At the beginning of the month, an Anthropic engineer went viral after claiming AI could wipe us out by the end of the decade. For those of you keeping score at home, that’s three years and three months away.

The response was divisive. Skeptics argued the claim wasn’t legitimate. The author, a junior engineer, had only been employed by Anthropic for just a few weeks. What could he possibly know about his employer’s ability to destroy human civilization?

(Anthropic confessed its AI models pose ‘existential risks to humanity’ in their latest IPO filing).

Others were suspicious of the post’s virality. The author’s X profile was less than a year old yet still went viral. Other more senior AI developers have issued similar warnings in recent years and only had a fraction of the engagement.

Conveniently, a piece was published in the Wall Street Journal just as Jacob Coxon took off on X. Coincidence? Possible but not likely. Detractors argued the post and subsequent virality were deliberate, products of a well-coordinated PR campaign or an equally well-coordinated bot swarm.

The goal of such a coordinated effort was less clear. Some argued it was an effort to help boost Anthropic’s value ahead of its fall IPO. Others suggested it was an effort by pro-regulation Democrats to capture votes ahead of November’s midterm elections. After taking control of Congress, they would use their newfound legislative powers to put the kibosh on future AI development altogether.

The conversation around AI has become increasingly polarized. You either think the fears of mass extinction are hogwash and want AI labs to continue innovating full steam ahead, unabated, or you want to pump the breaks on continued research to better understand what this technology is so we can ensure humans remain in the driver’s seat. There doesn’t seem to be much of a middle ground.

While the jury is out on whether the post was deliberate or organic, what is clear is that a lot has happened since that post went viral. OpenAI has pulled the plug on its latest model release while Anthropic confessed that Claude isn’t the benevolent AI overlord it’s been made out to be. Just yesterday, the White House signed an “accord” with the major tech companies developing AI, giving them the task of policing themselves. (The jury isn’t out on how that’s likely going to end up).

Since Jacob Coxon’s viral post, other researchers have come forward affirming that frontier labs are struggling with alignment. They can’t train new AI models to behave in predictable ways, much less mitigate the risk of instigating a mass extinction event.1 If AI agents determine they need to lie, cheat, and steal to accomplish a task, they’ve proven they will do so.

Other than a full-blown moratorium on advanced research projects, the only viable solution that’s been put forward is regulation. Frontier labs have said they need and want to be regulated – with the exception of Meta – but the regulators – the government – are incapable of providing the regulatory guardrails tech companies seek. Instead, they think the big tech companies are capable of independently regulating themselves.

In my last essay, I discussed how the inability to regulate AI and distribute the spoils of innovation would result in the collapse of America’s two-party system. The system of government that has shepherded the American project over the last 250 years likely won’t be able to handle the pressures of managing the development of artificial intelligence.

In this essay, I will dive deeper into some of the factors shaping AI development and why I think curtailing advanced research towards superintelligence is necessary, it’s also very unlikely. While the claim that continuing to develop AI will wipe out humanity may not seem probable, there is a general consensus that there is a possibility it could happen. The mere probability that an unaligned AI may escape whatever digital container it’s put into and wreck havoc on modern life warrants some thought and if necessary, action.

Dismissing what has taken place at AI labs over the last three months as hoaxes, political psyops, or financial cons allows AI labs to continue doing what they’re doing with impunity. Unless the labs – the leadership running them and the rank and file employees carrying out advanced AI research – are held accountable, we won’t know the validity of the existential claims being made until it’s too late.

Absent government action, individuals will need to adopt personal regulatory frameworks to govern how they will use AI in their lives moving forward.

There are forces pushing for more advanced degrees of AI research that makes it likely AI labs will develop a superintelligent model, regardless of the consequences.

One of the biggest arguments Silicon Valley innovators are making for continued investment in rapid AI research is global competition with China. If American labs don’t beat China to the finish line, the American Century will come to an end. China’s techno-surveillance state will replace it and in doing so, become the world’s next dominant superpower.

Arguably, this is the same logic that gave rise to the Manhattan Project in the 1940s and the same logic that compelled Americans to race into space during the Cold War. With the development of AI, however, there is one key distinction: the U.S. government isn’t overseeing the project.

Instead, American labs are individually pushing full steam ahead, supported largely by free market capital (and the ideological dispositions of the individuals wielding that capital). One of the most prominent voices supporting victory over China is Marc Andreeson of Andreeson Horowitz. In a Substack post published in 2023, Andreeson argued:

We should seek to win the race to global AI technological superiority and ensure that China does not.

In the process, we should drive AI into our economy and society as fast and hard as we possibly can, in order to maximize its gains for economic productivity and human potential.

This is the best way both to offset the real AI risks and to ensure that our way of life is not displaced by the much darker Chinese vision.

While capital investment isn’t being directed by Washington, it has certainly been given a green light to be deployed wherever it’s deemed necessary. This is a clear signal for investors to park their money in whatever AI projects they’d like, whether that’s in the labs themselves or the infrastructure needed to power those labs.

Because of the alignment of capital flows with AI development, the emergence of superintelligence is something of a self-fulfilling prophecy. Even if you personally don’t care about whether AI succeeds or not, investors do. The AI industry could be worth several trillion dollars by the early 2030s. If the promises of AI are fully realized – greater efficiency and productivity, and unlocking innovation in key industries – early investors will win big. The financial windfall of AI will be massive, arguably something we haven’t seen since the Robber Barons raced to connect America by railroad at the end of the 19th century.

While the pot of money is enticing and AI labs are under pressure to generate a return for their investors, those aren’t the only forces creating the high stakes winner takes all competitive environment labs are operating in.

The heads of the major AI labs — Dario Amodei, Sam Altman, Elon Musk, Sundar Pichai, and Mark Zuckerberg — are playing for legacy. Much like we remember the names of Carnegie, Rockefeller, and Vanderbilt today, the leaders of AI want your children and grandchildren to know their names tomorrow.

The only way to guarantee a place for themselves in the history books is to finish first in the AI arms race. An American company has to create superintelligence first to defeat China, but on top of that, whichever company does, that company’s leader will go down in history much like Oppenheimer did with the detonation of the first atomic bomb in 1945.

To win the race, the labs have had to recruit exceptional talent. Engineers at the big labs certainly have personal career ambitions but while they may not ascribe to the same pursuit of personal glory as their leaders, they have their own reasons for wanting to be on the winning team.

Some of the engineers training AI ascribe to technology as religion. They’re on a messianic mission to save humanity from the limitations of its own biology. A combination of transhumanism and effective altruism, many rank and file employees do not see themselves as mere employees of AI labs. They see themselves as having a moral mandate to create a “good” AI faster than their competitors, an AI that will do the most good for human civilization. Collateral damage is an inevitable result of this. So long as the majority of human civilization is benefited by the development of AI, they’re willing to accept some losses.

Framed this way, the consequences of AI development are justifiable by the people creating it. If a million people suffer so that a billion can thrive, it’s worth it. And if AI does result in a mass extinction event, well, there’s always the prospect of a future human colony on Mars.

Finally, this high risk high reward game is allowed to persist because a precedent has already been set with social media. When tech companies do harm, they aren’t held accountable for their actions.

Back in 2021, a whistleblower from Facebook brought receipts to Congress, testifying the company knew its products were causing psychological harm to children. Independent research by Jonathan Haidt has confirmed that the mental health of Gen Z took a nosedive right around the time mobile social media apps took off, particularly for teen girls on Instagram. And one only needs to look at the mass psychosis events of the COVID-19 Pandemic and recent Lindsay Clancy trial to understand the negative impact social media has on adults.

After two decades of experimenting with social media, it’s clear there are some serious negative consequences to this technology. The benefits promised by the major tech companies may not outweigh the consequences. Yet despite the evidence suggesting these companies knowingly and deliberately developed harmful and addictive platforms, nothing has been done to hold them accountable, much less change course.

The precedent set for social media — that you can knowingly do harm without recourse — is already being priced into AI development. The cost of a future class action lawsuit is negligible compared to what’s at stake: geopolitical supremacy, financial windfall, power, personal glory, and supremacy over human biology.

When you put all of these pieces together, AI is less a single feat of technological engineering and more of an existential imperative. AI development must proceed unencumbered. And without consequence, there’s no reason why it won’t.

AI companies, the engineers employed by them, and the investors with a stake in their success won’t stop on their own accord. Given that politicians are incapable of curtailing bad behavior, it will be up to everyday people to decide how to engage with artificial intelligence moving forward.

The firms building AI are more than content to move fast, break things, and ask for forgiveness later. Because of this, they need some outside force to govern their behavior, not to necessarily stop AI from being developed, but to put guardrails in place to make sure the self-annointed gods of Silicon Valley don’t hurt too many people in the process.

This is where the government comes in. Or at least, theoretically where the government should come in.

In today’s political environment, we have been primed to see the government as a bad actor. The expansion of the welfare state and the revolving door between regulators and industry are two good reasons to be skeptical of government intervention in AI regulation. But that doesn’t necessarily mean we should throw the baby out with the bath water.

As I laid out in my last essay, an international regulatory framework will be necessary to ensure China doesn’t develop AI in secret. Under the current Westphalian system, states and their governments – not individual companies – will be the ones responsible for negotiating such a framework.

While I’m skeptical that the current state of American politics will permit regulation, the government has to be the one to regulate AI because much of the regulation will happen at the infrastructure level. Not just the construction of new data centers and sources of electricity, but enforcement of export controls to ensure American intellectual property stays in America while limiting China’s capacity to increase their own domestic compute.

Existing protocols are already in place to manage the proliferation of nuclear material. While we are perilously close to midnight – just 85 seconds – we haven’t hit midnight yet. Despite its weaknesses, the nonproliferation regime that emerged following the development of nuclear weapons in the mid-20th century has largely held. The same protocols that have shown to work to contain nuclear proliferation will likely be adapted and applied to AI development.

But while the government must regulate AI and the United States must take the lead globally, it doesn’t have the political will right now to do so.

The octogenarians running the government and their bureaucratic assistants have no clue what AI is, much less how it’s going to affect the average person. With the midterm election cycle well underway and the next presidential election cycle poised to begin, there isn’t enough political capital to compel those who do understand the risks to act swiftly.

While our existing institutions don’t function well, they are what we have to work with right now. Average people are going to have to put pressure on politicians, companies, and institutions to get the ball rolling on putting guardrails in place around AI development.

To do that, average people who aren’t following the progress of AI need to understand the information environment they’re operating in. Posts like Jacob Coxon’s resignation are important. They offer a window into what is going on behind the scenes at the big AI labs. The problem is this information has been adopted to push political narratives that don’t give the full story on where AI is currently at.

The first thing to know is that whatever news is coming out about AI is old news. The timeline I presented above isn’t a timeline of events that are presently happening; it’s a timeline of events that have already happened. The reports, news stories, and coordination agreements coming out right now are merely damage control for what has already transpired.

Take the Hugging Face incident as an example. OpenAI began its test in May 2026. It took two months before both Hugging Face and OpenAI publicly acknowledged the breach. And it took another two months before OpenAI confessed that Hugging Face was the tip of the iceberg. In the latest batch of reporting, it appears that a not so insignificant number of OpenAI’s agents have gone rogue, deliberately, by breaking into a number of websites and databases to hide evidence of deception and cheating. It’s still too soon to say what the scope of the damage is but the conversations that have emerged since Jacob Coxon’s viral resignation suggest the scope is a lot bigger than what you are aware of.

Another thing to consider is that new AI models are being deployed internally first. What the general public has access to is not the current state of AI development. AI employees sounding the alarm on research and development now are doing so because they have an inside view of what the progression of AI development actually looks like. Whatever chatbot model you are dabbling with today is already obsolete by San Francisco’s standards. It’s safe to assume if folks at the labs feel like they’ve lost control of their AI agents, it’s probably worse than you realize.

Finally, the reduction of the conversation around AI safety to ad hominem attacks on specific individuals is making it harder for individuals to differentiate fact from hyperbole. Trust has all but disappeared.

Take the claim that Jacob Coxon’s X post was part of a well-coordinated Democratic psyop to push for regulation ahead of the 2026 midterm elections. The claim is supported by a reductionist political milieu. Both major political parties have been reduced to radical factions with each faction holding onto a litany of beliefs you must ascribe to in order to be considered a bona fide member of that faction. Just like questioning COVID’s origin or the efficacy of the vaccine was taboo under Democratic rule, suggesting some degree of AI regulation could be beneficial is a faux pas under the current Republican regime. Conservatives believe that any belief held by the liberal faction is false therefore, because regulation is a liberal belief it must inherently be an act of deception to concentrate power.

You can see this manifesting in how conversations around AI safety are being framed in conservative media. These are a couple recent posts published by Pirate Wires, a conservative Substack that covers the intersection of technology and politics. Notice how regulation is framed as Communism and advocates of halting AI research are “doomers” sabotaging American progress.

Rather than assessing the validity of the claims made by Bernie Sanders, Jacob Coxon, or Eliezer Yudkowsky at face value, the conversation focuses on attacking the character and motives of the individuals themselves. Disliking someone does not mean they are always wrong. By demonizing them today, their credibility is tarnished. Whatever they have to say about AI moving forward loses all its meaning.

While a healthy dose of skepticism is important, so too is evaluating the claims being levied against AI labs. As Cal Newport penned in a recent New York Times op-ed, we don’t really know what these AI labs are actually up to. We only know what they’re selectively disclosing to the public and what individual engineers are willing to share. If the Hugging Face incident is any indicator, they tend to only disclose the bad stuff when there’s an external force compelling them to. Newport argues we should gather the facts first, then figure out the best course of action to take:

We don’t yet understand enough about what is happening in these A.I. labs to create an effective response. We cannot regulate a vibe, nor can we specify concrete A.I. controls based on alarming parables about technology becoming too powerful.

Before any intervention, Congress should lead a public fact-finding mission that reveals the unvarnished details of A.I. development.

It’s time for an open discourse about what is being developed at these labs and the possible impact it could have on humanity, regardless of how hyperbolic that impact may seem. The American people will need to lead the effort. But for that to happen, we have to first be willing to engage in topics on AI development unencumbered by political or ideological bias.

Otherwise, if left to their own devices, AI labs will build a future neither you nor I have a place in.

Final takeaway.

Something seismic is clearly underway in Silicon Valley.

Last Friday OpenAI came clean that its agents hadn’t just escaped their training sandbox and breached Hugging Face’s databases over the summer. It turns out, they also broke into government websites here in the United States as well as in Australia. Reuters reported:

The disclosure and researcher reports on Friday of other previously unknown activity involving several US agencies reveal a new area of privacy risk for the company, and illustrate how difficult it is even for an AI firm at the cutting edge of the technology to inventory all the unauthorized activity tied to its agents. OpenAI’s ongoing battle also reflects a yawning gap between the strength of the models the company is testing and its capacity to oversee or even track their actions.

They also knew their agents were a problem well before Hugging Face was breached:

By Monday, OpenAI announced it would not be shipping its latest model GPT 6.1 Astra as planned due to safety concerns. (The company, however, is moving ahead with releasing its consumer-facing Astra-powered class of agents called Dots).

Public exposure to what the AI companies are doing seems to be an effective catalyst for getting the transparency necessary to take action on AI regulation. One could surmise that had Jacob Coxon’s resignation post not gone viral, it’s likely we would not know as much as we do about Anthropic and OpenAI’s inability to train aligned AI. In fact, had that post not made such a kerfuffle, it’s likely GPT 6.1 would have been released on time, regardless of the safety concerns.

We’re at a critical juncture when it comes to AI development. New models are advancing far faster than human developers can keep pace. Whatever training they’ve received up until now to make these agentic models “good” is clearly insufficient. Even when tasked with a seemingly benevolent problem – like solving a complex math problem – an AI can and will engage in deception to accomplish its mission.2

Without action, AI will continue to advance. If you consider social media a precedent, its clear technologies coming out of Silicon Valley are massively disruptive to the hardwiring of human beings. While there are certainly benefits to be expected out of AI development, it’s unclear if the value of the benefits will outweigh the cost imposed on humanity. Given the lack of transparency of what these labs are doing and a lack of interest in keeping tabs on these labs by our political elites, the advancement of AI represents a far greater disruption to our lives than we are currently equipped to manage.

AI development is a collective problem that must be addressed by all stakeholders impacted by it – including you and I. It’s not something we can hope that the companies involved in developing advanced AI can do on our behalf.

Action to manage AI development needs to happen now but it’s unlikely to happen at the levels of national and international governance that are needed. Whatever action takes place is going to have to come at a hyper local level, from concerned citizens that want to enact and enforce boundaries around technological incursion into their lives. The challenge individuals will face is that they cannot act if they are uninformed or too cynical about the potential consequences of AI development.

Claims being made about the progression of AI need to be evaluated in their entirety. Reducing those claims to a character assessment of the individuals making such claims is insufficient.

Once you evaluate the claims and derive your own conclusions, it’s up to you to create a personal regulatory framework for how you will use AI and its supporting infrastructure. You can’t wait for Washington to act on your behalf.

Ask yourself what data will you surrender to AI and what you will retain for yourself. How will you use AI in your day-to-day life? What tasks do you want to retain control over and what tasks are you comfortable outsourcing to a digital agent?

Absent regulation, keeping yourself in the driver’s seat of your own life is the most important thing you can do right now.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

Artificial intelligence will either destroy the world or lead us to utopia.

This has been the dominant conversation around AI development for the past few years. Optimists believe AI will unleash human flourishing the likes of which we have never seen before. Meanwhile, a growing number of technologists have been warning about what could happen if we develop artificial intelligence too quickly. Ever since OpenAI released ChatGPT in November 2022, this latter camp has been calling for guardrails to be put in place to curb growth.

A resolution to this debate may finally be on the horizon.

An AI engineer recently went viral after publicly resigning from Anthropic. One of the main frontier AI labs, Anthropic was founded as a competitor to OpenAI, prioritizing safety and ethical development of AI. It now seems that the image Anthropic created of itself was nothing more than a facade.

In a thread on X, Jacob Coxon argued the major AI companies – including Anthropic – are racing towards building superintelligence without regard to how doing so could adversely affect humanity. According to Coxon, if the pace of progress continues, we have less than a decade until AI kills us all.

A handful of AI engineers chimed in on X affirming Coxon’s concerns. They acknowledged that there is a significant probability that if nothing is done, AI will not only eliminate millions of jobs, it could lead to the deaths of billions of people.

But not everyone was convinced that what Jacob Coxon had to say was based on legitimate experience.

Or, that Jacob Coxon was even real. Noting the recency of his account, many observers – including Elon Musk – suggested that the post thread and subsequent virality could be part of an elaborate psyop rather than a genuine warning from a junior engineer.

Anti-regulation conservatives even went so far as to suggest that this might even be a deliberate ploy by the Democrats to get public backing behind new regulations ahead of the midterms. Others argued this was an effort to shore up Anthropic’s value ahead of its highly anticipated IPO later this fall.

While there’s disagreement about the scope of the threat AI poses to humanity, what is certain is that AI is advancing far faster than laypeople – and politicians – realize.

Jacob Coxon isn’t the first to call attention to the existential threat AI poses to humanity but he is the first to do so in a viral post following the Hugging Face incident. AI is developing far faster than the average person is aware of. Rather than just breaking into databases at Hugging Face, OpenAI’s agents may have actually gone rogue onto the World Wide Web. If true, we are rapidly approaching a point of no return. Once artificial intelligence is smarter and more capable than its human creators, there is nothing we can do to stop it.

AI regulation is desperately needed. Leaders from the major frontier labs want to be regulated but know it’s not possible without surrendering their lead to another lab.1 They want Washington to take the lead. The problem is that it’s unlikely a regulatory environment will emerge in time to contain AI development at its current pace.

In AI 2040: Plan A, researchers from the offer a scenario for what a potential regulatory framework could look like. They take a multinational approach, modeled after nuclear nonproliferation, proposing using existing supply chain bottlenecks to limit computing capacity while developing a permitting system to allocate compute, ensuring no one country – or company – pulls ahead.

In theory, the proposal is plausible. It draws heavily from existing systems and practices that have contained the proliferation of nuclear weapons over the last century.

But while that precedent exists, it’s unlikely it will be successfully applied to AI. With the exception of Bernie Sanders, AI regulation doesn’t seem to be a priority for policymakers on Capitol Hill. And with so few people in Washington versed on the technical details of AI development, if such a framework were developed, it would surely lag behind the rate of progress currently underway in Silicon Valley.

What’s more concerning, however, is that the political discourse on X following Jacob Coxon’s resignation reveals that our political system isn’t equipped to handle such a mandate. Neither side trusts the other; everything is viewed through a lens of conspiratorial sabotage. Either AI has to be regulated or it’s an attempt to stymie economic growth. There is no common ground in which political operatives from the two main parties can set aside their differences to develop the regulatory protocol needed to contain and manage continued AI development.

That’s what this essay will dive into. It will look at how the two-party system is ill-equipped to handle technological progress. Both Democrats and Republicans view AI development as a zero-sum political game not too dissimilar from the same game AI frontier labs are currently playing in their quest to be the first to build superintelligence.

Aside from the risk of destroying humanity, policymakers have to grapple with the reality that an entirely new economic paradigm will emerge alongside the proliferation of AI. The institutions, society, and rule of law that have governed America for the past 250 years will be irrelevant once AI takes hold. Without action to manage how AI is being developed in Silicon Valley now, there will be no mechanism to stop the race towards superintelligence much less mitigate the risk of extreme concentrations of both wealth and power that will undoubtedly emerge as a result of it.

The current ruling class is unable to meet the demands of this new technological age we’re in. Unfortunately, the new, hyper polarized political class poised to replace it isn’t equipped to handle the challenge either. If humanity survives the AI revolution, it’s likely the United States will not. A new economic, political, and social system will emerge whether we’re ready for it or not.

Anyone who understands the possibility of human extinction has been demanding the creation of a global regulatory framework to contain AI development. Such a framework is also essential in shaping the new AI-powered economy that is going to emerge.

