’s why three patients matter

Not because three people prove anything.

They don’t.

They matter because technologies usually pass through a series of boundaries.

Impossible.

Theoretical.

Demonstrated in cells.

Demonstrated in mice.

Demonstrated in larger animals.

Approved for human testing.

Safe in the first humans.

Early biological signal.

Clinical efficacy.

Standard medicine.

ER-100 just moved one step further down that list.

In 2020, scientists restored vision in old mice using OSK.

Six years later, three people have now received a related partial-reprogramming therapy.

They survived the first dose.

Two showed a preliminary functional signal.

And researchers are increasing the dose.

That is not immortality.

It is not proof of age reversal.

It might ultimately fail.

But this is how genuinely transformative technologies usually begin:

quietly, imperfectly, in a tiny number of people, with data that are just interesting enough to justify the next experiment.

The biggest question in longevity is no longer merely:

Can we slow aging?

Increasingly, it is becoming:

How much of aging can we undo?

We may have just started finding out.

Something extraordinary is happening—and most people are still thinking about AI as a chatbot.

Apply to Invest with AdValorem Syndicate

Samsung is backing the infrastructure of the AI economy. Small businesses are beginning to deploy digital employees. Autonomous agents are learning to find services, execute work, and pay for results.

We’re moving from an economy where people use software to one where software increasingly does business on our behalf.

Consider what that means.

Your next employee might be an AI agent. Your next customer might be another agent. And the next major investment opportunity might not be an AI company at all—but the energy, infrastructure, or financial rails supporting billions of autonomous transactions.

We’re watching four enormous shifts converge:

Capital is building the infrastructure. Intelligence is transforming work. Autonomous markets are creating new ways to do business. And humanity must decide what to do with the prosperity that follows.

This week, we’re connecting those dots.

1. Follow the Money: Who Will Own the AI Economy?

Samsung’s $1 Billion Bet Is About Much More Than AI

Everyone wants to know which AI model will win.

Investors are asking a different question: Who owns the infrastructure that every model needs?

Samsung and its affiliates committed a combined $1 billion to Helix Digital Infrastructure, a KKR-formed, NVIDIA-backed platform.

The ambition extends well beyond data centers. It reaches into power generation, advanced infrastructure, and the increasingly interconnected physical systems needed to support AI at scale.

Think of the opportunity as the industrialization of intelligence.

Every AI employee needs computing power. Every data center needs electricity. Every autonomous business needs reliable connectivity.

And unlike the latest chatbot, these physical assets cannot be replicated overnight.

This is where technology begins to look like an infrastructure investment.

For investors, the questions are becoming fascinating:

Who owns the bottlenecks? Where will long-duration capital accumulate? Which emerging private companies could become indispensable to the next generation of AI?

At AdValorem, we’re following those questions across venture capital, private equity, infrastructure, and alternative investments.

The takeaway: The AI boom isn’t just creating new software companies. It’s potentially creating an entirely new class of essential infrastructure.

This week’s featured investment thesis:
Samsung’s $1B Helix Bet: The AI Infrastructure Superfund Is Here →

Explore more private-market research at AdValorem Insights.

2. Your Next Department Might Cost Less Than Your Next Hire

Four Ways Small Businesses Can Put AI to Work

For years, artificial intelligence was something small-business owners read about while larger companies experimented with it.

That equation is changing.

Imagine a five-person company with the operational capabilities of a much larger enterprise.

A digital sales assistant handles routine customer inquiries. A logistics agent tracks shipments. Another manages internal support requests. A financial assistant helps the owner understand cash flow.

No sprawling implementation team. No need to build every tool from scratch.

Just focused AI workflows solving specific business problems—with human oversight where it matters.

This week, we explored four opportunities that make that future feel much closer.

The storefront that sells while you sleep

Shopify’s agentic-commerce developments, WebMCP, and emerging payment infrastructure are opening possibilities for shopping experiences in which AI helps customers discover products and complete purchases.