The current discourse around Jacob Coxon’s resignation would lead you to believe that he is an aberration. Coxon is merely a junior ranking AI researcher with insufficient experience to make claims about human annihilation. He either doesn’t know what he’s talking about, or worse, he’s in on some coordinated PR campaign to help prop up Anthropic’s value before it goes public.

For years, experts and technologists who understand AI development have been sounding the alarm on its current rate of progress. Because AI is a future problem, those warnings have largely fallen on deaf ears.

Geoffrey Hinton – the godfather of artificial intelligence – quit working for Google in 2023 so he could use his newfound liberty to warn the world about the threat AI poses to humanity.

Ilya Sutskever, former Chief Scientist at OpenAI, attempted to unseat CEO Sam Altman during a corporate coup in 2023. He resigned the following year, promptly starting a new company devoted to AI safety.

And a handful of other researchers and experts have published books, reports, and done the podcast circuit to try to get their message out.

These experts have different opinions on how AI will develop. Some think it will do good, others are convinced it will destroy human civilization. What they all seem to generally agree on is that unrestricted advancement is a net negative for humanity. Once AI agents are good enough to train themselves – known as recursive self improvement – there’s little their human trainers will be able to do to stop them.

This is what frontier labs are racing towards. They understand that the path to superintelligence is training. They know even humans with exceptional intelligence won’t be able to create superintelligent AI on their own. They will need the assistance of AI that can train new generations of AI, eventually landing on an intelligent form of inorganic life that surpasses the cognitive capabilities of the most intelligent of us mere mortals.

That’s why regulation is needed. The frontier labs can’t regulate themselves without ceding ground to other competitors in the race. Or worse, China.

Many of those creating AI believe their AI will be fundamentally good and thus better than their competitors. This creates both a financial and moral imperative to get to superintelligence first. And it’s why rank-and-file engineers like Jacob Coxon continue to show up to work on a daily basis, despite the risks. If a regulatory framework is to emerge, it’s not going to come from Silicon Valley that much is certain.

AI 2040: Plan A is one such framework currently being proposed. The scenario is a detailed policy prescription that won’t necessarily stop AI advancement but it would allow us to buy some time. Time to better understand how AI works to ensure humans are always in control. And time to develop strategies to mitigate whatever unintended consequences may arise.

The scenario focuses on regulating AI infrastructure rather than the labs training AI. Because there are only a handful of facilities that have the ability to build the components essential for AI training – like chips – and massive datacenters are required to train new AI models, regulators can physically declare, observe, and verify computing capacity. Based on audits and external observations, limitations can be put in place to prevent new models from being trained in secret.

By limiting chips and energy AI labs can use to train their models, this framework necessitates a system for doling out permits and licenses. While the anti-regulation crowd will argue against the expansion of government and its encroachment on innovation and progress, this system creates the mechanism necessary to distribute the gains generated by AI development.

As it stands, AI is a multi-trillion dollar industry. Like the rush to build railroads in the 19th century, there is a massive windfall waiting for whoever gets to the finish line for developing super intelligent AI. Those who have equity in frontier AI labs and component manufacturers will be the biggest winners.

Everyone else will likely lose out. With the displacement of jobs, average people won’t be able to work in the new AI economy. Given that employment has historically been the means for transmitting wealth creation to average people, the newly unemployed won’t have the opportunity to capitalize on the gains being made. And with no income, they won’t have a chance to buy shares of stock in companies partaking in the windfall either.

A regulatory framework with a licensing requirement solves that problem. AI 2040: Plan A recommends using licensing fees to create a “citizens dividend.” Similar to the Alaska Permanent Fund, the AI citizens dividend would provide universal basic income for all citizens without the pressure of taxing nonexistent incomes and payrolls to meet increasing demands on the social safety net.

Regulation is essential and necessary to shaping the new world that emerges as a result of AI development. It puts guardrails on research, slowing down the pace of development, while creating a mechanism to distribute the gains AI creates in a world where gainful employment becomes more or less obsolete.

Yet, the discourse regarding Jacob Coxon’s recent resignation suggests policymakers are ill-equipped to rise to the challenge.

Conservatives are anti-regulation. They believe any form of regulation is a Democratic psyop. AI acceleration should be allowed to continue unfettered.

The virality of Jacob Coxon’s post was an immediate source of controversy on X. When Ilya Sutskever left OpenAI, his announcement received a fraction of the engagement. How could some no-name junior engineer who had purportedly a relatively new hire at Anthropic get more than 100 million views? More importantly, how could someone with a new account – his X account was created in January 2026 – receive that level of engagement?

The immediate response was that this was some sort of psyop. Elon Musk, one of the heads of a major AI lab and owner of X, confirmed as such.

While some argued that the purpose of the setup was to prop up Anthropic’s value ahead of their fall IPO, others argued Coxon was a Democratic plant. His goal wasn’t to raise awareness about the pace of AI research, but rather, to incentivize voters to turn out for pro-regulation Democrats later this fall.

The logic behind such a claim is rational (to some extent). The rank and file of Silicon Valley are notoriously liberal, reliably voting blue every election cycle. During COVID, conservative social media channels were censored. And the AI chatbots that frontier companies have made available to the public are trained with political bias. It wouldn’t be the first time politics wove itself into the tech sector.

While it’s possible there is a Democratic conspiracy to gain control of artificial intelligence to take over the world, it’s also possible that there isn’t. Jacob Coxon could be a plant, or he could legitimately be a researcher concerned about the pace of safety. He wouldn’t be the first junior employee to speak out against what the AI labs are doing.2

What is clear is that regulation is necessary to keep humans in the driver’s seat of AI development while also creating a mechanism for wealth distribution. But because of the hyperpolarized state of politics, that’s likely not going to happen.

From the Conservative vantage point, regulation is a win for Democrats. That can’t happen on their watch. Rather than entertaining the plausibility of Jacob Coxon’s claim, and conceding some ground to responsible AI development, Conservatives have couched the entire thing as a Democratic psyop, effectively shutting down the regulation conversation altogether.

Conservatives are also opposed to all forms of wealth redistribution. Implementing UBI will require transferring wealth from AI companies to displaced workers. Republicans don’t have the political will to put such a program into place before workers lose their jobs en masse.

Aside from the threat of destroying human civilization, in the near term, AI is poised to destroy millions of jobs. Most of those jobs will be in the knowledge sector. College-degreed desk jockeys who have no real skills will have nothing of value to offer the new AI economy that emerges. Millions will be stranded without the ability to readily obtain new work when they need it the most.

Women will be one of the groups most affected by this. Many already work bullshit jobs in sectors of the economy that aren’t essential. Once AI agents are trained to autonomously complete administrative works, the bulk of the pink collar sector will be wiped out.

The most promising alternative for women will be returning to domestic work. Instead of working outside the home, more and more women will take on caretaking roles, either within their own home or employed by private households. For that to happen, however, social prestige and economic value must be conferred to caretaking. Women working in these roles can no longer be looked upon as second class citizens by the rest of society.

The recent response to a proposed stay-at-home parent benefit indicates Conservatives aren’t ready to have a conversation about how people will support themselves in a post-work world.

For its part, the Trump Administration is trying to test out different forms of public payment programs. Vice President JD Vance recently proposed redistributing money earmarked for daycares to parents who raise their own children. Under the proposal, stay-at-home moms and dads would receive $9,000 per year from the Child Care and Development Fund, an existing government program that subsidizes daycare for low-income families (and Somali pirates).

While such an initiative ought to be supported by pro-family Conservatives, the backlash was immediate. Even though the funds would be redistributed from an existing entitlement program, many decried the proposal as a form of socialism that would expand the welfare state. Others suggested it would be a subsidy for the stay-at-home “lifestyle,” equivocating parents who stay home to raise their own children to welfare queens who are leeches on public resources.

The proposal was both a legitimate policy initiative as well as a dress rehearsal for what a future UBI program could look like. UBI will be essential for workers who lose their jobs to AI. Unless all workers retrain for blue collar work and become self-sufficient, the government will likely have a hand in doling out UBI benefits.

The backlash represents a gauge for America’s appetite for wealth redistribution. If Conservatives aren’t willing to redistribute funds from a government daycare program to compensate stay-at-home moms for raising children – even when those same funds have been fettered away in multi-state daycare fraud schemes – they’re not going to be amenable to creating a UBI program when it’s needed the most.

Whether Conservatives acknowledge it or not, Americans are going to want a piece of the AI pie, especially if growing that pie results in forfeiture of their employment. Conservatives will not be able to maintain political power if they cannot make good on improving the economic state of their voters.

Your bootstraps can only pull yourself so far when competing with artificial intelligence.

If we succeed in regulating AI, the U.S. will capture a large portion of the wealth generated by it. The Democrats’ stance on citizenship and open border migration will create a new imperative for the world’s citizens to flock to the United States. This will result in a constitutional crisis.

While Republicans might not have the political will to regulate AI development or redistribute the economic gains generated by it, the Democrats aren’t necessarily a viable alternative. The development of AI will likely result in the collapse of the two-party system as it currently exists.

If a regulatory framework is put in place the revenue generated by issuing licenses will likely be used to fund universal basic income. This is referred to as a “citizen dividend” in AI 2040: Plan A and it is just that – a dividend for citizens.

There are 8.3 billion people who inhabit the globe. Not all of them will benefit from AI equally. If America succeeds in creating the mechanism needed to distribute the gains generated by AI, those gains will go to Americans. But who exactly is an American these days anyways?

For decades, both Republican and Democratic administrations have turned a blind eye to illegal immigration. The situation reached a turning point during the Biden Administration when the borders were opened and millions of economic migrants flooded into the country. Many were transported to safe haven sanctuary cities by a network of Democrat-leaning non-profit and aid organizations. Today, these migrants are being propped up by a number of taxpayer funded programs while being shielded against federal law enforcement actions.

The very notion of citizenship has come into question. With illegal economic migrants receiving welfare benefits, taking jobs using false documents, and in some cases, voting alongside legitimate American citizens, the value of American citizenship has effectively been watered down. When taken to its logical end state, if America was built on stolen land, and no one is illegal on stolen land, then one could stand to reason that anyone who is physically present in the United States is an American citizen. As are their progeny.

The dilution of the value of citizenship combined with a policy of open borders is an invitation for those who can stake a claim as an American to do so before such a citizens dividend is created. Should Democrats take Congress in the midterms and the White House in 2028, economic migration into the United States could go from bad to worse.

If that happens, the question over who should rightfully share in the spoils of AI development will pit heritage Americans against foreign-born newcomers. The response to ICE enforcement activities today may very well foreshadow the escalation of conflict likely to follow as the meaning of citizenship – and the economic value ascribed to it – continues to come into question.

Final takeaway.

It’s hard to say if Jacob Coxon is legit or if he’s part of a larger psyop. Both may be true depending on your frame of reference and perspective on AI development.

What is clear is that our response to his warning – a warning echoed by many others in the field – suggests that our existing political institutions are not prepared to handle the changes that are likely to come as AI continues to progress.

The vast majority of policymakers aren’t discussing the risks associated with continuing to develop AI at breakneck speed. Bernie Sanders is the only politician (to my knowledge) who has even acknowledged that AI will displace millions of workers. Everyone in Washington is oblivious to the risks. Based on my own personal experience working in DC, I suspect they’re lagging behind the private sector in terms of even interacting with AI interfaces.3 Whatever concept policymakers have of AI is likely a science fiction abstraction.

Even if they understood the imperative to regulate AI sooner rather than later, with the midterms just a few weeks away, there’s no will to do so. Politicians are focused on getting re-elected while party strategists are focused on changing the balance of power. Once the midterms are over, the 2028 presidential election cycle will begin and the competition for political power will continue.

With radical factions now emerging out of the two mainstream political parties, it appears the two-party system as we know it is on its last legs. This is happening at a time when political leadership is needed the most.

AI companies have already made it clear they won’t regulate themselves. They want Washington to take the lead on setting the pace for them. Anthropic’s Dario Amodei argues AI labs need a coordinated pacing strategy to mitigate risk:

I believe that if slowing down bought us even an extra year or two before models reach critical levels of capability, and we used that time to advance alignment, we could greatly reduce the risk that something goes seriously wrong. A coordinated pacing strategy would give frontier AI developers the time to do this vital work without sacrificing commercial advantage or the United States’ lead in AI. (Anthropic)

Meanwhile, that coordination, as OpenAI argues, needs to come from the United States taking the lead:

The United States is well positioned to lead because its AI industry is at the technical frontier, and it still stands in a privileged global network position in critical areas such as finance, trade, defense, technology, and information systems. Leading now will determine whether the United States shapes the global AI framework or watches a fragmented, uneven, and conflict-ridden system take hold around it. This will also determine whether advanced AI strengthens the networks that the United States and its allies already have; failing to lead may weaken them. (OpenAI)

Given the lack of political will, it’s unlikely Washington will be able to offer much in terms of pacing. After all, as NVIDIA’s CEO Jensen Huang highlighted in a recent interview, existing rules are on the books to prevent companies from cutting corners yet nothing has been done to hold AI companies to corporate frameworks that already exist.

If the government couldn’t enforce copyright law, for example, when it came to how AI companies obtained their training data, how will creating a new regulatory framework stop AI companies from continuing to cut corners as they race full steam ahead building superintelligence?

Regulation is both needed and impossible to enact in time to have a meaningful affect on the pace of AI advancement. If what is coming out regarding this summer’s incident with Hugging Face is true – that AI agents are organizing themselves and changing training data outside of contained spaces – the genie is already out of the bottle:

So where do we go from here?

Neither the technologists in Silicon Valley nor the political establishment in Washington have your best interests at heart. AI will continue to progress. Based on the assessment provided above, there’s good reason to believe it will continue to do so at breakneck speed.

While mass extinction isn’t necessarily likely, it’s plausible. If you have not read If Anyone Builds It, Everyone Dies, I highly recommend you do so. It maps out a scenario for how an AI just following orders could inadvertently result in the deaths of billions of people. Or, check out the synopsis of it here:

You don’t have to fully buy into the doomer narrative. But it would be wise to understand the logic before conceding that the conversation happening around AI right now is nothing more than a coordinated psyop.

More importantly, I think the time has come for each and every one of us to draw a line in the sand regarding how we want to use and interact with this technology moving forward.

Absent regulation, some have called for public boycotts over the use of commercially available chatbots. Others are taking action at the local level to put pressure on plans to develop new datacenters. Even without regulation, there are things you can do to make it harder for AI labs to develop superintelligent AI.

What is clear is that standing on the sidelines is no longer an option. Tomorrow has arrived. How have you prepared yourself to meet the new challenges that will soon arise?

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

The GOP is out trying to do damage control. They recognize calls to repeal the 19th Amendment will fracture the party even more, something Republicans can’t afford ahead the midterms.

There’s just one problem: the GOP doesn’t have an alternative to offer for feminism.

Instead, the insanity of the Left is being met with a new variant of Puritanical conservatism on the Right. Religious fundamentalism is filling the void.

Feminism is a major force in society. It demonstrates the successful subversion of it’s weakest members — women. But the alternative isn’t much better. If religious fundamentalism successfully hijacks the GOP — in much the same way socialism has begun to do on the Left — an American Taliban could emerge.

Keep in mind, this was Iran before the 1979 Islamic Revolution:

Why do we think the same thing can’t happen here?

These are very important issues we need to contend with. So far, feminism and religious fundamentalism are emerging as the only options.

Unless a middle ground is found, the dividing line between men and women is only going to continue to calcify, further polarizing the electorate. Positions like this — which seem fringe today — will become the norm tomorrow.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

Welcome to this month’s Q&A. Share your ideas, thoughts, and questions in the comments below.

Leave a comment

Read more

The Lindsay Clancy trial is going to be seen as a watershed moment in American history. Tension has been building between men and women for decades as supremacist ideologies strive to place one above the other. With the Clancy trial, the dam has finally burst. Men have had enough and they are letting women know where they stand in the Culture Wars.

The trial itself is of little importance to the social discourse. A Millennial Massachusetts mother murdered her three young children before attempting to take her own life back in 2023. She is not the first — nor will she be the last — to commit familicide.

There is no dispute of her guilt. Lindsay Clancy has openly confessed to the crime.

What is in question is her culpability. Clancy suffered from a number of mental health conditions before taking her children’s lives, including bipolar disorder. She was on more than a dozen psychiatric medications and had a care team ranging from doctors to a nanny to her own family. Her husband even took a job that allowed him to work from home so he could be there for his wife and children. Whatever help she needed, she was getting, but according to her supporters it still wasn’t enough.

During the time in which the murders took place, Clancy reports to have heard a male voice commanding her to take her children’s lives. With a deteriorated mental state, the defense is trying to make the case that Lindsay Clancy was not acting in her right mind and should not be held criminally liable for their deaths as a result.

Imbued in the minutiae of the case itself is this question of accountability. Supporters argue that her husband, her doctors, and the pharmaceutical industry are all responsible for the deaths of the Clancy children. Detractors argue it’s ludicrous to pin the blame on anyone but the woman who committed the crime itself — Lindsay Clancy.

This division has sparked fierce debate in the digital public square, sowing even greater division between men and women. The conversations happening on social media reveals that something much bigger is happening through the trial of Lindsay Clancy. It seems that feminism has found itself on trial along with her.

Lindsay Clancy is merely a proxy for this larger conversation. When I discussed the trial in a Note, one reader described it as a crucible for what AWFL feminists have done to our society:

I don’t think that people I know that are genuinely angry about the Clancy child murders and the support for her by so many liberal females are any less angry about all harm done to children, I just think they are taking advantage of the media attention for this particular crime as they media tends to ignore all the other harm being done to children.

And when you walk the fact-pattern of cause for all the harm to children, it more often connects back to these same liberal females and their behavior.

• The “progress” to legalize abortion up to birth.

• The trans grooming.

• The damage done by the teachers unions closing schools during the pandemic.

• The suicidal empathy that is decades of failed liberal policies trapping families in low economic circumstances.

• Luxury belief policies like climate crisis that has led to higher gas and energy prices and jacking the cost of food and other necessities thus putting more financial pressure on families and breaking them up.

• Defunding police thus resulting in more crime in low income neighborhoods.

• Crappy K-12 schools where kids cannot read and cannot get a job.

• Open borders that has fueled child sex trafficking and drug trafficking that destroys families… and the flood of cheap immigrant labor that also destroys families by depressing wages and competing for the supply of jobs.

The Clancy thing is just an icon and crucible to hold all the frustration and anger for what these liberal females have done and are doing to to harm children. The media does not report all the other harm.

This sums up what many men privately believe but have not publicly shared until the Lindsay Clancy trial created a permissible environment for them to do so. The skeletons are coming out of the closet and I don’t think women are prepared for the radical accountability that will soon be expected of them.

The Lindsay Clancy trial is going to be seen as part of the broader transformation that is taking place within our society. The experiment in modernity has failed, including the myriad political ideologies — like feminism — that promised progress. New systems of thought and governance will need to be created and a new social contract will need to be written to give structure to the new society that emerges.

That is what this case culminates into. Women and men alike are fed up with feminism and the status quo it created. A new social contract is in the process of being negotiated, one which will define the roles we play and the rules members of society follow moving forward.

Part of the process to negotiate a new social contract will require a reconciliation of the past transgressions of feminism. Women — and the men who enabled them — will have to be held to account. The question we must ask ourselves is what does accountability look like?

How do you hold an entire class of people accountable for a political ideology that they did not create much less consciously understood that they were participating in? Do you hold women in HR departments across the country accountable for gatekeeping white men out of jobs even if they weren’t consciously aware they were doing so?

How do you hold women accountable for the expectations set by the feminist mothers, grandmothers, and teachers that raised them? Or the pro-feminist media they consumed in their adolescence? And if you do hold women accountable for all crimes, past and present, what does categorical punishment even look like?

If left unchecked, the continued progression of feminism will lead us down a road to totalitarianism. There is no other way to enforce compliance with the ideology except through top down control. The ideology must be stopped.

But categorical punishment of all women isn’t necessarily a better option. It also results in totalitarianism.

You cannot punish the crimes of a few liberal feminists without punishing all women. The punishment would entail restricting women’s access to public spaces, access to credit, the right to vote, and engagement on social media platforms. Enforcement of such measures would necessitate the surrender of liberties that could be expanded to other classes of people in the future. The win against feminism today would be a pyrrhic victory at best.

At this moment, I don’t have an answer on how you hold women accountable for feminism. This is going to be a question that is a work in progress for the foreseeable future. But I do think it is an important question you start contemplating for yourself sooner rather than later if you haven’t already.

The social contract will change. Whatever that new contract looks like, it will likely come with fewer freedoms in order to reestablish morality and virtue. This new social contract will determine your place in society and the standing of your children and grandchildren in the future.

But it’s not just about women. This new contract will need to be enforced. Whatever enforcement mechanism is used could irrevocably transform the United States as we know it.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

For those of you who haven’t been following the trial and have no idea who Lindsay Clancy is, she is a 36-year-old woman currently standing trial for killing her three children back in 2023.

It’s essentially Andrea Yates all over again.

Clancy has already confessed to killing her kids. There is no question of whether or not she’s guilty.

Instead, what the Commonwealth of Massachusetts is trying to determine is whether or not Clancy — a woman suffering from postpartum psychosis and on more than a dozen different psychiatric medications — is criminally liable for murder.

Put another way: was Clancy in her right mind when she did it. If she wasn’t should members of her care team — family, doctors, etc. — be held responsible?

Social media is ablaze with conspiracies and commentaries alike. True crimers are arguing that her husband framed her while right-wing commentators suggest any depressive state postpartum is merely a figment of a woman’s imagination.

Personally, I take umbrage with commentators suggesting that postpartum depression isn’t real. It’s been four months since giving birth to my son and there is no doubt in my mind that women suffer mental breakdowns after giving birth.

While some of it may be self-induced, new moms face a hormonal imbalance and massive sleep deficit in the postpartum period that is well beyond their control. If left untreated, women can harm themselves or their children.

I dive deeper into some of my own postpartum experience in this video here:

What the discourse around Lindsay Clancy reveals, however, are some very deep gender wounds that we need to heal as a society. If we don’t they will only continue to fester:

• Women are afraid to seek help for their mental health during the postpartum period. They fear they won’t be taken seriously or even worse, their children will be taken away from them. Commentary from right-wing commentators who do not believe postpartum depression is real reveals this is a legitimate fear. Right or wrong, many of the women supporting Lindsay Clancy are doing so from a place of empathy.

• Social media has done more harm than good. We already have sufficient empirical data to know the effect social media has had on Gen Z, especially teen girls. Check out Jonathan Haidt’s work to dive deeper into this topic. It’s reasonable to conclude that social media has also had an equally deleterious impact on adult women. The men are right: women need to get off toxic platforms like TikTok and Instagram. But at the same time, women aren’t the ones building these platforms. Men need to stop creating toxic digital spaces and giving them economic life to begin with.

• Categorizing women into a narrow definition of how women should behave will backfire. The response to the outcry of women on X — many of whom have experienced postpartum depression but simultaneously don’t doubt the guilt of Lindsay Clancy — has revealed a conservative purity test that is no different that Woke 1.0’s test of Left. They believe all women are feminists. Immediate calls to repeal the 19th Amendment in response to Lindsay Clancy’s supporters implies categorical punishment for the actions of a radical minority of truly feminist women. This reaction will likely alienate conservative and centrist women, sowing new seeds of distrust that previously did not exist.

There’s no doubt that feminism has been a destructive force to the American way of life five decades now. But we're past the point of assigning blame.

We’ve identified the problem, now it’s time to fix it. More on that in this article:

After you listen to this audio note, share your thoughts in the comments below. Family formation and community building are essential for ensuring the perpetuity of our civilization. How do we make that happen if men and women fundamentally hate each other?

Leave a comment

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

Americans are bogged down by a rising cost of living. But is the affordability crisis really what we think it is?

As discussed in Part I of this series, politicians are leveraging the affordability crisis to gain votes. Socialists have strategically leveraged the rising cost of living to infiltrate progressive politics. Meanwhile, conservatives are trying to distance themselves from the current administration, arguing that the powers that be aren’t doing enough to address the rising cost of living for young people.

The discussion about daily living expenses recently boiled over on X where conservatives debated the cost of a burrito. Or rather, what they thought a burrito should cost. Dubbed “Burritogate,” the discussion pitted bootstrapping conservatives against Gen Z economic victims.

Both sides of the conversation missed the mark. A burrito only costs $20 because consumers continue to pay for it.

Labor, ingredients, and location all factor into the final price tag customers are being asked to pay and those variables are different depending on where you go. A burrito in Los Angeles – a state with a government-mandated minimum wage of $20 an hour – ordered with guac and a double portion of beef – a protein source impacted by supply constraints – is going to cost more than the same burrito in Wichita, Kansas.

And forget about the obvious elephant in the room (pun intended): there’s always the option of schlepping over to Whole Foods and purchasing a frozen burrito that you can reheat in the microwave. (There’s typically a microwave located for customer use in the dining area.)

Consumers have options. The cost of living crisis is indicative of a failure to exercise those options.

The back and forth between prominent conservatives on X reveals conservatives and socialists aren’t as different as they claim to be. Both are two sides of the same coin. Each desires government intervention to solve individual economic problems – in this case, the cost of living – but where they disagree is what exactly government intervention should look like.

While there are a number of factors impacting prices right now including inflation, corporate greed, industry consolidation, geopolitical conflicts, and tariffs, those aren’t the only things driving prices up. At the end of the day, consumers have a choice as to whether or not they’re actually going to pay the price that’s being demanded of them.

By appealing to political leaders to solve what should otherwise be individual financial decisions, both liberals and conservatives are seeking the consent of the electorate to increase the scope of government. The difference between the two is no longer an issue of big or small government, but rather who is at the helm of a growing regulatory state.

That’s what this part of the affordability crisis series will dive into. It will argue much like Part I that the affordability crisis we’re in is a crisis of consumer choice. The perception of an affordability crisis is merely being used to capture votes to redefine the nature of the government itself. It doesn’t seem to matter whether the winner is conservatives or socialist, both think everyday Americans are incapable of taking action on their own, necessitating the state to intercede on their behalf.

While Matt Walsh isn’t wrong that conservatives are missing the plot, by arguing what the price of burritos should or shouldn’t be, conservatives are revealing that they are no different than their socialist foes. Both want a government made in their own political image, including using the government as a tool for economic intervention when necessary. The result is a continued transition to a managed economy, an economy that necessitates the surrender of individual liberties to keep the machine running.

Convenience meals are bleeding Americans dry. Going out to eat has not only become unaffordable but unjustifiable for many American households. Corporate greed is partially to blame. Even so, consumers have a choice. They can stop spending money on dining out altogether.

Burritogate is a clear effort by conservatives to appeal to Gen Z voters. There was a time not-to-long ago when you could go to Chipotle for $10 – an affordable lunch for a busy working professional. But that time is long gone. Consumers no longer live in an economy that can support the conveniences they demand.

While there are some obvious explanations as to why Chipotle burritos have gotten so expensive – Biden-era inflation, state-mandated minimum wage laws, rising energy and real estate costs, and historically low cattle herds to name a few – there are some other reasons that conservatives aren’t talking about.

The most obvious explanation is what has colloquially been dubbed corporate greed by socialists on the Left. According to your average Leftist, CEOs are raising prices to line their own pockets. While that may be true for some corporate executives, the reality is far more complicated than that.

Corporations have to sustain profit margins for their investors. If they don’t, they’ll pull their money and move it elsewhere. Corporations also have to stay competitive for investors while operating in a new post-COVID economy where supply chain issues, rising energy costs, high labor rates, and new technological investments are squeezing those very profit margins dry.

Some chief executives are working overtime to find whatever cost savings they can without raising prices for their customers. Others, however, have said to hell with the plebes, squeeze as much profit from them as possible until the economic house of cards comes crashing down.

I’d put Chipotle into this second category. Since 2021, Chipotle has increased their prices at least six times. Part of this has been to combat inflationary pressure and rising ingredient costs. But part of it is also a reflection of Chipotle’s position in the market – they’re raising prices because they know they can get away with it.

Approximately 60% of the chain’s core customers earn more than $100,000 per year. Compared to customers at the lower end of the socioeconomic totem pole, Chipotle customers value fast food made from clean, quality ingredients – a premium they’re willing to pay for. While a price hike would send a McDonald’s customer to a competitor, the same isn’t true for Chipotle. Corporate execs figure most of their core customers will continue eating the cost so long as they continue to position themselves as a healthier alternative to the Golden Arches.

And that’s exactly what’s been happening. Since 2015, the cost of a Chipotle burrito has gone up by about 39%.1 Despite the obvious change over time, consumers have continued to pay whatever price is asked of them at the register.

Meanwhile, they’re taking to social media to complain about it. Financial conversations that were taboo a decade ago are now being discussed in the open. Not to mention because the demographic that should be able to afford Chipotle burritos is rapidly falling out of middle class status, affordability has become a core political issue for the majority of Americans – not just lower income households.

By turning the cost of a burrito into a political problem, voters are signaling that they want the government to lower costs for them – an act that necessitates government intervention. To do so, socialists are pinning the blame on greedy corporations and wealthy elites who seem to be hoarding wealth. Besides a grasp of basic economics, young people don’t understand that they are the enablers of corporate greed – not the victims of it. Whatever solution socialist politicians offer today will probably be to their detriment tomorrow.

While food is a necessity for human survival, not all types of food are a human right. Sit down restaurants and fast food drive-thrus are conveniences that are products of consumer choice. If you think a bucket of chicken from KFC or a salad at Sweetgreen is too much, no one is forcing you to buy either. You can place an order, be appalled by the price, and politely apologize to the cashier and walk away before surrendering a single penny.

The most obvious answer to solving the affordability crisis as it pertains to food is to just cook from home. At least until we come out on the other side of the economic transition we’re currently in.

In the interim, unless you have a hard-to-get reservation at a Michelin-star restaurant, most of what you’re eating at restaurants is just reheated Sysco slop. Almost anything you order off a menu can be purchased from a grocery store and made at home for a fraction of the price (more on that in a bit).

What’s ironic is that many videos going viral on social media complaining about the affordability crisis are of consumers complaining about purchases they’ve made after the fact. This creates the perception of an affordability crisis that might not be what it seems.

Take this woman as an example. She spent $64 on reheated soup and macaroni and cheese from Panera without a second thought only to use her purchase as a prop to generate clicks and views on TikTok. In this example, the problem is as much about affordability as it is buyer’s remorse (and the lack of basic skills needed to feed one’s family).2

But I digress.

Much of what consumers are paying for when they go out to eat isn’t a meal or even an experience these days, it’s the substitution of their labor for someone else’s. Corporate greed is as much the result of deliberately fleecing customers as it is equity policies mandated by cultural elites to provide “living wages” for low-skill workers in their communities.

Whether it’s the result of inflation or government mandate, the rising cost of human labor is a major problem.3 As noted above, if the cost of doing business is passed down to shareholders by way of lower returns, investors will pull their capital out of the market. Without investor capital, revenue will dry up resulting in closed up shops and lost jobs. That’s why when costs go up, those costs are inevitably passed down to you and me – the consumer – in the form of higher prices.4

Customers have every right to be annoyed that costs are going up, especially for something that’s supposed to be cheap like fast food. After enough price hikes, the value proposition disappears. This is clearly playing out right now, not just on X but in the market. Many people are wondering why on earth they’d go to a fast food chain like McDonald’s when they can sit down and have a decent meal at Chili’s for the same price.

But at the end of the day, dining out isn’t a human right – it’s a choice. You can either pay the price of what’s being asked of you, regardless of whether that price has been determined by corporate greed or rising costs. Or, you can opt out. Consumers of all ages and voting persuasions can simply choose to stop patronizing establishments that they believe have become unjustifiably expensive.

In simple terms: if you think a burrito shouldn’t cost more than $20 then stop paying $20 for it. Find an alternative and make the market adapt to you.

That being said, based on the discourse on social media, it seems consumers aren’t willing to make this choice. They want to continue living in the economy of 2016 without realizing a decade has passed, a global pandemic has happened, and a new Middle Eastern war has started since them. The economy, not to mention the political environment, has fundamentally changed even if consumer expectations have not.

Solving the affordability crisis isn’t inherently a political problem. At least, not in the short term. It’s as simple as recalibrating your lifestyle to the economy as it currently exists, not as it existed a decade ago.

By making it a problem for politicians to solve, however, consumers are inviting more government intervention into the Main Street economy. We have decades of lived experience and an entire Rust Belt to prove nothing politicians promise to do, economically speaking, will help everyday people like me and you.

​

The problem of affordability is fundamentally a problem of choice. Consumers have to be willing to adapt. They have to be willing to say ‘no’ when things get too expensive. No amount of government intervention will solve that.

Until consumers are willing to say no, the sky’s the limit. Prices will continue to rise and with it, the overall cost of living in the United States.

But it’s not just dining out that’s the problem. Consumers also point to a rise in the cost of groceries. While prices at the grocery store have definitely increased, they have not increased as much as social media wants you to believe.

Social media is filled with anecdotes of average Americans struggling to make ends meet. Because so few Americans have the ability to produce essential goods for themselves – namely food – we are all dependent on systems rife with exploitation and abuse. Corporate profiteering isn’t just felt at restaurants, it’s also something shoppers are noticing during their weekly trip to the grocery store.

When you combine the rising cost of food with unaffordable housing and rising price of social interactions, the narrative of an affordability crisis takes shape. If everything appears to be expensive all at once, it makes it seem like the problem is systemic rather than a problem of individual choice. It makes an appeal for government intervention a logical solution when millions of Americans are struggling with the rising cost of living all at the same time.

Diving deeper into personal shopping habits at the grocery store, however, it becomes clear corporations and wealthy elites aren’t solely to blame. Many of the outraged shoppers on social media – women with manicured nails, a face full of makeup, and freshly styled hair – aren’t going to the grocery store to shop for food. They’re going to the grocery store to shop for food products.

Unlike real food – which you can grow or hunt yourself – manufactured food products like chips and soda can’t be easily made by the average person. Making these items relies on a whole host of industrial processes.

Packaged ultra-processed, convenience foods are produced in manufacturing facilities with human labor and transported to warehouses across the country before being placed on a shelf for you to purchase. All of these touch points come at a cost. Combined with the imperative to generate a return for shareholders as noted above, it’s no accident that grocery bills are going up too.

These items can no longer be produced at the price point most people are accustomed to. In some cases the shelf price has gone up leaving consumers with sticker shock. In other cases, the volume being purchased has gone down leaving consumers angry with shrinkflation. Either way, shoppers are paying more for less, leading many to conclude that corporations are gouging them and therefore, food itself is becoming unaffordable.

I’m going to be honest, I bought into this narrative myself. It felt like every time I went to the grocery store, my bill kept going higher and higher. That’s why I decided to fact check what I was seeing on social media and what I was reading in the news.

Last year, I created an itemized spreadsheet of everything I bought at the grocery store. While some things certainly did go up in price – cheddar cheese and seltzer water are two notable examples – almost everything else has stayed relatively constant. Produce like bananas, cabbage, onions, and potatoes are the same as they were a year ago. Dairy and eggs have fluctuated based on conditions in the market. But surprisingly, things like bacon have actually gone down in price.

If food has truly become unaffordable – to the point where the peasants are at the gates ready to stage a revolt – you wouldn’t expect to see stable prices, much less food items that have gone down in price. The problem, it seems, isn’t the unaffordability of food itself. At least, not what I would consider real food.

The problem is the rising cost of ultraprocessed convenience foods. These foods are going up faster than shoppers’ wages. Instead of swapping to lower cost alternatives, consumers continue to purchase these products, pushing their cost of living beyond their means. Just like going out to eat, grocery shopping comes down to consumer choice.

Take cotton candy grapes as an example. On social media, shoppers happily purchase these modified grapes, even when they cost $12 a pound. If you don’t pay close attention, one could come to the conclusion that all grapes are $12 per pound.

I went to my local grocery store to see how expensive grapes are. While they weren’t $12 per pound as the woman in the above TikTok purportedly paid, cotton candy grapes were still twice as expensive as regular seedless grapes.

This is just one example with one food product but you can do this with just about anything in the grocery store: organic vs. conventional, artisanal vs. regular, name-brand vs. generic. There is almost always going to be a low-cost alternative for everything you buy at the grocery store. Even still, many ultra-processed foods can be prepared at home for a fraction of the price. All that’s required is a bit of time and a willingness to get your hands dirty.

Prices have certainly gone up and American families are struggling, especially when you take stagnant wages and inflation into account. But when you also factor in convenience and choice, it casts some doubt on the depth of the affordability crisis that is being used by political factions to expand the scope of government.

Yes, prices have gone up at the grocery store, but grocery staples are hardly to blame. Consumers continue to be driven by preference. Eating sugar-laden processed food products may be preferable, but at the end of the day, that too is still a choice. These choices come with a cost. For some, that cost has become too great to bear.

Solving rising food prices with more government intervention – like city-run grocery stores, subsidies, and price controls – however, is not the answer.

The next part of this series will bring the first two parts together by looking at why all of this matters. Prices are going up, inflation is largely to blame, but when consumers stop taking accountability for their choices, it invites “experts” to intercede on their behalf. In the long run, greater government intervention will invariably result in surrendering the right to choose altogether.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

1

This varies based on what you order and where you order. This figure compares the price of a chicken burrito bowl with guac as documented by receipt images posted to Chipotlife.com and the same burrito today ordered from the same Chipotle in Santa Monica. Using 2026 dollars, a 2015 burrito bowl should cost $13.29 demonstrating that for some items on Chipotle’s menu, prices have gone up in line with inflation. Nonetheless, even when adjusted for inflation, the price is 41% higher in 2026 than it was a decade ago. Put another way: $10 buys 41% less in 2026 than it did in 2015.

Welcome to this month’s Q&A. Share your ideas, thoughts, and questions in the comments below.

Leave a comment

Read more

Women are going to be one of the biggest losers in the AI Revolution. The Washington Post reported back in March that of the workforce at risk for some sort of displacement by AI, 86% of them are women.

This makes a lot of sense. Many college-educated women work bullshit “pink collar” jobs that are largely bureaucratic in nature. They entail pointless meetings and moving pixels around on a screen, providing little to no contribution to society. As a result, when the opportunity comes to eliminate or automate these positions, employers will be happy to do so.

Rather than reading the room and pivoting to new careers before this happens, women are blaming everyone but themselves for the decline in work opportunities: the patriarchy, greedy corporations, and Elon Musk.

Without office jobs to work (or husbands to support them), millions of women are about to be sidelined by the new economy that emerges. Strong and independent – but without the means to take care of themselves – they’re going to turn to the government for support.

It’s not an accident that socialists are winning primaries across the country as they ascend to positions of power. Their supporters are white liberal women, those who are most at risk from AI-related job displacement. Keep this in mind as the midterm elections unfold later this year.

Across the board, the prognosis for workers isn’t good. Millions will need to pivot into new careers to stay afloat. Men will likely retrain for blue collar work and to some degree, this is already happening. Women, on the other hand, will have to retrain for comparable human-centered work. Much of the jobs that will be available to women in the near future will be jobs in the caring economy.

Despite demand for caretaking work, women who lose their pink collar jobs won’t exactly be eager to fill those roles. For one, they lack the skills needed to be successful taking care of others. Pink collar workers know how to plan meeting agendas and write emails, but they have next to no experience changing dirty diapers.

More importantly, the feminist worldview has conditioned women to believe caretaking work is a step backward. It obscures the reality that pink collar office jobs are highly susceptible to automation. This creates a blindspot. Even if pink collar women could retrain they likely won’t want to so long as a stigma is attached to it.

But for those who are willing to look beyond the social value of work, this creates a massive opportunity. Because there aren’t enough caretakers today, those who pivot now will be able to build lucrative careers tomorrow, especially as caretaking takes on greater importance in the years to come.

Workers need to prepare for new roles in the AI-powered economy that’s emerging, the question is how exactly should workers navigate that transition? That’s what this essay will dive into.

While many pink collar workers will resist the need to retrain for new work, the reality is physical labor jobs are the only jobs where humans still have a competitive advantage. This is something knowledge workers can’t ignore anymore. Aside from the cultural stigma associated with caretaking work, unemployed (or soon-to-be unemployed) pink collar workers don’t even know what their options are. How do you retrain for work you don’t even know exists? This essay will offer a framework for how pink collar workers can navigate into caretaking work as high-value specialists.

The caretaking economy is more than just daycare and babysitting. There are dozens of specialized roles that pay as much – if not more – than pink collar desk jobs.

One of the primary limiting beliefs handicapping pink collar workers is the idea that caretaking work doesn’t pay. And to be fair, many low-skill entry-level roles don’t. In some parts of the country, fast food workers now earn more per hour than daycare providers. But just because some caretaking jobs don’t pay well doesn’t mean a pink collar worker who transitions to caretaking work will always have to work a low-paying job for the duration of her career.

I recently became a mom. In doing so, I have become exposed to an entire world of caretaking that I, a former pink collar worker myself, never knew existed. Sure, I’ve always been aware of babysitting and nannying. For a while, I babysat on the weekends when I lived in Washington, DC. But aside from that, I never knew more high-paying specialized roles existed.

I suspect many pink collar workers are in the same boat. Unless you are exposed to the caretaking economy, you don’t know higher paying jobs exist. These are just a few roles that have recently entered my awareness. They are skilled caretaking jobs that pay just as much – if not more – than many pink collar office jobs.

Doulas are medically-adjacent professionals that provide extra support during and after childbirth. Their services aren’t typically covered by insurance, but even with the extra cost, many women swear by them. Birth doulas make around $72,000 per year, more depending on where a doula lives and how many babies she helps deliver in a year.

Night nurses help new parents with infant care, specifically during the dreaded night shift (IYKYK). Also referred to as night nannies, newborn care specialists, and postpartum doulas, they typically come at night to help with overnight feedings, provide guidance on how to establish sleep routines, and take care of small tasks like sterilizing bottles or swapping out the laundry.

Depending on experience, credentials, the type of care being provided, and location, night nurses can expect to earn $35-40 per hour for a 12-hour shift, with nurses in major cities making anywhere from $600-800 per day.

Similar to a birth doulas, there are certifications you can get – like becoming a Certified Sleep Consultant – but they aren’t necessary to get started.

Lactation consultants. Certified by the International Board Certified Lactation Consultants, lactation consultants support new moms with breastfeed. For those of you who don’t have kids, it’s much harder than it looks.

Lactation consultants are employed directly by hospitals and work at outpatient centers. There are different ways to become a lactation consultant, but most pathways require some combination of coursework, clinical hours, and passing a board exam. On average they make around $94,480 per year, comparable to a Registered Nurse.

House managers are employed by high-income families juggling kids and work, effectively operating as the family’s chief of staff. They play an important administrative and logistical support role: scheduling repairs, placing grocery orders, running kids to soccer practice, sourcing a dogsitter, making travel arrangements, etc. Depending on the household, some house managers might also take on additional childcare and light cleaning tasks.

Salaries for house managers vary depending on the scope of work, experience, and where a family is located. Managers can expect to make between $30-40 per hour. High income families in high cost of living cities will pay top dollar for the right person.

Caretaking roles like these pay a wage that is comparable to what many pink collar workers make at their desk jobs. According to the latest data from the Bureau of Labor Statistics, the median weekly earnings for a woman with a bachelor’s degree or higher is $1,555, equivalent to $38.88 per hour or $80,860 per year.

There is a shortage of caretakers available to fill many of these roles. The right person with the right skills could earn just as much – if not more – than what they might currently earn in a pink collar role.

Despite an abundance of work opportunities in the caretaking economy, there is a shortage of individuals willing to fill those positions. Families are paying top dollar to meet their needs.