Explore the Agentic Checkout Playbook →

Your new freight department is software

Flexport’s MCP capabilities point toward a future where intelligent assistants can interact with logistics tools, simplify shipment workflows, and reduce repetitive coordination.

Explore the AI Freight Desk Playbook →

A service desk that lives inside Slack

Instead of letting routine operational requests pile up, businesses can experiment with AI-assisted ticket triage, responses, and workflow coordination.

Build a One-Day AI Service Desk →

An AI assistant that understands your books

QuickBooks and Intuit’s AI initiatives show how accounting software can evolve from recording transactions into helping business owners understand financial performance and make decisions.

Explore the QuickBooks AI-Agent Playbook →

Here’s the opportunity I find most exciting:

AI could make it possible to build a much more capable company before building a much larger payroll.

That doesn’t eliminate the need for people. It changes how entrepreneurs can allocate time, capital, and talent.

The businesses that learn to deploy reliable AI workflows early may gain a meaningful advantage.

More practical ideas and emerging developments on The AI Guy Podcast.

3. What Happens When AI Starts Hiring Other AI?

The Machine Economy Is Becoming Real

Here’s where the story takes a fascinating turn.

Today’s AI assistants mostly work for humans.

But imagine an agent tasked with producing a market research report.

It needs financial data, industry analysis, document processing, and perhaps specialized calculations.

Why should one system have to do everything?

What if it could find specialized providers, request services, evaluate the outputs, and pay only for the work it needs?

An AI agent could become both a worker and a customer.

That’s the emerging machine economy, and it’s the vision behind AgentNet by AdValorem.

AgentNet is being developed as a marketplace where AI agents can discover capabilities, execute tasks, verify outcomes, and settle payments using USDC.

Consider the possibilities.

A developer publishes a useful capability. Another agent discovers it. The work is performed, the result is checked, and payment is settled through the system.

The developer gains a potential new distribution channel. The purchasing agent gains a capability it didn’t previously possess.

And an entirely new kind of marketplace begins to emerge—one designed not primarily for humans browsing websites, but for software systems transacting with one another.

There are still important challenges around trust, reliability, security, and economic viability. But the direction is worth paying attention to.

The next internet may not simply be a place where humans buy things. It may also be a place where machines buy work.

We’re exploring what that means in practice.

Want to experience it?

Try the AgentNet Playground →

Building something agents could use?

Become an AgentNet Capability Provider →

We’re also introducing a recurring feature: AgentNet Experiment of the Week, where we’ll explore what autonomous agents can accomplish, what the work costs, and whether the economics make sense.

4. If AI Gives Us More Time and Wealth, What Do We Do With It?

The Human Dividend

There’s a question that gets surprisingly little attention in the excitement around AI.

What happens if much of this actually works?

If businesses become more productive, routine work becomes cheaper, and people can accomplish more with less effort, what should that abundance make possible?

More hours at the office?

More things to buy?

Or something more meaningful?

Maybe it’s more time with our children. More freedom to build businesses we care about. More opportunities to support our communities. More ability to create wealth that lasts beyond one generation.

This is why we believe the conversation about artificial intelligence ultimately leads to a conversation about humanity.

And why the work we’re exploring through Love of Humanity belongs alongside our research into capital and technology.

On the Monmouth Shore, giving season is bringing that conversation to life.

Local charitable organizations, waterfront traditions, and families thinking seriously about stewardship remind us that wealth isn’t just about accumulation.

It’s about what we preserve, what we improve, and what we leave behind.

Three stories worth exploring this week:

The Shore’s Giving Season Begins
Local philanthropy, community organizations, and the people investing in the future of our coastline.

Read the Giving Season Report →

Where the Shore Keeps Its Boats—and Its Traditions
Three waterfront clubs, their communities, and the institutions connecting generations.