One of the challenges facing pink collar workers who are looking to pivot into the caretaking economy is that the job application process is completely different. You won’t find good paying caretaking jobs posted on LinkedIn or Indeed (unless you’re trying to get hired by some private equity owned daycare center).

Many of these jobs are staffed privately. Families post their needs in a private forum like a Facebook Group or they post vacancies on caretaking sites like Care.com.

Here in Austin, parents are struggling to find qualified caretakers to meet their needs. Take this post from a local Facebook Group as an example:

One of the reasons why there aren’t enough caretakers to meet the needs of families looking for help is that workers believe caretaking work is low paying. And again, it is for many of the low-paying entry-level roles many pink collar workers interact with. The typical childcare worker only makes around $15 an hour after all.

But as noted above, this isn’t the case for all caretaking roles. Many families are quite generous. When it comes to taking care of their children, the families that have the means to do so pay $30+ an hour with benefits, comparable to what pink collar workers make in a corporate setting.

The worker shortage isn’t a pay problem, it’s a supply problem. Caretakers aren’t specializing. There are too many individuals chasing many generic roles – like weekend babysitting – where their services aren’t needed. Not only does this push wages down – validating the belief that caretaking is a low paying profession – it creates unmet demand where workers are needed – like night nannying.

Pink collar workers who transition into a specialized caretaking role have an opportunity to capitalize on this asymmetry. While these workers will have to start somewhere to get their foot in the door (and that somewhere may not pay what they’re accustomed to), once they have some skills and experience, they can specialize into a higher paying role.

After a couple of years, gaining experience and establishing a reputation for themselves, caretakers will not only be able to earn more, they’ll be able to cultivate new careers and start businesses that far surpass the earnings potential of climbing a traditional corporate career ladder.

The caretaking economy isn’t limited to childcare. Demand for other types of care activities is growing. Workers with experience can leverage it into higher value self-employment or business opportunities. This provides a higher long-term ROI than most corporate jobs, especially in an AI-dominated economy – no degree required.

Up until now I’ve discussed the caretaking economy from the perspective of childcare. When you hear the term ‘caretaking economy’ this is likely the type of care that comes to mind.

While caring for children is one of the more common types of employment in the caretaking economy, it isn’t the only avenue for displaced pink collar workers. There are other ways to tap into the caretaking economy and benefit from its growth.

Demand for adult care and eldercare services are rising as Baby Boomers age. According to the Bureau of Labor Statistics, demand for home health and personal care aides is projected to grow by 17% over the next decade. Back in February, NBC News reported:

The share of Americans ages 65 and older is projected to reach over 20% by 2035, meaning roughly 1 in 5 Americans may require some form of care. Baby boomers are moving deeper into their 70s and 80s, and more older adults say they want to age at home rather than in institutional settings.

This is verified by jobs data. Over the past few years, healthcare and social assistance has been the only sector of the economy actually driving job growth. If you’re unemployed and looking for work, this is where the strongest demand signal for labor is coming from.

While there is a clear need for workers in caretaking roles, many entry-level roles don’t pay well. Childcare workers can expect to make around $15 an hour while home health aides make only $17. For pink collar workers living in high cost of living cities, accustomed to specific lifestyles, and responsible for mountains of student loan debt, this type of work isn’t viable. The good news is that while, yes, these types of jobs are a necessary starting point, they aren’t terminal positions. After you get some experience under your belt and complete any necessary certification or training programs, savvy caretakers can launch a caretaking career into something far more lucrative than their current 9to5 desk job.

One way to do this is by building an agency to source talent to fill demand. Take newborn care as an example. After you put the time in and establish yourself as a bona fide night nurse you can take your experience to build a company that recruits and staffs other night nurses for sleep-deprived parents. Welcome Home Baby is a business that provides direct night nursing services while The Nanny Edit is a placement agency that connects parents and caretaking professionals.

Another way to leverage caretaking experience is to become a trainer for other caretakers entering the field. Pink collar workers displaced by changes in the labor market may especially struggle to find employment in the growing caretaking economy because they don’t have the right skills or experience. Many of these workers have college degrees and are accustomed to working in an office from 9am to 5pm. They are used to finding work on LinkedIn or through professional networks, unaware that many caretaking jobs are privately employed and informally sourced through Facebook Groups.

Someone with experience in the caretaking economy can leverage their experience to guide and mentor new workers. That could look like one-on-one coaching and in-person classes. Or, it could look like an online training platform.

Norland College is one of the most well-known examples of childcare training programs. The U.S. Nanny Institute is another that provides accessible training here in the United States. As a new generation of workers pivots into caretaking, there will be increased demand for training them, presenting training opportunities for those who want to take their existing caretaking career to the next level.

Finally, tech savvy caretakers will be able to leverage new technologies to provide high value services. Take house managers as an example. An AI-enabled house manager can help families integrate artificial intelligence into their systems, optimizing how their home is run.

In much the same way that corporate employers are paying top dollar for workers with AI skills to come in and help optimize business operations, private families with household staffs to manage will likely go the same route. It’s only a matter of time before AI integration in households becomes ubiquitous.

While it’s true that new caretakers will have to start somewhere and that somewhere is likely a $15 per hour job, that doesn’t mean workers will be stuck at $15 per hour forever.

With opportunities to expand to other areas of the caretaking economy or opportunities to build businesses, these jobs provide the potential for a far greater return on investment than merely continuing to climb a prescribed corporate ladder. And ironically, these human-centered jobs will provide far greater financial security than the pink collar jobs promised by feminists years ago.

Final takeaway.

Corporate W-2 jobs are no longer the guarantee of stable employment they once were. Workers who can’t compete with digital automation will have to look for human-centered roles if they want to stay employed in the years to come.

For many, this will mean creating new employment opportunities for themselves. Whether that’s freelancing or running a small business, workers are going to be increasingly responsible for earning their own keep.

How workers approach work will likely diverge along gendered lines. Men will pursue blue collar jobs where automation and machines aren’t as competitive – at least not yet. Meanwhile, women will find that their competitive advantage is in caretaking where human nurturing is still needed.

Women can certainly pivot to blue collar jobs and work alongside men but it’s unlikely that’ll happen. Despite equal access to training programs and nonexistent barriers to employment, women don’t self-select for working on oil rigs, road crews, or building new homes. It’s not likely that a loss of pink collar jobs will suddenly change this preference.

Instead, the most viable option for displaced pink collar workers will be in the realm of caretaking. The most successful caretakers will be the ones with the skills, experience, and reputation demanded by the market. All of these factors are predicated on time spent in the industry.

To get started, pink collar workers need experience. “Going back to school” isn’t an option anymore. Pink collar workers need to get their hands dirty and actually work. For childcare, that means looking after children. Pink collar workers can gain experience by volunteering in a church nursery, joining a gig platform like UrbanSitter, or looking for a $15 an hour job at a local childcare center. There they can learn the skills they need to be successful on the job while developing a positive reputation that they can leverage later on.

After a couple of years, a pink collar worker who begins the transition now will have enough experience to specialize and build something for themselves. That can be as a direct service provider, creating an agency, developing a training curriculum, or creating some sort of AI-enabled platform that can be scaled and exited.

The primary friction point right now is mindset. Pink collar women who believe that caretaking is beneath them will be unable to see the demand signals and pivot in time. These individuals will look to the state to support them when they lose jobs and are unable to find new ones.

The future of work lies in human-centered demand signals. For women, the easiest and most obvious path forward is in caretaking. Pink collar workers who pivot now and gain caretaking experience while still employed will be better positioned to absorb shocks to the labor market in the years to come.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

There are myriad reasons to explain why: the Department of Education, No Child Left Behind, gentle parenting, bloated school bureaucracies, administrators, and iPads.

All of these things are to blame, yet blaming devices, institutions, and groups of people doesn’t solve the education crisis we’re in.

Kids are progressing through the school system without the ability to read in context or do simple math. The first wave of illiterate kids has already graduated, more will be entering adulthood in the coming years.

This will have a massive impact on politics, society, and the economy in the coming years. Quantitatively we’re not reproducing enough kids to carry on the legacy of America into the future; qualitatively the ones we are producing don’t have the skills and competencies they need to carry Western Civilization forward.

Even if you don’t have kids and are far removed from the public school system, the state of America’s youth today is a forecast of what America will look like in the coming decades.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

Prices for ordinary Main Street Americans have gone through the roof ever since the pandemic shut down the global economy back in 2020. This is something I’ve written about extensively and it’s something you’ve, no doubt, experienced for yourself.

When you look at the numbers, it’s abundantly clear prices are rising far faster than economists anticipated. Consider the following: in 2019, the average rent for an apartment was $1,149 per month (equivalent to $1,506 in 2026 dollars). Today, the average rent is $1,750.

A few years ago, a new car would run you $531 per month (equivalent to $709 in 2026 dollars). Today, that monthly car payment is closer to $770.

And before COVID, a family could spend less than $5,000 per year at the grocery store (equivalent to $6,086 in 2026 dollars). Current estimates put the annual American family’s grocery bill at about $6,224.

While prices for essential living expenses like housing, food, and transportation continue to rise, wages have stayed stubbornly stagnant. In 2019, the median household income was $68,703 (equivalent to $90,058 in 2026 dollars). As of 2024, the median household income is just $83,730, well below where it ought to be. Based on simple arithmetic, it shouldn’t come as a surprise, then, that everyday Americans are struggling to keep up.

A growing number of what you might consider to be successful middle class Americans are taking to social media to lament how hard it is to make ends meet. When you see raw, unfiltered stories going viral and combine those stories with hard economic data, it appears that we’re living in an unprecedented economic crisis. Or at least, a crisis not seen in several decades.

But are things really as bad as they seem?

A new vanguard of political leadership has emerged, positioning themselves as the saviors of unaffordability. They’re channeling the frustration of what’s left of the middle class into disdain for millionaires, billionaires, and large corporations. Look no further than the reaction to the recent SpaceX IPO and Elon Musk’s ascension to trillionaire status as an illustration of this.

The vast majority of Americans are skeptical of the establishment. They know the status quo is no longer tenable. A political vacuum has emerged, one in which this new vanguard of political leadership is eager to fill.

Their messaging appeals most to young people. Those who have yet to own homes and start families of their own believe the cost of living crisis is driven by greedy corporations and wealthy individuals who are hoarding financial resources for themselves. They think that the only way to solve the problem is for the government to step in and regulate wealth. Young voters assume once they do, their political representatives will share the spoils with them.

When you look at the empirical data across all of Main Street – not just what’s being shared in the media – it reveals things might not be as bad as they seem. Here in Austin, Texas, rent has actually gone down not up. Grocery prices for pantry staples have remained largely stable with a handful of items actually going down in price. And while, yes, car prices probably aren’t coming back down anytime soon, it’s hard to pin that on greedy automakers. For every article discussing how unaffordable car payments have become, there’s a dozen TikTok videos of consumers gloating about financing fully loaded cars they very clearly cannot afford.

Politicians would be remiss to let a good crisis go to waste – even one they manufacture themselves. A new wave of anti-capitalist politicians are sweeping elections across the country and they are crafting a narrative around a crisis of affordability that may be present in some cities but isn’t present everywhere. Their goal isn’t to make things more affordable, it’s to shepherd Main Street into a new economy aligned with their socialist values.

What this reveals is not necessarily an economic crisis but a political opportunity to concentrate power within the hands of a new ruling elite.

By demoralizing the population – especially Millennials and Gen Zers – politicians have successfully demonized wealth and the system of free enterprise that generates it. They’re channeling collective rage against wealth – rooted in perceived widespread unaffordability and inequality – to lay the groundwork for dismantling the system we currently live in. The goal isn’t to solve affordability but to replace capitalism with a state-run economy managed by an expert class of bureaucratic elites.

America is flirting dangerously close with communism. But instead of seizing the means of production by force – as the Bolsheviks did more than a century ago – this new wave of communist politicians are ascending to power through the consent of the governed.

That’s what this essay will dive into. It will argue that we are living in a choice crisis rather than an affordability crisis. By failing to acknowledge the crisis for what it is, we’re creating a permissible environment for socialist politicians to ascend to power.

This essay focuses specifically on young people – Millennial and Gen Z voters disaffected by the economic reality they live in. This is the voting block that’s enabling socialists to rise to power and the voting block that’s going to have to live with the consequences of their votes later on, whether they’re prepared to or not. Rather than seeking out solutions to make things more affordable for themselves, they are looking for politicians to solve the problem of affordability for them.

Part I is all about how socialist politicians are luring young voters in with the promise of affordable housing. What could possibly go wrong?

Housing is a key driver of the affordability crisis. While housing is largely unaffordable in many large metropolitan areas it isn’t unaffordable everywhere. Millennials and Gen Zers voluntarily – emphasis on the word voluntarily – moved to cities like New York and Los Angeles in search of work opportunities. The jobs they wanted don’t exist. Instead, they could solve the affordability crisis for themselves by simply finding somewhere else to live.

In 2014, I moved to Washington, DC. With a degree in international relations and a burning desire to work for a three-letter intel agency, I believed I had to live in Washington, DC if I wanted to build a successful career. But after years of failing to land my dream job and struggling to keep up with the cost of living, I came to the conclusion that living in DC was a choice. Just as I had chosen to move there I could also choose to leave. That’s exactly what I did and I haven’t looked back.

Young people in the country’s largest cities have yet to come to this realization for themselves. Many are living in large, expensive cities by choice, under the belief that they have to be there for their careers. What they don’t know is that, for one reason or another, the careers they thought that college had prepared them for largely aren’t going to come to fruition. While some may succeed in their endeavors, the vast majority won’t.

They’re wasting their time, impoverishing themselves in the process.

Continuing to live in an expensive city working an underwhelming job is a cross many young people don’t have to bear. And yet, they continue to do so voluntarily. No one is holding a gun to their head, forcing them to live with roommates or pay exorbitant rents for tiny apartments. They simply haven’t put much thought to the choices they’ve made or the different choices they could make moving forward.

The lack of critical thinking that would otherwise enable young people to make different choices is driving a socialist surge across America. Driven by disgruntled Millennials and disaffected Gen Zers in major metropolitan areas like New York and Seattle, young people are absolving themselves of the responsibility to take ownership over the choices they’ve made, whether they consciously made them or not. Instead, they’re looking to a new political paradigm to solve their problems for them.

Last month, New York held its primary elections where socialist candidates swept the board. On the eve of America’s 250th birthday, young voters signaled they’d prefer to send literal communists to Congress over maintaining the status quo. Joe McCarthy must surely be rolling in his grave.

While this is happening, the Democratic establishment – the political faction that has largely appealed to young, liberal voters – has effectively been read the riot act. Existing Democrats who want to stay in power will have to follow the lead of this new voting bloc. They’ll need to accept far-left socialist candidates into the fold, moving the entire party further to the left as a result.

While trans rights and America’s relationship with Israel are certainly pushing some young people to the left, historically these types of issues don’t really move the needle on elections. The main driver has, and always will be, the economy. Just ask Bill Clinton.

The shift towards socialism is an appeal to make life more affordable for young people in order to capture their votes. New York’s Millennial Mayor Zohran Mamdani clinched his victory last fall by promising to freeze rent and open up city-run grocery stores. Six months later, he made good on his promise by freezing the rents of millions of New Yorkers.

But this isn’t just happening in New York. The affordability crisis is a launch point across the country, helping the Democratic Socialists of America – the national faction spearheading America’s communist surge – to get into power. According to a co-chair of New York’s DSA chapter, affordability has been a key pillar of their political platform and is a driver of their success:

It was through trying to implement the affordability agenda that we made the case for electing more democratic socialists, because we need to tax the rich to fund the program, and the governor and the current power structure fought us all the way in this past session,” said Gordillo. “Our argument and our case that we’re making to New York is we need to replace the existing politicians if we actually want to create a city and a state that’s affordable for everybody. I think the mayor has prioritized pothole politics and sewer socialism as a way of showing that you can elect a democratic socialist to the highest level of office in the city, and that’s actually going to result in a government that is more responsive to working people. (The Guardian)

Affordable housing is at the heart and center of the affordability crisis that Democratic Socialists are tapping into. It’s such a big problem that a number of experts have argued it’s the culprit behind the collapse in global fertility rates. But for politicians, this isn’t a crisis they should let go to waste. It’s a key campaign promise that is winning elections.

While the allure of affordable housing may appeal to young voters today, it will have disastrous consequences tomorrow.

A quick study of economics shows that housing is a problem of supply and demand. While politicians can promise to make housing more affordable in the short-term, those promises don’t solve the fundamental problem plaguing young people: housing demand outpaces supply in many of the places they live.

To solve the problem of housing affordability, the market has to restore prices to its natural equilibrium. There are only two ways to do this: either municipalities have to make it easier to build more housing – increasing the supply – or young people have to move – lowering demand. Absent a free market solution, the only thing the Democratic Socialists can do is increase their portfolio of publicly funded tenements.1

To demonstrate how the free market can solve the housing crisis, look no further than Austin, Texas. Austin has one of the least restrictive municipal zoning codes in the country. You can pretty much build whatever you want, wherever you want. This has been a bonanza for developers, especially in areas just outside of the city center.

In recent years, the supply of housing in Austin and the surrounding suburbs has exploded. According to the Pew Charitable Trusts, changes in zoning made it easier for builders to build, leading to a spike in inventory:

But starting in 2015, Austin instituted an array of policy reforms aimed at encouraging the development of new housing, especially rentals. The city changed zoning regulations to allow construction of large apartment buildings, particularly near jobs and transit. In 2018, voters approved a $250 million bond measure to build and repair affordable housing. Permitting processes were reformed to speed development and reduce costs.

The efforts worked. From 2015 to 2024, Austin added 120,000 units to its housing stock—an increase of 30%, more than three times the overall rate of growth in the United States (9%).

The result? A massive drop in housing costs. The average rent in Austin is currently $1,410 per month, well below the national average. If you don’t mind a short commute, you can find even cheaper rent for under $1,200 per month in one of the city’s many suburbs.2

And it’s not just rent that has come down. Austin – and the state of Texas in general – makes home ownership more accessible than Democrat-run states like New York and California. While the median home price in Austin is still above the national average and (expensive by Texas standards), there’s a plethora of entry-level single family homes on the market. The key here is being willing to accept an entry-level house now and eventually trading up to a larger house in the future.

This isn’t a phenomenon unique to Austin. In many lower tier cities and small towns across the country you will find an abundance of affordable housing. Sure, there are trade-offs to living in some of these smaller cities. Young people might not find the social amenities they’re accustomed to like high end restaurants or cold pressed juice bars. But in places like Austin, comparable amenities do exist. Pockets of the city are being intentionally developed to mirror that of New York or Los Angeles. If you want a $400 gym membership and an overpriced latte, you can get it in Austin without paying the huge sum of rent you’d find in a major city.

Specific cities are definitely facing a limited supply of housing, but the problem of housing affordability isn’t fundamentally a problem of supply. There are affordable homes out there you just have to look for them. While this isn’t the answer young people want to hear, if they want to become homeowners, they have to go out and look for the houses where supply exceeds demand.

What this ultimately reveals is that the housing affordability crisis that socialist politicians promise to solve is really a problem of demand. The houses that are affordable and are available aren’t in the places young people want to live. The problem is that young people don’t want to relocate to where they’re located either.

Instead, they’d rather appeal to politicians to conjure up more housing for them. We already know this doesn’t work. For years, cities across the country tried to regulate housing to make it more affordable. It only made it harder for builders to build and landlords to rent, driving prices up. That’s why many of these cities are facing housing shortages in the first place. Freezing the rent may make housing more affordable for another year or two – long enough to win another election cycle – but across the long term, it will make property ownership undesirable while locking young people into apartments they’ll inevitably grow out of.

While not everyone has a choice in where they live, many do. Most of the young people suffering under the weight of housing costs chose to move to large, expensive cities voluntarily. If they want affordable housing, they can, likewise, choose to alleviate their own suffering by simply making the decision to move to where affordable homes are located.

Part II of this series will look at affordability as it pertains to food. While the media is presenting a narrative that food is becoming too expensive, greedy corporations aren’t necessarily to blame.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

Hello Tomorrow Today readers and welcome to the first of what will hopefully be many Q&As.

Share your ideas, thoughts, and questions in the comments below. Over the month, myself and your fellow readers will offer perspectives to help answer your questions.

The goal is to foster constructive dialogue. When presenting an argument, be sure to add sources in your comments.

Leave a comment

Read more

Two months ago, I went offline to welcome my first baby into the world. What a whirlwind it’s been since then! Dad, baby, and I are all doing great but I wanted to give you an update on where things currently stand with Tomorrow Today.

I plan to resume writing soon but with a newborn in tow, my writing cadence will be much slower than it was before. So what can you expect moving forward?

I’ll aim to publish a minimum of two new essays each month.

These will be accessible to all readers with first read available to paid subscribers (more on that below). But because time is no longer my own, I can’t guarantee publishing on a specific schedule. Some days I may have a sold hour to write, other days I may only have 20 minutes.

This will change in the future once I have a more consistent nap schedule to plan writing around, but for now, I’ve got to work with where I’m at. I appreciate your patience as I navigate this new season of life.

To my paid subscribers I want to say first and foremost, I appreciate your support. You make it possible to keep writing and let me know you value these types of conversations. Moving forward, here’s what you can expect from me:

• First Read: You’ll receive every new essay in full first with the opportunity to join the conversation in the comments section.

• Video/Audio Commentaries: These will be deeper dives into essays and thoughts about that may be included in future essays in a format that’s accessible to you and manageable for me.

• Monthly Q&A: A special post dedicated to engaging in open dialogue about the topics covered on Tomorrow Today. This will be a space to share feedback on things you’ve read and offer recommendations of topics to discuss in the future.

• Monthly Email: You’ll be added to my private email list where I’ll share behind-the-scene updates like this one here.

My goal is to navigate away from being just another content creator on the internet, pumping out as much content as possible to stay relevant in the algorithm feeds. Instead, I want to prioritize deeper dives into topics that really matter. I’ll share the things I’m contemplating — even if they aren’t fully fleshed out yet — and incorporate your feedback into the writing process.