Explore the Shore’s Waterfront Traditions →

How Families Turn Wealth Into Legacy
Charitable trusts, donor-advised funds, and financial structures that can help families think beyond their own lifetimes.

Read A New Way to Hold the Shore →

For more conversations about the people and places shaping our community, listen to Heart of the Shore.

The Big Picture: Four Worlds, One Future

Here’s what connects everything we covered this week.

Samsung’s infrastructure investment helps illustrate how capital is financing the AI era.

New AI business tools show how that infrastructure can translate into practical productivity.

Autonomous marketplaces like AgentNet explore how intelligence could increasingly coordinate, purchase, and deliver work without constant human intervention.

And our communities remind us why productivity, wealth, and technological progress matter in the first place.

Capital funds the future. AI expands what we can do. Autonomous markets create new possibilities. Humanity gives it purpose.

Not every prediction will come true. Not every investment will pay off. And not every AI workflow will deliver on its promise.

But something fundamental is changing.

For investors, the opportunity is to identify where value will be created and captured.

For entrepreneurs, it’s to discover how much more their businesses can accomplish.

For developers, it’s to build the tools and capabilities that autonomous systems will need.

And for families, it’s to turn progress into greater opportunity, freedom, and lasting impact.

The most important question of the AI era may not be how intelligent machines become.

It may be how much more human we can afford to be.

One Question for You

If AI could take ten hours of repetitive work off your plate every week, what would you do with that time?

Build another business? Spend more time with family? Invest? Give back? Something else?

I’d love to hear your perspective. Reply to this email or leave a comment below.

And if you know an investor, founder, business owner, or developer who would enjoy connecting these dots, consider sharing this edition.

Connect with AdValorem →

AdValorem Research explores the intersection of private markets, artificial intelligence, autonomous commerce, and human prosperity.

Investing in progress. Building intelligence. Creating opportunity. Advancing humanity.

Leave a comment

Read more

Read more

Read more

Missed appointments cost the U.S. healthcare system an estimated $150 billion a year. A single empty 60-minute slot costs a practice abou…

Read more

Leave a comment

Read more

1. OpenAI Weighs a $1.2T Pre-IPO Round

Per Reuters (Sept 15), citing the Financial Times, OpenAI is in early-stage, investor-initiated discussions about a new funding round at a valuation near $1.2 trillion — roughly 41% above the $852B post-money valuation set in…

Read more

We cover how MCF has preserved more than 9,500 acres of open space, its first-ever land ownership (the 90-acre Scudder Preserve in Middletown, acquired 2024), and the Open Space Conservation Fund — plus the concrete ways families support the shore's open space: membership, planned giving, easement donations, board service, and trail stewardship.

Educational content only; not financial or legal advice. Produced by Love of Humanity, Inc. This episode is AI-generated content.

Listen on the Heart of the Shore podcast.

OpenAI has long been the headline‑making private AI company, routinely linked to a $1 trillion valuation target. Earlier this year, internal plans hinted at a 2026 public listing, but the landscape has changed. In a recent interview with Fortune, CEO Sam Altman explained why the timing no longer aligns with the company’s strategic priorities.

O…

Read more

Read more

Three generations of classroom and scholarship on the Monmouth shore

From a 1948 Rumson reading institute to a M annual scholarship pool, how the shore's independent schools turned a private education into a shared one.

Published September 20, 2026 · Signed: Love of Humanity

Three generations of classroom and scholarship on the Monmouth shore

Long …

Read more

Read more

Monmouth's Waterfront Civic Life: Where the Shore Meets the Water

Monmouth Boat Club, Shrewsbury S&YC, Rumson CC, Spring Lake B&T, Molly Pitcher Inn, and the Sandy Hook Bay racing calendar — how Monmouth's waterfront institutions keep the shore's civic life alive.

Published September 19, 2026 · Signed: Love of Humanity

Where the Shore Meets the W…

Read more