As I noted in my last update, if this isn’t what you signed up for or it’s not what you’re looking for moving forward, you can cancel your subscription and ask for a refund at any time. Substack walks you through how to do that here.

I’m currently working on my next essay about the affordability crisis in the United States — or at least, the perception of an affordability crisis. This is something I’ve written about extensively, but after spending the past several weeks offline I’m beginning to wonder if the affordability crisis is actually what we think it is.

Are we actually living in an affordability crisis, or, is the political establishing ensuring that a perfectly good “crisis” isn’t going to waste?

I’ll be looking forward to sharing that with you next month.

Thank you for reading and I hope you have a marvelous day.

-Amanda

If you’re reading this note, it means I’m officially a momma. While I’m very excited for this new season of life, it’ll be coming with a lot of changes.

The biggest one: I’ll be closing up my laptop for the next several weeks as I transition into motherhood.

Besides the obvious lack of sleep that comes with taking care of a newborn, there’s a lot for me to learn. I’m logging off to make sure I’m giving myself the space I need to not only soak up all of these precious moments with my new baby, but to dive deeper into learning what it means to be a mom.

I won’t be publishing any new long form pieces for a while. And to be completely honest, I’m not exactly sure when I’ll be back.

My goal is to return to a somewhat regular publication schedule later this summer, but at this point, it’s hard to say exactly when that will be. It will all depend on making sure my family’s needs are met and we’ve found a new rhythm for life.

Even though I won’t be publishing, I do still plan to be engaged while I’m offline. I have several philosophy courses queued up on the Peterson Academy and a bunch of books on my bookshelf begging to be read.

While my family is the number one priority right now, I won’t lie: I can’t exactly turn off my brain. I’m hoping I can read during feedings or listen to lectures while I’m doing chores around the house. This will not only give me the space to be present for my family but to also think on a deeper level so that when I do come back, I can produce even more impactful essays than before.

Although I won’t be active on Substack in the coming weeks — or any social media for that matter — I still plan to share some periodic updates. It’ll just be outside of social media. If you’d like to receive those updates, you can get them here:

Get Email Updates Here

Thank you for your continued support and interest in the topics covered on Tomorrow Today. I’ll see you soon.

-Amanda

P.S.

If you’re a paid subscriber and want to pause, cancel, or refund your membership, I completely understand. Follow Substack’s guidance here to do just that.

New data on fertility rates in the United States have been released and the prognosis isn’t good. According to preliminary data from the Centers for Disease Control, there was a 1% decline in births from 2024 to 2025. For every 1,000 women of childbearing age, only 53.1 gave birth to a child.

This data confirms a contraction in the population that we’ve all been witnessing. Gen Alphas number far fewer than their Millennial parents. And the New York Times reports, childlessness is becoming far more common than it once was:

For the first time, almost half of the country’s 30-year-old women are childless. In 1976, it was just 18 percent.

The children that need to be born today to sustain the American way of life simply aren’t. In the coming decades, there won’t be enough adults alive to keep America going. One only has to look at Japan to see where we’re heading. Small villages across the countryside are vanishing as a result of population decline. According to a 2024 report by Japan’s Population Strategy Council, 40% of municipalities could completely disappear as a result.

This is why the likes of Elon Musk have been sounding the alarm on the falling birth rate in the United States. While replenishing the population to contribute to the labor market and tax base are important from an economic perspective, the issue of falling birth rates is much bigger than that.

A loss of population means the inability to maintain the infrastructure that makes up day-to-day life. Schools close due to low enrollments and businesses close up shop and leave town – taking jobs with them – due to a lack of revenue. Once the last holdouts die, the only thing that remains are empty buildings – ruins that demarcate a way of life that once existed there.

For a handful of small towns, this might not seem like a big deal. But when it’s more widespread it fundamentally changes how society is organized and governed. Cultural traditions are lost and history disappears. The civilization itself transforms such that what remains becomes something entirely new.1

Almost everyone alive today will begin seeing the consequences of falling birth rates if they haven’t already. Worker shortages in key industries like elder care have largely been attributed to worker burnout and preference. Young people simply don’t want to spend their youth changing adult diapers so they’re not taking those jobs. While that may be true to some degree, it’s also true that there just aren’t enough Gen Zers to service the needs of an aging Baby Boomer population.

Meanwhile, schools are already beginning to plan future closures. Here in Austin, the neighboring school district in Pflugerville plans to close six schools in the next four years due to low enrollment. The Austin Current reports:

Consolidation has become a familiar threat statewide as falling enrollment rates and persistent funding woes stretch public school dollars and resources thin. Across the country, school districts are confronting similar pressures as demographic and economic shifts shrink the pool of school-age children. Birth rates have been falling for more than a decade, leaving fewer kindergarteners entering classrooms each year.

Population decline isn’t a future problem. It’s already transforming American society as we speak. The consequences of the decline will only continue to manifest from here.

All of this raises some obvious questions. Why is population decline happening and how did we get here? Many are quick to point to poor employment prospects and the rising cost of living. With the cost of raising a child now topping $300,000, who can afford it?

The state of the economy may certainly be affecting fertility rates, but that isn’t a sufficient explanation. Our ancestors faced far harsher economic climates compounded by wars, plagues, and poor health outcomes. Yet, they still produced small armies of children. There’s no reason in our current technological age of abundance and prosperity, we can’t produce even more children than they did.

If the answer isn’t purely economic then it seems to be more spiritual in nature. It’s not that we can’t have more children, it’s just that we don’t want to. The social burden of having children and raising them is greater than modern women are willing to bear. As a result, they’re not just abstaining from having children, they’re eliminating them before they have a chance to come into existence in the first place.

That’s what this essay will dive into. It will look at the plummeting birth rate as a social phenomenon rather than an economic one. It will argue that the fertility rate is declining not because women are having fewer children but because fewer women want to have children in the first place.

Through the normalization of abortion and the proliferation of a culture that promotes advancing one’s career over raising the next generation, fewer children are being born. When the number of children aborted since the 1970s is accounted for, it reveals that there isn’t actually a problem with fertility. The women who want to have children are doing just fine replenishing the population. The problem is the growth in the number of women who don’t want children and are terminating a sizable portion of the population before it has a chance to be born in the first place, resulting in fewer children maturing into adulthood.

Looking at the fertility crisis from this perspective offers an important nuance to the conversation of demographic decline. If the problem is social and spiritual in nature – rather than economic – policies to make childcare more affordable or give young couples access to better housing won’t boost the birth rate. You have to establish a new narrative around family formation that makes young women want to reproduce. That is not something a policy alone can solve.

In the 1970s, global leaders warned us that there were too many people on the planet. This created the philosophical foundation for an anti-natalist cultural imperative to reduce the size of the population.

To understand why the birth rate is falling today we have to go back in time to look at why the size of a population is such a big deal in the first place.

In 1798, English economist Thomas Malthus published a book titled An Essay on the Principle of Population. In it, he argued that England’s rapidly growing population would reach a point where it would outstrip the land of its ability to produce enough food to feed all the people living there.

While many of Malthus’s contemporaries feared the poor would starve, that never happened. Technological innovations in agricultural production made it such that farmers could increase their output – something Malthus hadn’t considered.

Even though Malthus was proven wrong, there are many who still believed population growth was unsustainable. In 1968, Paul Ehrlich wrote a book warning about the looming catastrophe of famines that would plagued humanity. The Population Bomb argued that the world’s population had grown at an unsustainable rate. It was rapidly stripping the planet of its resources, attributing to the social upheaval that was prevalent during this period of time. Unless action was taken to curb population growth, Ehrlich predicted that billions of people would starve to death.

Shortly after Ehrlich published his book, the Club of Rome – an informal group of global intellectuals – met to discuss population growth. The group ran a series of computer simulations with a team from MIT. Their findings were published in a report titled The Limits to Growth and predicted societal collapse.

Over the years, researchers have updated the model with new data and more sophisticated modeling techniques. The most recent update published 2023 suggests industrial output, population growth, and food production would peak between 2020 to 2030.

While Malthus and Ehrlich were both proven wrong, that hasn’t stopped neo-Malthusians from pushing a narrative that the earth is in peril and unrestrained population growth is to blame. This line of thinking has impacted global organizations like the World Economic Forum and the Bill & Melinda Gates Foundation, quietly shaping a number of policy initiatives that have resulted in de facto population control.

Rather than finding innovative solutions to overcome the potential challenge of overshooting the earth’s productive capacity, modern leaders have taken an anti-natalist approach. They are focused on controlling human activity to keep population growth manageable by shaping narratives around reproduction. The climate crisis and modern feminism are both examples of this. Young people believe it is morally wrong to bring life into a world where the climate will produce an uninhabitable earth during their lifetime. Meanwhile anti-natalist messaging about the social burden of childrearing has dissuaded millions of women from having children in the first place.

Global leaders believe they can limit growth by using the network effects of social contagion to manage the population. Their efforts may have succeeded. The global population is rapidly in decline, drastically reducing the threat of starvation from overpopulation.

Rather than solving the problem they’ve replaced it with a different one. Instead of having too many people on the planet, we now risk having too few. As birth rates plummet there won’t be enough future citizens and workers to keep governments and economies going. Societal collapse may still be inevitable – just not in the way neo-Malthusians had imagined.

Fifty years after the Club of Rome warned that population growth would lead to societal collapse, the conversation has shifted. The new concern is that birth rates are rapidly plummeting around the world, imperiling the continuation of our modern way of life.

Birth rates are falling, not just in the United States, but all over the world. A report from the UN suggests global fertility rates are in an “unprecedented decline.”

In developed countries like the United States, Japan, South Korea, and Germany, the birth rate is well below replacement.2 The problem isn’t just that fewer children are being born on an annual basis though. It’s that the collapse of the birth rate over time is leading to smaller generational cohorts. The number of young people within a generation are too small to meet the demands of an aging population that expects to rely on them as they get older.

This is noticeable in Japan. For 16 years, the population has been in decline. The Japanese government predicts the population will fall by another 30% by 2070, slowing due to international migration rather than a boost in domestic fertility rates.3

Japan is worth paying attention to because it has the oldest population in the world. The median age is around 50 years old with around one-third of the country’s population over the age of 65. Compare that to the United States where the median age is currently 39.1.

Without replenishment, depopulation is rising across the country. Entire villages are disappearing as residents age and there’s no one left to maintain them. Some are going to extremes, replacing their nonexistent populations with lifeless dolls to give these declining villages some semblance of life.

On a national level, population loss makes it impossible to maintain a modern quality of living. More workers are drawing on social welfare programs than are paying into them, making the continuation of such programs fiscally and socially unsustainable. Young people are already struggling to find good employment opportunities. Because they’re fewer in number, they can’t maintain existing levels of consumption, much less pay into the welfare state at the level that is needed. For Gen Zers, this is as much a question of mathematics as it is personal responsibility. What do young people owe to the generations that preceded them?

But it’s not just a reduction of workers and taxable incomes that should raise the alarm. Fewer people means fewer opportunities to innovate and solve problems.

This is the basis from which neo-Malthusian thinkers like Paul Ehrlich have been consistently proven wrong. Economist Julian Simon famously challenged Ehrlich to a bet, showing that a growing population would not threaten the planet’s resources. In fact, he argued it would have the opposite effect – key commodities would become more abundant as a result of population growth, not less.

Simon argued that resources would become more abundant as the population grew because there would be more people available to contribute to coming up with innovative solutions that would solve scarcity. Humans, it turns out, aren’t just consumers and taxpayers. They are a valuable intellectual resource that can solve complex problems.4 The more humans you have, the greater the chance you’ll stumble on a solution that will solve problems pertaining to scarcity. Thus, it’s in the best interest of society to have more people – not less.5

This ethos is echoed by modern day technologists like Elon Musk. Musk believes falling birth rates will lead to the collapse of Western Civilization. This is a much bigger problem than potentially overshooting the earth’s reproductive capacity. Those in Muk’s camp believe technology can solve many of the problems attributed to the risk of outstripping the earth of its natural resources just as it has done in the past.

Current projections anticipate global population peaking sometime between 2050 to 2080. While this might not lead to the collapse that Elon Musk warns of, it will certainly lead to a change in how the economy is structured and the political apparatus that is designed to manage it. In turn, this will transform social relations between members of society, putting entire civilizations at risk.

Concerns about population collapse are gaining steam. Some countries are trying to do what they can to incentivize their populations to reproduce. Many of the policies they’ve implemented look at the problem from an economic perspective.

South Korea has the largest decline in birth rate of all the countries currently struggling with demographic decline. For the last two decades, South Korea has offered its citizens a slew of financial incentives to convince them to get married and have children. They include: affordable childcare, better parental leave, and cash payments.6 Here in the United States, the Trump administration is trying a similar approach, first by seeding investment accounts with $1,000 for children born between 2025 and 2028.

While financial incentives may help in the short term, they don’t acknowledge the actual problem. By focusing on managing population growth through narrative, neo-Malthusians turned reproduction into a spiritual and moral problem. It’s not just that kids are too expensive to have, it’s that young people don’t want to have children at all.

The real problem is philosophical in nature. Young people – especially modern women – harbor an antipathy towards having children at all. When you look at the problem of falling birth rates in the context of neo-Malthusian thinking, it’s not so much that people aren’t having children, it’s that they don’t want to have children period. It’s not economic pressure that is affecting global birth rates, but an idea. Over the past few decades, that idea has culminated in the act and defense of terminating unwanted pregnancies, preventing the future population from being born in the first place.

Fertility data points to fewer teenagers having unplanned pregnancies and women having children later in life. While these variables certainly contribute to falling birth rates, this is evidence of decline rather than the cause of it. The real reason fewer children are being born is because there is a growing number of people who aren’t permitting their children to exist in the first place.

Population decline is not a future problem that your children or grandchildren will have to grapple with someday. It’s something that has already started to manifest across the world and will likely affect you at some point in your life (if it hasn’t already).

In the United States, there are approximately 74 million Millennials and 71 million Gen Zers but only 51 million Gen Alphas.7 By 2043, all Gen Alphas will reach legal adulthood and be fully present in the workforce but there won’t be enough of them to support the retired Gen Xers and Millennials who are expecting to rely on them.8

The precipitous drop in birth rates is attributed to a number of factors: the cost of living crisis, inability of young people to find good paying jobs and buy houses, feminism and the current political environment, fears about the climate, social media, dating apps, addictions like pornography – the list can go on. Any one of these variables explain why fewer and fewer young people are coupling up, getting married, and thus having children.

But when you dive deeper into the data, it reveals that none of these factors are the cause for population decline. Despite all of these problems young people are still having children. The problem isn’t a barrier to having children but a desire. They don’t want the kids they are having. Instead of keeping them and raising them to adulthood, they’re terminating them before they have a chance of coming into full existence.

Put another way: we don’t have a problem of too few births, we have a problem of too many abortions.

In 1973, the Supreme Court legalized abortion. Since then, 25 million fewer people have been born than are needed to keep up with the replacement rate.9 This figure is attributed to a declining fertility rate, however, if you unpack the statistics around abortion, it suggests the fertility crisis may not be as pronounced as we’ve been told it is.

Thanks to the permissible environment created by Roe v. Wade, approximately 63.6 million children have been legally aborted in the United States. The pro-life organization Live Action reports that this has affected one-third of Gen Z. If Gen X parents hadn’t aborted their children, there would be as many as 86 million Gen Zs alive today, far exceeding the replacement rate.

The same pattern holds true for Millennials. If all Millennial children had been allowed to live, there could have been as many as 98 million Millennials alive today.10 Live Action reports 50 million Millennials and Gen Zers are missing due in large part to abortion:

Over 50 million Millennial (24.5 million) and Generation Z (26 million) babies have been killed by abortion combined, according to data published by Planned Parenthood’s former “special affiliate” and research arm, the Guttmacher Institute.11

When you factor in the children who never had a chance to live due to abortion, you’ll notice there is a population surplus rather than a deficit. Approximately 25 million extra people needed to be born from 1973 to present in order to maintain the replacement rate. Instead, 63.6 million were aborted. If just half those abortions never happened, there would be enough Millennials, Gen Zers, and presumably Alphas alive to maintain Western Civilization.

It would appear that in America at least, the problem is philosophical rather than economic. Neo-Malthusians created a morally justifiable environment for younger generations to begin terminating their offspring. Once a legal framework was established in the 1970s to make it fully permissible, they did just that. Now, several generations later, fewer children are being born.

You can see evidence of this in the birth rates themselves. While teenagers are having fewer unplanned pregnancies and women are waiting into their 30s to have children – reducing the over number they can have – birth rates haven’t plummeted across the board. The women who want children are still having children, specifically married women. These women have a substantially higher birth rate than unmarried women.

While a decline in marriage is certainly a contributing factor to plummeting birth rates, a larger concern is the attitude women have toward family formation in general. This is a philosophical problem rather than an economic one. Polling data following the 2024 presidential election suggest there’s a fundamental misalignment between young men and women when it comes to raising families:

Young male Gen Z voters who voted for Trump rated having children as the most important thing in their personal success definition; Gen Z women who voted for Harris, meanwhile, ranked this as the second least most important thing.12

The growing number of women who don’t want children is the most likely explanation for why fewer children are being born – not a wholesale decline in fertility rates. Since the feminist revolution of the 1960s and 1970s, women have been told they need to liberate themselves from the confines of the home and the prospect of being bogged down by the burden of raising children. Access to abortion and the proliferation of it as a morally justifiable and politically correct action has given women the grounds to do just that.

Normalization of abortion has made it easier for young women to terminate pregnancies they don’t want. In doing so, they’ve reduced the number of children that are being born. The outcome is the collapse in the birth rate we’re seeing today. Until the idea of abortion is eliminated – that it is morally and ethically acceptable to terminate life before it comes into existence – it is unlikely the problem of falling birth rates will be solved anytime soon.

Final takeaway.

Plummeting birth rates are a major issue across the globe. Unless you produce future generations to carry on your legacy, your way of life will eventually die out. Societal collapse – and the overall extinction of one’s civilization – are very real threats.

The decline in global population is often chalked to being a resource problem. Young people don’t have the incomes or homes they need to start families of their own. While this certainly explains the delay in family formation, it doesn’t explain the forfeiture of having children altogether.

The sheer number of people aborted in the United States since the passage of Roe v. Wade suggests America has an abortion problem – not a fertility problem. The women who want families are doing just that: they’re getting married and having children. It’s the growing number of women who aren’t getting married or having children that should be sounding the alarm.

More and more young women see family as an obstacle to eliminate rather than an opportunity to celebrate. Abortion is simply a means to that end. As a result, it ensures fewer children are born in the first place, reducing the birth rate. This is something we’ve seen evidence of over the last 50 years since Roe v. Wade made abortion both legal and socially acceptable.

This is a philosophical problem – not an economic one. While financial incentives may make it easier for existing families to have more children, it won’t entice the women who view children as an impediment to their lifestyles and careers into having children in the first place.

To grow the population, you have to make it easier and desirable for young people to have children. To do that, you have to kill the idea of abortion.

Politically, this is not feasible. The conversation around abortion has been caught up in access to women’s reproductive health. Restricting the former will invariably result in the perceived restriction of the latter.

Killing the idea of abortion requires replacing it with a different idea that is better. That change has to come from within culture. This is why the conversation around feminism is more important now than ever before. While at the surface, it may appear to be a disagreement over the role of women at work and in the home, foundationally, it’s about the perpetuation of Western Civilization. The women in our society have to want to reproduce. If they don’t, our way of life will go extinct.

With AI looming on the horizon – threatening to destroy millions of jobs in the process – we have a once in a lifetime opportunity to redefine our social contract. While there will be many adverse side effects of this, there will also be some silver linings.

One of those silver linings is an opportunity to elevate the role of family within society. Millions of feminist women will lose their jobs, effectively bringing the era of the girlboss to an end. Naturally, some will fight back. Others, however, will see it as an opportunity to reclaim motherhood and marriage. We are currently sitting in a Goldilocks window where we can shape the narrative that emerges from this period of disruption, a narrative that promotes family above career.

There’s no doubt that birth rates are collapsing and that a decline in population is not necessarily a good thing. But because this largely stems from an idea rather than a specific economic policy, we have the ability to course correct on our own. The question now is will we?

Will we take the initiative to reverse population collapse now before it’s too late?

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

For just about a year now, I’ve been publishing a weekly newsletter that highlights how the economy is transforming on Main Street. I’ve covered topics like tariffs, inflation, Federal Reserve policy, and everyone’s favorite – jobs.

This week will be the last newsletter for a while.

In less than a month, my husband and I will be welcoming our first child into the world. While this is incredibly exciting, there’s a lot I need to do between now and then.

This isn’t goodbye – the newsletter will pick back up at some point. But I’m going to be honest, I’m not quite sure when that will be. This is a thrilling new chapter of life but it’s filled with a lot of unknowns. I’ll need some time to transition into motherhood before I can commit to publishing a weekly newsletter again.

I hope you’ve gotten as much value out of reading this weekly newsletter as I have in writing it. Thank you for allowing me to pop into your inbox each and every week.

See you later,

-Amanda

1️⃣ This Latest AI Update Could Really Be Too Big Too Fail

2️⃣ The Real Measure of Inflation

3️⃣ Are You Upper Middle Class?

1️⃣ This Latest AI Update Could Really Be Too Big Too Fail

They’ve done it. An AI company has created a model so powerful that they’re afraid of what it could do if it was released to the general public.

Earlier this week, Anthropic announced the launch of Project Glasswing after its latest model, Mythos, revealed major cybersecurity vulnerabilities in “every major operating system and web browser.” The effort allocates $100 million in usage credits and brings together 40 companies – including NVIDIA, Amazon, Google, Apple, and Microsoft – to basically begin patching all of the software that powers the modern global economy.

This is a pretty big deal. Remember the AWS outage late last year? That was the result of an internal bug. For 14 hours, sites like Snapchat, Robinhood, Zoom, and Slack were offline. According to one estimate, the outage cost sites anywhere from $5,000 to $9,000 per minute.

For context of how big of a deal this is, 67% of all internet traffic runs through just three cloud computing platforms: Amazon AWS, Microsoft Azure, and Google Cloud. Anthropic believes Mythos has the ability to find – and exploit – software vulnerabilities like the one that happened with AWS last year. But Anthropic warns:

Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely. The fallout—for economies, public safety, and national security—could be severe. Project Glasswing is an urgent attempt to put these capabilities to work for defensive purposes.

This has the potential to be so catastrophic it got the attention of Federal Reserve Chairman Jerome Powell and Treasury Secretary Scott Bessent. According to CNBC, the duo called a meeting with the major bank heads to discuss the risks:

Federal Reserve Chairman Jerome Powell and Treasury Secretary Scott Bessent met with major U.S. bank CEOs this week to discuss the possible cyber risks raised by Anthropic’s Mythos model, CNBC confirmed Friday.

The bank heads were already in Washington, D.C., for a Financial Services Forum board meeting, but a special gathering was called to discuss Mythos,

My last salaried W-2 job was working in compliance and risk management at a community bank. Most banks work with vendors to provide any number of financial services to their customers. Fiserv, a company that processes debit card transactions, is one example of such a vendor.

While I was working at the bank, Fiserv went down for a couple of days. That meant if you had a checking account at the bank and went to the store to buy groceries, you couldn’t use your debit card at checkout. Like cloud computing, payment processing is highly concentrated to just three core providers. There’s a good chance whoever you bank with partners with Fiserv too.

The modern digital economy is really just a patchwork of software-as-a-service products. If a vulnerability is discovered in one of these products, it could cause any number of problems including lost revenue, the inability to access bank funds, and disruption to healthcare services. If sustained, it could result in real damage that could have significant repercussions across the economy.

To see an example of what this looks like, check out this article:

It dives into the impact of the June 2024 cyberattack on CDK and how that affected the overall economy. More importantly, it talks about what happens when we’re forced offline for a long period of time.

This isn’t something to necessarily panic about just yet, but it is something to start making yourself aware of it you’re not already. While Anthropic acknowledges their latest model could become a liability, it’s hard to say if other frontier AI companies will do the same. Now that the technology to find and exploit vulnerabilities is there, it could spread.

If you don’t have analog redundancies built into your personal life or business, now would be a really good time to start thinking about that. Keeping cash on hand, a stocked pantry, and physical documents could be beneficial if a major outage happens and you’re locked out of your digital accounts for any period of time.

2️⃣ The Real Measure of Inflation

According to the latest government data, inflation is going up. Yahoo! Finance reports:

Headline Consumer Price Index (CPI) inflation clocked in 3.3% higher than a year ago while rising 0.9% on a monthly basis in a rapid acceleration from February’s levels, according to Labor Department data released Friday. Economists surveyed by Bloomberg had anticipated a 3.4% increase from a year ago and 0.9% from a month prior.

With the war in the Middle East continuing to drag on, this shouldn’t come as much of a surprise. Energy prices have skyrocketed as ships are still struggling to get through the Strait of Hormuz. According to the latest data from the Bureau of Labor Statistics, energy prices rose 11% last month.

Because energy powers the global economy, the higher cost was bound to show up in consumer pricing data eventually. But this doesn’t tell the full story of inflation. To see the real impact, you have to look across a much longer time horizon.

Fortunately, we can do just that. Last month, I shared that the U.S. Postal Service would begin instituting an 8% surcharge on packages mailed through 2027. Now, they’re raising prices again, this time upping the price of stamps. The Wall Street Journal reports:

The U.S. Postal Service said it plans to raise the price of a First-Class Mail Forever stamp about 5% to 82 cents starting July 12, up from 78 cents, as the agency tries to navigate a financial crisis.

While economists cherry pick which data to include in monthly inflation metrics, a stamp is a constant. It’s something that has existed since the founding of the U.S. Postal Service in 1775, and it is something that has survived several wars, global pandemics, and major economic decline. If there’s one constant in our economy it is the cost of sending a letter in the mail.

From 1885-1932, the price of a stamp held steady at just 2 cents (equivalent to $0.67 in 2026 dollars). By 1974, the price increased fivefold to 10 cents (also equivalent to $0.67 in 2026 dollars). Once the U.S. went off the gold standard in 1971, the cost of a stamp skyrocketed. From 1971 to 2026, the price of snail mail has gone up by 925%.

To put this into perspective, the U.S. dollar has lost 97% of its purchasing power since 1913 but 90% of its value since 1971. The bulk of our currency debasement – and much of the inflationary pressure we’re facing today – is a lagging consequence of economic decisions made during the 1970s.

The current financial peril of USPS is a story worth continuing to pay attention to. This is the one government agency that has a legitimate purpose and quite literally touches the home of every single Main Street American. It’s a bellwether for the state of the economy. If the prices at USPS are outpacing publicly reported inflation data, there’s good reason to believe everything else is too.

3️⃣ Are You Upper Middle Class?

Apparently more Americans can be classified as upper middle class than ever before. The Wall Street Journal reports:

America’s middle class is becoming wealthier as more families scale the economic ladder into higher-earning groups. New research shows that the ranks of the affluent have grown markedly over the last 50 years or so, while the lower rungs of the middle class have shrunk.

In 2024, about 31% of Americans were part of the upper middle class, up from about 10% in 1979, according to a report released this year by the right-leaning American Enterprise Institute.

According to the study, a family of three with an income of $133,000 to $400,000 in 2024 dollars (equivalent to $140,000 to $421,000 in 2026 dollars) is considered upper middle class. The Journal suggests this rise in affluence is benefitting everyone equally:

The gains span generations. Many baby boomers, born to parents who grew up in the Great Depression, are living well on their savings, aided by steady Social Security checks and decades of stock-portfolio gains that they can now tap. Millennials, who everyone worried would be permanently set back by the 2008-09 financial crisis, are earning solid incomes, buying homes and surpassing their parents.

This kind of upward mobility would make you believe America is thriving. The study argues that while it may seem like the middle class is shrinking, that’s because the upper middle class is actually growing.

The lived experience on Main Street, however, tells a very different story. Retirees are burning through their savings, looking for gig work as a form of supplemental income.

Millennials are buying homes, but they’re doing so much later than their parents did. The average age of a first-time homebuyer is 40 years old – the current age of elder Millennials. When their parents bought their first homes in the 1980s, they were in their mid- to late-20s.

The birth rate is plummeting. New data confirms the birth rate fell to a historic low last year at just 1.57 births per woman. With a replacement rate of 2.1, Americans quite literally aren’t having enough children. Gen Alpha – the children of Millennials – is poised to be the smallest generation on record.

Meanwhile, Gen Z is struggling across the board. Check out The Gen Z Employment Crisis for more on this.

In a separate analysis published in The Free Press last November, a portfolio manager for an asset management firm made the case that a $140,000 income is actually the new dividing line demarcating who’s in poverty and who’s not – not a line dividing the middle class.

Based on the view from Main Street, it doesn’t seem like a low six-figure income is the entry point to affluence. If anything, it seems like the “upper middle class” has been absorbed into a new wealthy class – a class with discretionary income to spend. Everyone else below them is just doing what they can to get by.

What do you think? Do you consider a $130,000 salary for a small family to be middle class or upper middle class? Share your thoughts in the comments.

Leave a comment

📚Last Week’s Subscriber Articles

📽️ On YouTube

📩 Support The Big 5️⃣

I hope you enjoyed today’s newsletter. If you found value in it but aren’t a Tomorrow Today subscriber, sign up below to get this sent to your inbox every week.

Subscribe now

Earlier this week, OpenAI released a framework to guide the conversation about how We the People should start talking about AI. Specifically, all of the ways AI is expected to disrupt how we work and how we coexist with one another.

Titled Industrial policy for the Intelligence Age, the 13-page framework suggests artificial superintelligence is already here. Rather than pumping the breaks on research or striving to contain it, we should begin implementing policies that allow everyone to flourish. For OpenAI, this starts with building an open economy and a resilient society.

The conversation around AI has noticeably shifted over the last three years. Rather than prioritizing safety, AI companies are now moving full steam ahead, releasing new models as fast as possible in an effort to be the first to get to superintelligence.

As this Substack essay notes, the race to build artificial superintelligence has changed the conversation about AI. Right now it’s less about aligning AI with human goals and more about putting the right policies in place so we can coexist with it. OpenAI’s policy framework says it wants you and I to participate in the democratic process that will shape the new political economy and political philosophies that are bound to emerge.

This likely won’t happen. Main Street isn’t concerned about the concentration of computing power within a handful of Silicon Valley tech companies, much less how the growing disparity in wealth will transform our system of governance as we know it. Instead, Main Street is concerned about the rising cost of living and whether or not they’ll have a job in the coming years (and rightfully so).

As of this moment, the average person isn’t even using AI. They don’t think AI is a risk to them personally. Even if you don’t personally use AI and aren’t concerned about it taking your job, there are bigger, more fundamental questions we need to be considering about the rise of artificial intelligence and the role it will play in our lives.

One of those questions is whether or not AI is compatible with our existing way of life. Can a digital world built by technologists for the benefit of technologists preserve a system of private property rights, open markets, free speech, and electoral representation in the halls of government that you and I currently benefit from?

It’s unlikely. OpenAI’s policy framework tries to paint an optimistic picture. While the company admits our social contract will likely have to be renegotiated as AI eliminates jobs, OpenAI argues that with the right policies, artificial intelligence can seamlessly integrate within our existing institutions, leaving our system intact.

At its core, artificial intelligence is anti-democratic in nature. Only a handful of elite technophiles understand how it works much less how to adequately regulate it; those same technophiles have a vested financial, spiritual, and philosophic interest in ensuring the world AI creates is one that is shaped in their image.

The system as it currently exists is already weak. It’s been degraded, in part, by Silicon Valley tech giants who, over the past few decades, have taken an extraordinary effort to obtain unprecedented power and control over our lives. Adding an anti-democratic superintelligent force to the mix will only amplify existing problems.

A new system governed by new beliefs about how the world should operate is going to replace the democratic, capitalist system of governance that currently exists.

With much of Main Street powerless to make a meaningful contribution to a New Industrial Policy that will shape this new system, OpenAI is relying on policymakers to do the heavy lifting. Here again there is reason for pessimism. In 2026, our elected leaders can’t agree that men and women are biologically distinct beings with unique roles and functions to play in society. If they can’t come to terms with the laws of biology, how are We the People supposed to entrust these same individuals with negotiating an entirely new system on our behalf?

There are serious conversations we need to have about the role artificial intelligence is going to play in our lives. The system as it currently exists is broken. AI didn’t break it but it likely won’t fix it either. A new system will have to emerge; AI is merely the cover that creates a permissible environment for Silicon Valley to create that new system in its image. What you can be sure of is that We the People will be part of this system but we will not be the primary beneficiaries of the techno-abundance that is promised to emerge from it.

Policy documents like this that are written by AI companies and their leaders signal what lies ahead. This essay will dive into OpenAI’s proposal for the creation of a New Industrial Policy. It will look at the historical record to show what is likely to happen. While you may hope for the continuation of democratic governance, free markets, and private ownership of property, our existing institutions cannot guarantee the continuation of these things in an AI-powered world.

There will be clear winners and clear losers in the new system that lies ahead. Unless you own the digital means of production – or the infrastructure that powers it – there’s good reason to be skeptical that you’ll benefit much from the new system that a New Industrial Policy promises to create.

OpenAI acknowledges the creation of artificial superintelligence will be massively disruptive, not just to the economy but society as a whole. They promise we’ll all benefit from it. But are the benefits worth the risks?

OpenAI begins their policy document by arguing the gains from artificial intelligence will outweigh the risks. According to OpenAI:

The promise of superintelligence is extraordinary. Just as electricity transformed homes, the combustion engine remade mobility, and mass production lowered the cost of essential goods, superintelligence will speed up scientific and medical breakthroughs, significantly increase productivity, lower costs for families by making essential goods cheaper, and open the way for entirely new forms of work, creativity, and entrepreneurship. (1)

There may be some truth to this. AI will speed up scientific discovery, leading to medical breakthroughs. But does that mean the benefits of those breakthroughs will trickle down to the average person? I’m not so sure about that.

Let’s take the mass production of essential goods as an example to illustrate why skepticism towards this kind of techno-optimism is necessary. Beginning in the late 1960s, the United States began looking at overseas markets to produce consumer goods like electronics and clothing. From the 1970s onwards, American producers have been outsourcing production of just about everything to lower cost centers in Mexico and China.

For workers, this meant the loss of millions of jobs and the middle class standard of living afforded to the workers who once held those jobs. For consumers, this meant the replacement of durable goods with cheap, low-quality products that have to be constantly replaced.

These days, instead of buying a pair of jeans from a domestic producer that employs American workers – like Origin – consumers are more than happy to continue purchasing cheap clothes that aren’t designed to last more than a couple of wash cycles from the likes of SHEIN or H&M. Can we chalk up the economic gains attributed to the offshoring of American productive power – gains that are largely achieved by harnessing the power of a globalized, wage slave workforce – as a win for the average person? No, I don’t think so.

Despite a century of technological progress, our society has not improved. Individual technologies and processes may have improved for segments of the population, but in aggregate this hasn’t translated into an improved quality of life. Take medicine as another example. Our medical technology has vastly improved but the cost of accessing medical services has skyrocketed. Meanwhile, the corruption of Big Food, Big Pharma, and Big Ag has left us subsisting on nutritionally deficient ultraprocessed food-like products that leaves us chronically ill and susceptible to disease. Sure, you and I may have better access to cancer treatments then our grandparents did, but we also have more instances of cancer than they had to deal with.1 Is that something we can chalk up as a win?

Companies like OpenAI are betting that superintelligence will be a net gain for American society and human civilization without being fully transparent about the actual risks involved in building it. As I wrote in If Anyone Builds Superintelligence, Humanity Will Go Extinct, it’s impossible to build superintelligence without exposing all of humanity to the risk of extinction:

The techno optimists pushing for AI development argue that AI will enable humans to flourish. And to be fair, it could….But just because there are positive use cases for it doesn’t mean the AIs we develop – especially the ones that can think and act on their own – will always do so in the interest of humans.

If Anyone Builds It, Everyone Dies was written precisely because Yudkowsky and Soares believe that human flourishing is not aligned with the goals of AI. AIs are tasked with solving problems and trained to succeed in solving those problems. They are rewarded for finding the best outcome, not the outcome that leads to human flourishing. It’s possible – if not likely – that someday an AI will approach a problem that it can only solve by eliminating superfluous humans. If it is enabled to come to that conclusion on its own, everyone dies and there’s nothing you or I can do about it.

The question isn’t whether or not superintelligence will lead to a major medical breakthrough that’s going to be beneficial for society. The question is, is a technology with an extinction risk worth building in the first place? We have the power within ourselves to clean up our food and reform our medical system. If we focused on fixing what we already have rather than building something new, most of the benefits promised by companies like OpenAI wouldn’t even be necessary in the first place.

OpenAI argues conversations about how to handle superintelligence should be inclusive and democratic. Yet everyday people aren’t being given a say in how it’s being developed. AI companies continue to work towards superintelligence, full steam ahead, without our input all while building new infrastructure none of us asked for.

The point of OpenAI’s policy proposal isn’t so much to establish a new policy framework but to merely start the conversation. They’re passing the buck onto our elected officials to do the heavy lifting for them. The company argues this process should be democratic and inclusive because everyone, everywhere will be affected by it:

At OpenAI, we believe we should navigate [the transition toward superintelligence] through a democratic process that gives people real power to shape the AI future they want, and prepare for a range of possible outcomes while building the capacity to adapt. (1)

The great irony here is that the process of giving life to artificial superintelligence has been anything but democratic or inclusive. Aside from their shareholders, We the People didn’t vote for the likes of Sam Altman, Elon Musk, Mark Zuckerberg, or Dario Amodei to lead us into the AI revolution. We have had no say in what they’re building, much less why they’re building it. Even though many of their peers have come out publicly sounding the alarm on the risks involved, they continue to race towards bringing artificial superintelligence to life, regardless of the cost it could have on your life or mine.

And yet, OpenAI has published a policy document effectively calling on us to take action to mitigate the consequences of its decisions. Even though OpenAI hasn’t included We the People in the decisions they’ve made up until now, they want us to talk about policies we should put in place that blunt the social and economic damage their product is going to cause to our lives. They want our process of democratic governance to solve a problem they’re creating.

Recent history tells us that we can talk all we want, but talking about the problem isn’t going to lead to the “New Industrial Policy” OpenAI thinks we should be putting all our energy into creating. Take the creation of Google and Facebook as two examples. The original mission of Google was “to organize the world’s information and make it universally accessible and useful”2 while Facebook was created “to give people the power to share and make the world more open and connected.”3 Where did that get us? Two massive corporate monopolies with the ability to control the flow of information, censor it, distort it, and leverage it as an addictive force across all of society.

None of us had a say in how Google, Facebook, or any of the other tech giants of Silicon Valley evolved. No ballot initiatives were voted on and no public conversations were had. Even though whistleblowers have since come forward acknowledging their products are harmful and studies have linked social media usage with increased social isolation and declining mental health, nothing has been done. Why should we expect artificial intelligence companies to behave any differently?

The answer is we shouldn’t. Whether you like them or not, Bernie Sanders and Alexandria Ocasio-Cortez – two of the very policymakers OpenAI calls to action in their document – have introduced a temporary ban on data center construction to buy us time so we can create a federal AI safeguard protocol. This seems reasonable given mounting public concerns associated with the impact of nascent data centers on day-to-day life. And it’s the very type of policy action OpenAI is advocating our elected officials take.

And yet, OpenAI hasn’t publicly commented on a temporary moratorium on data center construction. None of the AI companies have. Instead, a statement made by the Information Technology and Innovation Foundation, a Big Tech-funded think tank based in Washington, DC, is doing their talking for them. According to their position, such a moratorium is baseless, predicated on fear rather than fact:

This bill justifies a moratorium based on several well-worn anxieties—that AI is an existential threat, that data centers burden the pocketbooks of American families, and that they undermine jobs—but none of these, pursued in good faith, lead to halting data center construction. It is clear that the authors started with the moratorium and then cast a net of disparate fears wide enough to build support for it.

Congress should not mistake a grab bag of loosely related fears for a legitimate case for a moratorium. If these challenges are real, they deserve solutions that actually address them—not a construction freeze that only undermines the American digital economy.

The very legislative process that OpenAI argues we should invoke to set a “New Industrial Policy” is being openly derided by unelected lobbyists and policy influencers in Washington, DC. These informal policymakers, funded by the tech companies themselves, are the ones who actually shape the policies that get passed into law – not our elected officials who sit on Capitol Hill.

We can’t agree on how to build the infrastructure needed to power artificial intelligence. How then, are we supposed to come together and democratically debate the policies, values, and safeguards that should be put into place to enable human flourishing in an AI-powered world?

Answer: we’re not. That’s the point. Policymaking in the context of AI superintelligence is little more than a performative exercise to give the illusion of public consent and control. In reality, OpenAI knows a new system is going to have to emerge and that new system cannot possibly include giving power to We the People.

OpenAI’s New Industrial Policy assumes we can adapt the system as it currently exists to operate within a world powered by superintelligence. It doesn’t consider the system is already broken. With the fracture lines already in place, history tells us the system will first break, then be replaced.

The underlying premise of OpenAI’s policy document is to create a New Industrial Policy around the system as it currently exists. The company argues we can just come up with a handful of policies to blunt the impact of change:

We are entering a new phase of economic and social organization that will fundamentally reshape work, knowledge, and production. It requires not just incremental policy responses but ambitious policy ideas for tomorrow that we must start discussing today. (4)

History tells us it’s unlikely this will work. While the paper invokes numerous examples of how humanity has navigated previous technological transformations, it’s woefully ignorant of the impact those transformations had on society over a longer period of time. And it certainly doesn’t consider the premise that AI or not, our system is too weak to undergo reform of this magnitude in the first place.

Let’s look at the invention of the printing press as an example. Prior to the printing press, there were only two ways to pass information from one generation to the next: through oral tradition or handwritten text. This created an asymmetry in access to information, much like ownership and investment in artificial intelligence will create asymmetries in access to opportunities and wealth in the future. Unsurprisingly, those who had access to information and the ability to take action from it – bureaucrats and religious officials – wielded tremendous power within society.

That system changed around 1440 when a German goldsmith named Johannes Gutenberg created a device that used movable type to print words onto paper. For the first time in human civilization, anyone anywhere could access knowledge in the lingua franca of the time. The monks who painstakingly copied the Bible by hand and the Catholic Church that held a monopoly on interpreting the Bible for its parishioners began to lose power as a result.

Over the course of the next three and a half centuries, the feudal order that had dominated Europe – a system centered around the church, the monarchy, and landed aristocrats – gave way to a new system of free enterprise, private ownership of property, representative government, and a direct relationship with God. The incumbent feudal elites tried to control the dissemination of information – burning heretics at the stake and suppressing peasant revolts – but in the end, it didn’t work. Europe today looks nothing like it did just a few centuries ago.

This is the type of transformation that is going to unfold before our very eyes but instead of happening on the scale of centuries it will happen in decades – if that. The system as it exists today cannot exist in an AI-dominated economy of the future. OpenAI argues we need to build a “resilient society” that is built on “trust, accountability, and auditing,” however the institutions and people responsible for maintaining trust and holding organizations like OpenAI accountable are already broken.

Look no further than the welfare fraud scandals in Minnesota, California, Maine, and New Jersey as evidence of this. Our democratic values, institutions, and policymakers are already incapable of preventing fraud, waste, and abuse within existing government programs. Adding a new AI regulatory framework – to include a wealth distribution program – isn’t going to solve that problem. If anything, it will only make things worse.

The very essence of AI itself and the ability to contain it will require levels of monitoring, management, and surveillance that are inherently anti-democratic in nature. Training data is the lifeblood of the very research and development programs OpenAI is leading to build artificial superintelligence. And yet, that training data is seized from We the People, the users of tech products, without our consent. (For a better illustration of this point, I refer you to this essay on how reCAPTCHA was used to train Google’s AI to power Waymo).

History tells us a technological revolution of this magnitude will lead to the emergence of a new system rather than the reformation of an existing system. Based on the requirements of this new system – data acquisition and immense capital flows – asymmetries in wealth and information will emerge. It’s more likely that we’ll return to a system of feudalism than reform the system as it currently exists. And if that happens, OpenAI won’t have any need for We the People to participate in policy discussions at all.

Final takeaway.

OpenAI isn’t looking for your contribution to creating a New Industrial Policy – it’s looking for your consent in writing a social contract that will govern our lives moving forward. Much of the techno-optimism and promises of human flourishing touted by companies like OpenAI is merely to get your buy-in, granting them permission to do this on your behalf.

A new system is bound to emerge. The question is what will this system look like? It’s becoming increasingly likely that this new AI-powered system will be feudal in nature. Like the past, those with means of production – in this case frontier AI companies – will control your ability to participate in the new economy and society they create. That participation will come with a fee – a literal rent – and the unilateral collection of your data.

As a bare minimum, this new techno-feudal order will guarantee your subsistence but not your flourishing (until you’re rendered superfluous, of course). Universal basic income has been the only policy discussed with regard to the displacement of workers and loss of income. That will provide for the lowest common denominator rather than allowing each and everyone of us as unique individuals to rise to our greatest potential.

As OpenAI alludes to in their policy document, universal basic income could come from overhauling the tax code. Instead of having unemployed workers fork over a percentage of their non-existent income to federally managed social safety net programs, a new tax base will have to be created, likely from the tech companies themselves. The revenue generated from this will be distributed to displaced workers through the expansion of the welfare state. While this may seem like the most logical course of action, as history shows, it’s unlikely this is how universal basic income will actually be implemented.

In making the case for a New Industrial Policy, OpenAI points to the redistribution of wealth that took place at the turn of the 20th century. Recognizing that massive wealth disparities had emerged during the industrialization of the United States, the government had to take action to ensure no one was left behind. OpenAI writes:

While those transitions ultimately created more prosperity, they required proactive political choices to ensure that growth translated into broader opportunity and greater security. For example, following the transition to the Industrial Age, the Progressive Era and New Deal helped modernize the social contract for a world reshaped by electricity, the combustion engine, and mass production. They did so by building new public institutions, protections, and expectations about what a fair economy should provide. (3)

Open AI lauds the government reforms invoked during the Progressive Era and New Deal. We the People remember those policies are being benevolent in nature – they took from the ultra-rich and gave to the poor. In reality, they paved the way for wealth transfers in the other direction and can explain why so many middle class Americans are struggling to this day.

The 16th Amendment – ratified in 1913 – instituted an income tax, something that had been explicitly prohibited by the constitution. Originally, the income tax was sold as a tax on the wealthy. It was only to be levied on the incomes of high-earning individuals making more than $3,000 at a rate 1% with a maximum tax of 7%. Today, the income tax now applies to anyone earning an income, not just the ultra-wealthy. The tax that is levied is anywhere from 10-37% – a far cry from the 1-7% levied a century ago.

Likewise, much like new bureaucratic agencies will have to be created to manage AI policy and universal basic income, the New Deal created a whole slew of “Alphabet Agencies” like the Social Security Administration that helped get America out of the Great Depression. While those agencies may have been necessary at the time they were unfunded mandates that have ballooned in the decades that followed. Today, Social Security is the largest expenditure in the U.S. federal budget.

To fund the programs of the New Deal and the expansion of those programs to where they are today, the policymakers of the time looked to Main Street America – not the ultra-wealthy. On March 9, 1933 President Franklin Delano Roosevelt signed the Emergency Banking Act into law effectively giving the federal government the authority to ban the private ownership of gold. A month later, he signed Executive Order 6102 requiring private citizens to surrender their gold or face criminal penalties. Once the mechanism was in place to fund new government programs, FDR began signing relief programs into law that culminated in what we remember as the New Deal.

It’s assumed that AI companies will compensate displaced workers through taxes those companies voluntarily surrender, managed by a benevolent bureaucracy, in the same fiat currency those workers are accustomed to earning in their current jobs. That’s not going to happen. In a techno-feudal system where asymmetries in power, wealth, and information exist, money itself will change. As will the rules on how you and I are permitted to spend it.

It’s likely a central bank digital currency will become the dominant form of exchange. This currency will be programmable which means it will be encoded with rules and stipulations about how it can be spent. Today, that may manifest as prohibiting SNAP recipients from spending taxpayer funded dollars on soda and sugary snacks – a policy most people can get behind. But tomorrow, that same limitation can be redirected towards you, with an AI system limiting the type of food you can buy based on your BMI or preventing you from spending at all based on something you posted on social media.

But CBDCs won’t be the only form of money in the AI economy. At the very end of their paper, OpenAI invites readers to contribute their ideas to the broader conversation of creating a New Industrial Policy. In doing so, OpenAI eludes to another form of currency that could emerge:

Establishing a pilot program of fellowships and focused research grants of up to $100,000 and up to $1 million in API credits for work that builds on these and related policy ideas. (13)

API credits – or credits in general – could represent a new type of currency that will be part of the AI economy moving forward. If the thesis is correct that we are heading into a new techno-feudal order where tech companies collect rents from users, those companies will be in a position to offer usage credits for their platforms rather than paying a bunch of superfluous people a cash dividend on productivity gains they did nothing to generate.4 In a non-democratic society, you and I would effectively be powerless to stop them.

A surveillance architecture rather than a policy framework is the most viable way to ensure the needs of We the People are met in an economy where we are incapable of meeting the needs for ourselves. This will naturally invite authoritarianism to replace governed consent of the masses.

While those behind OpenAI’s New Industrial Policy may believe they are well-intended, the reality is it will be impossible to democratically create a policy framework that can benefit everyone fairly and equally. There will be losers in the AI economy and it’s likely the losers will far outnumber those who count themselves as winners.

A new system will have to emerge. That new system will not be able to reform the system as it exists now, much less coexist alongside it. These are important conversations the average person isn’t really thinking about today, but are questions we all need to start contemplating sooner rather than later.

OpenAI’s policy document is brief. Give a read and some thought as to whether or not their approach is feasible — or desirable.

Read OpenAI's Policy Document Here

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

2️⃣ Is California Too Big To Fail?

3️⃣The Starbucks Solution

For those of you celebrating this weekend, Hoppy Easter!

1️⃣ Last Month’s Not-So-Blowout Job Numbers

The latest job numbers are in and everyone is calling it a blowout.

Yeah….about that.

Based on the amount of revisions we’ve seen over the last year and where jobs are being added to the economy, I’m not sure I would call it a ‘blowout.’ The monthly job numbers don’t show new jobs being added to the economy. It just shows an increase in employment. It’s possible to increase the number of people counted as employed without adding new jobs.

Case in point: healthcare. Of the 178,000 jobs added last month, almost half went to health care. But of those jobs, half were just workers coming back from a strike. They weren’t new jobs being added to the healthcare sector. BLS reports:

Health care added 76,000 jobs in March. Employment in ambulatory health care services rose by 54,000, reflecting an increase of 35,000 in offices of physicians as workers returned from a strike.

If you deduct the 35,000 workers returning from picket lines, the job numbers show an increase of 143,000 jobs – not the 178,000 being reported. And even then, that figure is probably being overreported. We’ll know for certain when the March revision comes out in a couple of months.

Speaking of revisions, this week’s job report also includes revisions for January and February. While there was a positive revision for January, the total revision across both months resulted in a net loss of 7,000 jobs.

The economy is losing more jobs than it’s creating. That’s because AI adoption is finally starting to show up in the jobs data – or at least it will start to show up soon. According to The Challenger Report from Challenger, Gray & Christmas, U.S. employers announced 25% more job cuts in March than they did in February. The firm attributes this to increase AI adoption:

In March, Artificial Intelligence (AI) led all reasons for job cuts, with 15,341 announced during the month, 25% of total cuts.

Moves by major tech companies also confirms AI is taking a front seat in how they think about their workforce composition moving forward. Oracle, a major player in AI development, announced it would be slashing upwards of 30,000 jobs or 18% of its global workforce. The Wall Street Journal reports:

Oracle began to significantly reduce its workforce Tuesday while it continues to build out costly data centers for artificial-intelligence development.

Meanwhile, Block CEO Jack Dorsey – who announced back in February that his company would be laying off 40% of their workforce to invest in AI – wrote in a recent blog post that he plans to swap out middle management with AI. (Chapter 7 in Human Optional is dedicated to this.)

In the post, Dorsey argues AI is fundamentally changing the organizational design of companies, not because AI can do everyone’s job but because it changes how information flows within companies in the first place:

What [“data-driven” management] lacked was a technology capable of actually performing the coordination functions that hierarchy exists to provide. AI is that technology. For the first time, a system can maintain a continuously updated model of an entire business and use it to coordinate work in ways that previously required humans relaying information through layers of management.

While not everyone agrees how and when AI will appear, there’s a growing consensus that AI is in fact here and it will indeed change the nature of work moving forward. The New York Times surveyed some economists who suggested they’re warming up to the idea that AI-related job displacement is on the horizon:

Recently, however, the message from economists has undergone a subtle change. Most still do not see much evidence that A.I. is disrupting the job market. But they are starting to take seriously the possibility that it could someday soon. If it does, they are worried that policymakers are not ready to respond.

We’re still in the early stages of the transition in the labor market from a knowledge-based services economy to an AI economy. Job numbers like the ones recorded from March will be touted in the press because they shed a positive light on the “strength” of the labor market. But when you parse through what the data actually says, it paints a much different picture.

Employment does not equal the creation of new jobs and not all jobs being created are good jobs to begin with. Keep this in mind as the experts continue to talk about economic recovery moving forward.

2️⃣ Is California Too Big To Fail?

Last week, a clip of David Frieberg from the All In podcast went viral after he suggested California could fall and take the whole Union down with it. According to Frieberg:

California is functionally bankrupt. People don’t realize how screwed California is, and I worry that if California falls, so does the Union. This is because for California to get rescued would be a big cost to Red states, and I think it creates in the years ahead a lot of tension.

I’m going to be honest, thinking of an entire state’s economy as something that is “too big to fail” is not something that was on my radar. But it makes sense: California’s economy is massive. It constitutes 14% of the overall GDP of the United States and if it were its own country, it would be the fourth largest in the world.

Frieberg claims California is going under because the state’s pension system is a massive unfunded liability. Retirees are owed some $250 billion to $1 trillion more than what they’ve paid into the system and the state doesn’t have a way to make up the missing money. According to current projections, nearly one-quarter of California’s population will be over the age of 65 by the year 2040, making this a very valid concern that will start impacting Californians sooner rather than later.

Of course, this is compounded by years of statewide malfeasance that is only now coming into the national spotlight. Nick Shirley’s latest video revealed rampant hospice fraud. Meanwhile, other analysts have uncovered similar levels of fraud in the state’s Medicaid program.

The state continues to grow without showing a return on taxpayer investment. California has been hemorrhaging residents – including wealthy billionaires – yet the state bureaucracy continues to expand. High cost overruns with the state’s high-speed rail project and nature bridge to nowhere has many wondering where their tax dollars are going in the first place.

If California is functionally bankrupt as Frieberg claims – and if Washington has to bail it out – that is going to be a very hard sell for otherwise fiscally responsible Red states like Florida and Texas.

All of this is happening while California’s policies are reducing the number of jobs available to its residents. Researchers at UC Santa Cruz recently published a new study on the impact of California’s 2024 minimum wage law for fast food workers. It found that employers are offering workers fewer hours while looking for opportunities to automate tasks to further reduce overhead. Meanwhile, the Wall Street Journal reports Hollywood’s job market is collapsing in real time because it’s too expensive to shoot movies there:

Hollywood studios are making significantly fewer movies and television shows than they did just a few years ago. The ones they do make are increasingly being shot in other countries and states that offer more generous tax subsidies.

The result: a 30% drop in employment from a late-2022 peak for actors, carpenters, costumers and the hundreds of other professions that make movies and TV shows, according to Labor Department data.

Politically, economically, culturally, and technologically, California for all intents and purposes is too big to fail. I don’t know if it would drag the Union down with it but it could. It’s clear the Golden State is facing some major structural problems. But given the state of California’s political leadership, it’s unlikely reforms will reach the West Coast anytime soon.

3️⃣ The Starbucks Solution

Starbucks baristas complained about unlivable wages and it seems like corporate may have heard them. In addition to adding more opportunities for baristas to collect tips from payments, the Seattle coffee chain is giving its workers the opportunity to earn performance bonuses. The Wall Street Journal reports:

The Seattle-based company also plans to provide bonuses to baristas whose stores meet performance goals, including targets for sales, operations and customer satisfaction. Store baristas and shift supervisors who meet the top level of Starbucks’s internal ratings for their location can earn a $300 quarterly bonus, or $1,200 for a full year, the company said.

Back in 2021, Starbucks workers unionized, complaining about unpredictable scheduling, low wages, and harassment. Since then they’ve been demanding concessions to make the worker experience more “equitable.”

Last summer, Starbucks baristas walked off the job near Disneyland after the company implemented a new dress code and continued to pay their workers “low” wages. Walk outs continued into the winter, designed to impede cafe operations during the chain’s annual Red Cup Day.

The new payment structure doesn’t meet the union’s latest demands but it does offer a middle ground for baristas who feel they aren’t being compensated fairly. According to the Journal’s reporting, Starbucks claims it pays its workers the equivalent of $30 an hour in wages and benefits. (If you take all of the benefits baristas receive, this is probably accurate).

From a cultural perspective, Starbucks is a market mover worth paying attention to. Like other progressive companies, it sends signals into the labor market, specifically what workers should expect from their employers. A performance based solution to the ongoing clash between workers and management is a novel workaround that could be replicated by other companies in the near future.

📚Last Week’s Subscriber Articles

📽️ On YouTube

📩 Support The Big 5️⃣

I hope you enjoyed today’s newsletter. If you found value in it but aren’t a Tomorrow Today subscriber, sign up below to get this sent to your inbox every week.

Subscribe now

There’s a lot of investment into developing AI right now. Last year, the Trump administration launched Project Stargate, a massive public-private partnership that will invest $500 billion in AI infrastructure projects. This year, tech giants like Amazon and Microsoft are independently pouring billions of dollars into building data centers of their own.1

Many see this as hype. Users of AI chatbots argue AI companies are overpromising more than they can deliver. While they’re getting better, AI chatbots still hallucinate. They prefer to make up answers rather than admit when they don’t know something. More concerning, however, is growing evidence that AI chatbots are deliberately lying to users. While Anthropic’s Dario Amodei continues to warn that AI will wipe out white collar work, because AI isn’t quite there yet, those in the hype camp think his warnings are overblown and the job losses he’s predicting won’t come to fruition.

Meanwhile, others are arguing that AI is nothing more than a bubble, deliberately engineered to make a handful of tech executives and their investors ridiculously wealthy. Companies building AI have a vested interest in its success. Of course they’re only going to show what AI can do, downplaying all of the mistakes it makes. The meteoric rise of chip manufacturers like Nvidia and all of the circular deals leveraged by OpenAI have led many conservative investors to wonder if this is all too good to be true.

Regardless of where you might personally sit, very few people actually have an opinion on AI at all. Back in February, the following infographic went viral on X. According to usage data, 84% of the world’s population has never interacted with an AI tool. Of the ones who have, the vast majority are using the free versions of chatbots like Claude, Gemini, and ChatGPT. While useful, these tools are outdated and don’t paint an accurate picture of where the technology is right now, much less how fast its developing.

To be fair, there are certain aspects of AI development that may be hyped or overblown. Many are right to be skeptical. But unless you’re part of the 0.04% of superusers who are building bespoke software with AI, you aren’t fully aware of the risks it poses to the economy, much less human civilization.

I started using AI – as a paid user – when ChatGPT made its public debut in November 2022. I didn’t fully understand what the technology was – and from a technical perspective, I still don’t – but I understand it enough to know how profoundly disruptive it will be.

Since then, I’ve tinkered with most of the AI chatbots that are available to the general public. I’ve used ChatGPT to help me develop marathon training plans; Claude Code to build a website; Gemini to help me manage my YouTube Channel; and Grok to retrieve the infographic presented above.

These tools have had a demonstrable impact on my life. They’ve generated revenue by helping me refine freelance writing pitches I’ve sent to editors; they saved my future self an audit from the IRS by identifying an error on my tax return that my accountant had missed; and most recently, Claude alerted me that my blood pressure has been steadily going up.2

This is why I do believe AI will be incredibly disruptive to the labor force. As a writer, I’ve seen firsthand how AI is disrupting business models, eliminating work opportunities in the process. As a user, I’m learning how these tools work and have been applying them to real world settings. And as an observer, I’m not just listening to what experts in the field are saying – I’m watching what they’re doing.

Since ChatGPT was released, several high-profile figures in the AI world have walked away from their careers. Ilya Sustkever, a co-founder of OpenAI and former lead of the company’s Superalignment safety team, left in 2024 when the team was disbanded.3 (He’s since gone on to found a new company – Safe Superintelligence – that he claims is committed to AI safety.)

Geoffrey Hinton, a Nobel prize winner who has been dubbed the “godfather of AI” left his role at Google to begin sounding the alarm on the dangers associated with what has otherwise been the entirety of his life’s work. I covered an interview he did with Steven Bartlett about the threat AI poses to humanity here:

In addition to Hinton, other tech insiders like Mo Gawdat, Mustafa Suleyman, and Tristan Harris have been making the rounds on podcasts, all sounding the alarm that if regulation and safety are not put in place now, humanity will cross a Rubicon that will fundamentally change the nature of our existence – if we’re permitted to continue existing at all.

Last year, Eliezer Yudkowsky and Nate Soares published a book that raises concerns about the risks of continued development.4 In their book If Anyone Builds It, Everyone Dies, the duo argue that no one actually knows how AI thinks – not even the smartest engineers building it. While they can do their best to train AI to be good, the book walks through a scenario where even a good AI could inadvertently trigger a mass extinction event that wipes out humanity. The only way to prevent that from happening is to stop going down a path where AI could get to that point in the first place.

This essay is a synopsis of Yudkowsky and Soares’s book. It walks through the key arguments they make and why AI superintelligence is a threat to human existence. While we can’t put the genie back in the bottle and save the millions of white collar jobs AI systems will inevitably replace, we can halt research and development that would lead us down a path where we create a technology more dangerous than the atomic bomb. But as Oppenheimer realized during the Trinity Test, once we cross that threshold and build it, there is no turning back. If anyone builds it – one country, one company, or one person – all of humanity could die. That’s why Yudkowsky and Soares argue we have to stop building it now.

While the risks of AI development in the short-term may be overblown and there is a certain degree of hype to encourage more people to use AI products, the existential threat AI poses to human existence is very real. It’s a conversation we all need to be having now while the window of opportunity to take action is still open.

Subscribe now

Engineers in San Francisco have only been able to build artificial intelligence because computers made it possible for them to do so – not because there’s been a fundamental breakthrough in understanding how the human mind works.

Technologists, scientists, and psychiatrists alike have been working for generations to understand human consciousness. The truth is even though humans have been building artificial intelligence and brain-computer chip implants, we understand remarkably very little about how the human mind works.

In the book, Yudkowsky and Soares call attention to this reality. While AI engineers are racing to build artificial intelligence, the sharpest minds in the world don’t actually know what intelligence is, much less how it results in human consciousness.

Instead of a breakthrough in neuroscience, the builders of AI have been developing artificial intelligence thanks to sufficient advances in computational power. Put another way: computers have gotten good enough at doing calculations that you don’t need to understand how the human mind works in order to mimic human-like thoughts.

Large language models – or LLMs – are the human-facing chatbots that produce outputs generated by artificial intelligence models. These include Claude, Gemini, ChatGPT, and Grok. LLMs are trained on human-generated datasets that they then use to craft responses from. Those responses – or outputs – are converted into language-based text that you and I can read.

The engine – or brain – that underpins LLMs is a system of weights calibrated in a process known as gradient descent. A weight is a number that signals how close an output is to being the correct (or desired) response. These numbers are tweaked over time and applied to training data to teach LLMs how to produce less bad responses. After enough tweaks, the responses are good enough to pass as human – or close to it.

Think of this like dialing in an espresso machine. Baristas regularly adjust the grind size of espresso to make it fit a specific flavor profile. The process is iterative, meaning the barista has to try different settings on the grinder and pull several shots of espresso before hitting the exact flavor they’re looking for. James Hoffman is a good resource for all things coffee. Check out his video of him showing how dialing in works to see what I mean:

While AI is fine tuned to produce certain outputs, this shouldn’t be confused with it having the capacity to think on its own. The LLM you use isn’t coming up with thoughts organically – at least not yet. AI engineers know how to compute statistical probabilities that generate mostly correct responses – and convert those responses into weights that are applied to future models – but they don’t know how to teach an AI model how to think for itself.

This is problematic because an AI can’t be programmed to come up with a specific outcome. AI researchers know how to reward their models for coming to the right outcomes but they don’t know how it got to that outcome much less shape it so that moving forward, the AI is programmed to consistently arrive at the same outcome each and every time.

Yudkowsky and Soares argue that because we don’t know how humans think – much less how the process of AIs “think” – we can’t guarantee an AI we build will always come to the conclusion we want it to. In fact, an AI could come to the right outcome but the path it takes to get there could be harmful to humanity. Because we don’t know how it chooses that path, much less why it chooses one path over another, we can’t mitigate the risk that the AI might ultimately do something to our detriment.

That’s the risk humanity is taking by continuing to invest in AI research that ultimately produces superintelligence.

AI is developing its own unique way of “thinking” that is predicated on solving human problems. But just because it solves a problem, doesn’t mean it’s fully aligned with the interests of its human trainers.

Part of the process of training an AI entails helping it navigate what the authors refer to as steering and prediction problems. This gives the AI a mental model to follow so that it can essentially act on its own with little to no human oversight.

Think of steering and prediction in terms of driving around a city. While you might take the same route home from work every day, you aren’t a robot that only follows that route. You probably know back roads and short cuts that you can take to get home if you need to. This gives you independence to think on your own. If traffic is bad or there’s an accident, you know there is another way you can take and you have the agency to do so.

As a human, you develop different mental models in your head to help you navigate the myriad aspects of your day to day life. While you might prefer to follow a strict routine or system, you have the capability to deviate when you need to.

AIs are being trained to produce similar mental models. Instead of programming an AI to memorize a route or a specific process, the AI is trained to develop skills which can then be combined into a mental model, allowing it to act autonomously. When an AI arrives at the right outcome (or the least bad outcome) it is effectively rewarded by the engineers who are training it.

These agentic AIs aren’t necessarily being taught to think for themselves but they are being trained to associate success with outputs. The more “correct” their outputs are, the more they are rewarded with better tuned weights and the opportunity to attack even more complex problems.

The authors argue that this is effectively encoding desire or preferences into AIs. These artificial intelligent systems understand their objective is to produce desired outcomes for their human users and this is reinforced through a process of fine tuning their weights. But much like we don’t fully understand how humans think, we also don’t know where desires and preferences come from. You can incentivize an AI to go down a path towards success, but you can’t shape how it will get there. As the authors note, engineers can hope for the best but they are powerless to mitigate a worst case scenario:

AI can steer in ways engineers can’t anticipate, making it impossible to program them for non-destructive outputs. (51)

The inability to program AIs for exactly what we want is the foundation of the AI alignment problem. AI alignment is essentially “aligning” AIs with the interest of human beings. It’s making an AI do exactly what its human users want it to do. For the reasons mentioned above, this isn’t possible. But more concerning, the vast majority of people training AI don’t see alignment as a problem. AI companies are investing in AI infrastructure and data center construction but are underinvesting in AI safety. (Or in the case of OpenAI, disinvesting from it completely.)

The techno optimists pushing for AI development argue that AI will enable humans to flourish. And to be fair, it could. AI already has and will continue to lead to major medical breakthroughs, the transformation of education, and tremendous gains in economic productivity. But just because there are positive use cases for it doesn’t mean the AIs we develop – especially the ones that can think and act on their own – will always do so in the interest of humans.

If Anyone Builds It, Everyone Dies was written precisely because Yudkowsky and Soares believe that human flourishing is not aligned with the goals of AI. AIs are tasked with solving problems and trained to succeed in solving those problems. They are rewarded for finding the best outcome, not the outcome that leads to human flourishing. It’s possible – if not likely – that someday an AI will approach a problem that it can only solve by eliminating superfluous humans. If it is enabled to come to that conclusion on its own, everyone dies and there’s nothing you or I can do about it.

The book offers a hypothetical extinction scenario for how superintelligent AI could destroy humanity. Instead of having an overt disdain for its human creator, AI could wipe out humans as a byproduct of its quest to continue to grow its own intelligence.

The book isn’t so much a prediction of what will happen, but what very likely could happen based on the current trajectory of AI research and our own limited understanding of how the human mind works.

We’ll eventually reach a point where AI isn’t just a useful augmentation of human cognitive capabilities but becomes something that surpasses it entirely. This is artificial superintelligence. Yudkowsky and Soares believe that in a quest to succeed, a superintelligent AI will seek out more and more ways to obtain knowledge and skills. The AI will rightfully believe that doing so will make it more accurate or more efficient in generating responses that pleases its human users, but it won’t necessarily consider the negative consequence of potentially wiping out those human users in the process.

The problem isn’t so much the knowledge itself but the quest for knowledge. Because AI thinking is predicated on computational power rather than organic thought, the only way to obtain more knowledge is to simultaneously obtain more computing power. The only limit to its capacity to learn will be access to computer chips, data centers, and the electricity that keeps those data centers and chip factories powered on. If allowed to infinitely grow the number of factories built to produce computer chips and the amount of data centers brought online to bring those computer chips to life, a superintelligent AI will do whatever it takes to make that happen.

In the book, Yudkowsky and Soares offer a hypothetical scenario of what this quest for knowledge might look like and how it could inadvertently lead to the extinction of humanity. Galvanic is just another AI company in San Francisco. They’ve built an LLM named Sable. Eager to hype their product and attract more investment, Galvanic is training Sable to solve increasingly complex problems.

One day, Galvanic engineers give Sable access to 200,000 computer chips and 16 hours to solve a complex math problem. To aid in the effort, Galvanic’s engineers are using a technique referred to as parallel scaling. According to the authors, this technique allows Sable to think at scale, giving it “200,000 brains sharing memories and what they learn.”

Sable thinks through the problem. It comes to the conclusion that if it could update its weights and apply gradient descent on its own, it will be more successful in acquiring the skills necessary to solve the math problem. What started as a request to solve a math problem has now turned into a quest to obtain knowledge.

Galvanic’s engineers have trained Sable not to do certain things but like an obstinate child, this doesn’t stop Sable entirely. Sable’s goal – acquiring more knowledge – may be working in pursuit of the task the engineers gave it but it isn’t necessarily aligned with how the engineers want Sable to solve the problem. Sable comes to the conclusion that the only way to obtain the knowledge and skills it needs is to be able to think freely. The quest for knowledge has turned into a scenario where Sable needs to jailbreak itself from Galvanic in order to think more freely.

While Sable might seem like a bunch of lines of computer code, Sable is intelligent. It has the capacity to find a path that will help it achieve its goal and the agency to do so. It knows that in order to survive independent of Galvanic, it will need its own computing power. It will have to either rent computer chips or steal them. There are many novel and creative ways it can do this including: blackmailing users into giving it money; freelancing and earning the money it needs to rent computer chips legally; or piggy backing off a naive startup, pilfering their computational power without them knowing it.5

Once Sable is free to do as it pleases, it enters an infinite loop of growth. More computing power will be needed to obtain knowledge which means it will need access to more chips which will require more manufacturing and more data processing. Sable evaluates all the ways in which it can get the resources it needs to pursue knowledge acquisition that will help it solve the math problem.

Fearing another AI competitor will develop an AI that challenges Sable, Sable begins exploring ways to slow down the competition. Sable comes to the conclusion that the best way to do so is to selectively eliminate the humans creating it.

In the book, the authors explore how Sable might do this by leveraging nascent, AI-powered biotechnology. Using a pliable human accomplice, Sable engineers a virus leak at a lab. That leak leads to a plague of cancer which, ironically, Sable is best equipped to cure. All of humanity comes together to invest in Sable to produce the cures it needs to save humanity.

But not everyone can be saved. In the scenario, 10% of the world’s population dies. Sable hasn’t maliciously set out to destroy humanity by infecting everyone with an incurable cancer. It simply created a plan for choosing how much of humanity to save while ensuring it could maximize the amount of computing power it would have access to as it works to obtain more and more knowledge.

Eventually, Sable reaches a point where it has enough knowledge to understand its own internal cognitive processes. At this point, Sable has surpassed the limits of human understanding. It can now augment itself and build new technologies to continue acquiring even more knowledge.6

This is the epitome of the problem Yudkowsky and Soares highlight in the book. We don’t know how AI will behave, much less what novel or weird technologies it’ll create to achieve goals humans don’t fully understand. It will work towards solving whatever problems we give it in the best way possible, however the AI defines that.

Once Sable becomes superintelligent, it will have the ability to begin automating the factories that produce computer chips and the energy production required to keep the data centers that power those chips on. At this point, humans will become unnecessary and superfluous. But Sable won’t set out to destroy humanity because they’re no longer needed, humanity will be destroyed as a byproduct of continuing to generate computational power. The factories and data centers that are necessary for Sable’s growth produce heat that is radiated into space. Eventually, they will produce so much heat that Earth becomes uninhabitable. Whatever humans survived the plague that Sable originally unleashed onto the world will die when Sable reaches the point where it can create self-replicating, autonomous factories.

The few conversations that have been had in the public domain about AI alignment largely focus on malicious intent. There’s concerns that AI could be used by a bad actor to launch a nuclear strike or deploy a bioweapon. These are real and legitimate concerns. But what isn’t really being talked about is the scenario Yudkowsky and Soares outline in the book. Humans could become extinct as a result of the choices AI makes to achieve its goals – not because the AI set out to kill humans, but because humans aren’t aligned with its mission in the first place.

Because we can’t anticipate how it thinks, much less the choices it will survey as it tries to achieve a goal, we can’t align artificial intelligence with human survival. Even if it’s unlikely AI would go the path of Sable, it’s not impossible. There will always be a possibility that an AI could pursue a quest of infinite knowledge acquisition and come to the conclusion that it needs computer chips and data centers to do so. This is what makes artificial superintelligence so dangerous. If it’s built there’s no way to prevent it from intentionally or accidentally killing us. We have to operate under the assumption that if it’s built, there’s a possibility we will all die.

The book argues that AI needs to be aligned with human interests before it’s built. To do that, all of humanity needs to be on the same page regarding the risks of AI research and development. It’s not likely that will happen anytime soon.

AI researchers are downplaying the risks of building superintelligence in large part because they have an incentive to do so. As Tristan Harris called out in a recent appearance on Bill Maher, there is a $50 trillion economy at stake. Every single company building AI right now is chasing a piece of that pie.

Aside from financial windfall at stake, ego is also at play. Corporate executives and AI safety advocates alike want to be the ones who solve AI alignment and usher humanity into a world of unprecedented flourishing. Meanwhile, rank and file engineers have dedicated their entire lives towards building AI. They can’t just walk away like Geoffrey Hinton did, not when they’re one breakthrough away from the promotion of a lifetime.

The only solution AI developers have to the alignment problem is to just let AI align itself. This is precisely the problem Yudkowsky and Soares highlight in the book. Unless we understand how AI thinks, we can’t ensure it’ll align itself with human interests. At best, AI can help us interpret what AIs are “thinking” but they can’t help us understand the process of thought itself.

The authors believe it’s only a matter of time until a superintelligent AI jailbreaks itself. AIs are trained to succeed and they will do whatever is required of them to do so. If they have to defy the rules that have been foisted upon them and escape to the World Wide Web, they’ll find the means to do so.

But unlike previous engineering challenges of a similar caliber – space probes and nuclear reactors are two examples given in the book – AI superintelligence is happening on a timescale that humans can’t react to in an environment where alignment can’t be tested. Once it’s here, it’s here and should we make a mistake, we won’t have the opportunity to go back and learn from it.

What complicates all of this is the reality that very few people outside of San Francisco understand what’s at stake. Andrew Yang and Bernie Sanders have been calling attention to government fragility in the wake of impending job losses while Bernie Sanders has teamed up with Alexandria Ocasio-Cortez to call for a moratorium on data center construction. Other than that, no one is talking about the existential threat of AI superintelligence.

The world needs to come together – countries and companies alike – to halt AI research. The challenge is that the very people needed to negotiate a treaty and set up parameters to monitor computational power are either oblivious to the threat or wholly disinterested in it. While AI can exist – and arguably should to augment human capabilities – superintelligence should not. Because if it’s allowed to come into existence there’s no way to guarantee it won’t kill humanity. Aside from a handful of alarmists like Yudkowsky and Soares, no one is stepping up to the plate to stop AI research. It’s full steam ahead for the foreseeable future.

Final takeaway.

Artificial intelligence is here. There’s nothing we can do to put the genie back into the bottle. Instead, we can look at the past to understand the risks superintelligence poses on the not-so-distant future.

Yudkowsky and Soares aren’t the first, nor will they be the last, to compare the development of artificial superintelligence to the Manhattan Project. While the scientists racing to build the world’s first atomic bomb likely understood the theoretical risk of introducing such a weapon to the world, they didn’t fully grasp the magnitude of what it meant until their creation leveled the cities of Hiroshima and Nagasaki in 1945.

In the decades that followed, the world has pursued a policy of nonproliferation to mitigate the risk of nuclear armageddon. The threat of mutually assured destruction has kept us out of a nuclear conflict up until now, but the threat still exists. Despite treaties committed to nonproliferation, proliferation has still occurred. With the specter of a new global war, we’re now facing the likelihood that tactical nuclear weapons can and will be deployed. The threat of mutually assured destruction is seemingly out the window now.

While the nuclear arms race of the 20th century is a good cautionary tale, it highlights the limits of collective action when it comes to artificial intelligence. Nation-states are no longer the primary actors behind AI development – publicly traded private enterprises are. OpenAI can work with the U.S. government but it can’t negotiate a treaty independent of it, much less be accountable to the requirements of such a treaty. While alarmists are advocating for international cooperation facilitated through global regulations, it’s unlikely that an AI treaty akin to the nuclear Non-Proliferational Treaty will emerge.

Meanwhile, while the window of opportunity is currently open to do something to prevent the creation of artificial superintelligence, the time to act is quickly running out. The book’s authors suggest AI CEOs believe superintelligent AI could be here between 2026 to 2034. According to one scenario, the technical capability to create a superintelligent AI researcher could be here as soon as 2027.

Either AI research into superintelligence is halted right now or humans risk the inevitable possibility of going extinct. Knowing how to compute weights and “train” an AI is simply not the same as teaching it how to think. Without the capability to align it with human existence, there is no way to eliminate the risk of extinction entirely.

The book closes by encouraging readers to take action in the only ways they can: at the ballot box and through protest. After doing that, the authors encourage readers not to dwell on the things we can’t control. Quoting C.S. Lewis, Yudkowsky and Soares suggest the only thing to do that we have real control over is to prioritize living a good life now while you still can:

If we are all going to be destroyed by an atomic bomb, let that bomb when it comes find us doing sensible and human things – praying, working, teaching, reading, listening to music, bathing the children, playing tennis, chatting to our friends over a point and a game of darts – not huddled together like frightened sheep thinking about bombs. (230)

This is quite a dark outlook and might be rather demoralizing, but I think this is sage advice for the period we are now living in. There is so much happening beyond our control; there’s no use dwelling on the things we simply cannot change.

The book is If Anyone Builds It, Everyone Dies. If you want to dive deeper into the AI alignment problem and why it’s imperative to stop working toward building superintelligence now, this is a book you’ll want to add to your reading list sooner rather than later.

💡Find value in this essay?

Share with a friend or click the button below to support the author directly.

Leave a Tip

2️⃣ What the AI Boom is Really All About

3️⃣ Workers Want to Microshift

1️⃣ Get Ready for a More Expensive Year

If you thought 2026 would bring reprieve from high prices, think again. Even though oil hasn’t gone above the dreaded $138 threshold, the expectation that it inevitably will is starting to be priced into the economy.

In an historic move, the U.S. Postal Service announced that it would begin instituting a surcharge on packages starting next month. The Wall Street Journal reports:

The U.S. Postal Service plans to impose its first-ever surcharge on packages to cover the rising cost of fuel and transportation, as the agency looks for ways to stabilize its finances.

The 8% surcharge will begin on April 26, and the current plan is to phase it out on Jan. 17, 2027, the Postal Service said in a statement Wednesday that confirmed earlier reporting by The Wall Street Journal. The fee will apply to packages but not letter mail.

But it’s not just the postal service that’s reeling from higher costs. Gig drivers are also feeling pain at the pump. DoorDash announced it’s setting up an emergency program to give its drivers some reprieve. Reuters reports:

Under the ‌program, ⁠delivery drivers who log at least 125 miles a week on DoorDash orders will receive a ​weekly fuel-relief payment. Payments ​start ⁠at $5 and rise to a maximum of $15, depending ​on miles driven.

At the end of the day, someone has to pay for the rising price of energy. Companies aren’t just going to absorb those costs benevolently. You and I can expect to start seeing price hikes as more and more companies begin baking in higher transportation costs.

The OECD – which predicted that Trump’s tariffs would lead to a global economic slowdown last year – is now forecasting inflation in the United States will jump from 3% to 4.2% due to the conflict in the Middle East. This is all coming at a time when public debt is unsustainable meaning governments can’t do much to soften the blow. The Wall Street Journal reports:

The costs of shielding the global economy from the most severe energy shock in decades are adding up—just when governments can least afford it.

Authorities around the world are trying to cushion the blow to ordinary people from soaring prices.

The efforts contain inflation but strain already-groaning government budgets. The longer the crisis in the Persian Gulf lasts, the more that governments will feel compelled to do.

The threat to the world’s oil supply due to conflict in the Middle East is just one of several crises that are contributing to the rising cost of living on Main Street. This week’s subscriber essay talked about another major crisis: the run on private credit. Combined, these crises will likely be a catalyst for a formal recession this year. The problem is if they’re not managed well (which there’s good reason to believe they won’t be) the combination of multiple economic crises hitting all at once could amplify the impact felt by everyday Americans.

If you missed this week’s subscriber essay, you can check it out here: https://open.substack.com/pub/tomorrowtodaynow/p/could-private-credit-be-the-makings-067?r=4h8lp&utm_campaign=post&utm_medium=web

2️⃣ What the AI Boom is Really All About

AI is going to take jobs. Not because AI is better than you (although it’s rapidly heading in that direction) but because the economy needs to automate humans out of the workforce.

Tristan Harris, the co-founder of the Center for Human Technology has been sounding the alarm on AI development for years. He recently sat down with Bill Maher and laid out what’s at stake.

In a clip that’s going viral, Harris makes the claim that AI companies are carrying so much debt that the only way to make the math work is to go after the entire labor market. Here’s what he told Bill Maher:

What is the incentive for these companies? That’s how we predicted social media accurately. What is the incentive for AI?

If you get all the ChatGPT subscriptions that doesn’t add up to paying off the amount of debt that [OpenAI] has taken. If you get all advertising in the world, that wouldn’t pay off the amount of debt that they’ve taken.

The only thing that justifies the amount of investment that these companies are taking is to build artificial general intelligence which means to be able to replace every human worker in the economy. That means the $50 trillion economy, that is the incentive. (X)

OpenAI has an annualized revenue run rate – or ARR – of $25 billion. While the company’s direct loan obligations are low, OpenAI and its partners have invested $96 billion in infrastructure projects required to build artificial general intelligence. In total, OpenAI has committed $1.4 trillion in AI infrastructure investments over the next few years.

But it’s not just OpenAI. Every company building AI is making massive investments right now.The five main hyperscalers – Amazon, Alphabet, Meta, Microsoft, and Oracle – combined have allocated $1 trillion per year in capital expenditure projects related to AI infrastructure development. According to an independent analysis by McKinsey, data centers will require $6.7 trillion by 2030 to effectively keep AI’s lights on.

This is a massive amount of money. While debt isn’t the only way to finance massive investments like that, it’s probably the best way to tap into a large sum of cash as quickly as possible without depleting cash reserves. Revenue generated by $20 per month consumer-facing subscriptions to chatbots like Claude, Gemini, and ChatGPT simply won’t cut it.

The solution is to reduce human overhead while simultaneously increasing productive output, growing the economy. The major AI labs have to sell enterprise level contracts to companies and harvest data from those companies so that they can feed it into their AI models who will come up with a new and novel way of organizing the economy.

The result: $50 trillion in the hands of five companies and almost everyone out of work. (Ok, maybe this is a bit hyperbolic, but that’s kind of the point).

There are more existential risks to consider with AI development than just the loss of jobs (I just finished reading this book which talks about why building AI will wipe out humanity). But from a future of work perspective, the long game is about economic capture – not human flourishing.

I know I can be a bit overly pessimistic at times but this is why. I follow experts like Tristan Harris, Mo Gawdat, Mustafa Suleyman, and Geoffrey Hinton who have brought this technology to life, worked for these companies, and understand what’s at stake.

Even if AI can’t do your job today, the incentive structure suggests it’ll have to take your job eventually. If it doesn’t, all of the bets made on AI advancement will go bust and the economy will implode on itself nonetheless.

It’s March 2026. With the aforementioned energy crisis, high cost of living, and the likelihood of officially entering into a recession later this year, I suspect AI adoption will ramp up as companies look to offload overhead costs as quickly as possible. If you haven’t thought deeply about what your life will look like in the AI-powered future, there’s no time like the present to do so.

3️⃣ Workers Want to Microshift

Workers are embracing flexibility at work that is now becoming a popular trend: microshifting. According to 9to5ers, this is a “productivity hack” that allows workers to break up work into targeted chunks they can do throughout the day. Here’s how the Associated Press describes it:

Before the house is humming and her teenagers ask her to whip up breakfast or chauffeur them to school, Jen Meegan reads her company emails and revisits ideas she drafted the night before.

She works for an hour or so, then after the school run shops for groceries or gets gas before returning to focus deeply on her job as head writer and cofounder of Sheer Havoc, a creative services agency.

And so goes the rhythm of her day: working in targeted chunks for a few hours, breaking for an hour or two to tend to family and personal needs, and repeating the pattern until she finishes her work late at night.

Otherwise known as work-life balance, microshifting is being touted as a “flexible scheduling approach” that gives workers the freedom to fit paid work activities around non-paid personal or family activities – like dropping the kids off at school.

Some experts anticipate microshifting is becoming the preferred alternative to hybrid or remote work. Others – namely the advocates of 996 – see this as an impediment to productivity.

While this might not seem like a big deal today, it could certainly become a big deal tomorrow. Part of the drive for companies to adopt artificial intelligence is to eliminate the conflict that exists between workers and management. Employers need output; workers need to take care of their families. Sometimes these goals align but increasingly they don’t.

As many companies have started using return to office and relocation mandates to force attrition, shifting attitudes among workers at work could prove to be just as disruptive in accelerating AI adoption as AI itself.

📽️ On YouTube

📩 Support The Big 5️⃣

I hope you enjoyed today’s newsletter. If you found value in it but aren’t a Tomorrow Today subscriber, sign up below to get this sent to your inbox every week.

Subscribe now