--- title: "The American Economy Isn't Ready for AI" description: "\"CEOs Start Saying the Quiet Part Out Loud: AI Will Wipe Out Jobs.\" That was the Wall Street Journal headline, in July of 2025, atop an article that future generations may remember as a turning point. In it, Ford Motor Company's chief executive warned that AI would replace \"literally half of all white collar workers\" at the manufacturer. And he is far from alone in thinking that way.\n\nFast forward several months, and that displacement is no longer theoretical. It is showing up in employment data, particularly in the United States, as hiring slows to a crawl and unemployment begins to tick up for the first time in years. What makes this so different from previous downturns is that these companies are not shedding employees because times are tough. For some of them, times have never been better.\n\nAI is going to improve living standards for billions of people, and it carries enormous potential to be one of the most profound forces for positive change in human history. But that promise holds only if we get the transition right, because the consequences of mishandling it would be equally large in the other direction — possibly fatal.\n\nThe defining risk of this moment is not whether AI eventually makes the world richer, but whether the American economy can survive the disorderly, under-planned road between here and there.\n\n## Key Takeaways\n\n- US hiring has collapsed to near zero, with Federal Reserve Chairman Jerome Powell acknowledging \"job creation is pretty close to zero\" and citing AI as a primary driver of layoffs and freezes.\n- Companies announced 153,074 job cuts in October 2025 alone — the highest for that month in over two decades — even as firms like Microsoft and Amazon posted record profits.\n- AI differs from past innovations because it replaces cognitive work without generating comparable new categories of employment at scale, severing the centuries-old link between corporate growth and hiring.\n- Consumer spending is 68% of US GDP, and signs of household distress are mounting: subprime auto delinquencies hit a record 6.65%, credit card debt reached $1.23 trillion, and over 9 million borrowers are delinquent or in default on student loans.\n- Entry-level work is being hit hardest, with recent-graduate hiring at major firms down 35% versus 2023 and some tech-hub postings down as much as 50% from pre-pandemic levels.\n- Policy discussion is nearly absent, dominated by national security framing, while young people make life-altering financial decisions based on a labor market that is disappearing.\n\n## The Displacement That's Already Begun\n\nArtificial intelligence crossed a threshold somewhere in late 2022 that very few people saw coming. When ChatGPT debuted that November, it took the world by storm. It was not very good at much of anything, but it technically worked. Competitors quickly followed in its wake: Anthropic released Claude, then Google released Gemini, both of which made memorable mistakes — including the moment Gemini advised people to eat a small rock each day for its supposed nutritional benefit.\n\nThe advancement since then has been remarkable. We are barely three years into the life of these systems, and what began as a curiosity — something to play with, and perhaps to worry about in some distant future — has become something else entirely. Meta is now signing deals to draw power from a nuclear plant, entire technology ecosystems are being built and operated around these models, and OpenAI's Sora can produce short videos nearly indistinguishable from real footage.\n\nWarnings about AI transforming work have circulated for years. What has shifted recently is the perception of that displacement: it has moved from a looming, far-out threat to something front and center, and that trend looks unlikely to reverse.\n\nThe labor data is beginning to reflect that shift. The US job market has been weakening for some time. There is genuine dispute about exactly when the slide began, but the consensus holds that it took a real hit in 2024. The economy was initially thought to have averaged a gain of 186,000 jobs per month that year — not terrible, but a noticeable drop from 2023's healthier 251,000 per month. That assessment was overturned when the Bureau of Labor Statistics revealed that the prior year had actually added 911,000 fewer jobs than previously believed. The revision, one of the largest in recent history, meant that what had looked like modest growth was actually closer to stagnation.\n\nThe end of 2024 is when the bottom really began to give way. The economy had entered what analysts were calling a \"no hire, no fire\" dynamic, in which companies were not eager to take on new personnel but were not yet confident enough to begin layoffs at any meaningful scale. This unusual stability reflected both a declining need to replace departing workers and a kind of corporate post-traumatic stress following the great labor shortage of 2021, when firms were bending over backward to attract applicants for roles they desperately needed to fill.\n\nThat dynamic is now changing, and the labor market has begun to soften further, especially in the United States. Job creation in recent months has essentially collapsed. Even Jerome Powell, the chairman of the Federal Reserve, conceded that \"job creation is pretty close to zero,\" noting that \"a significant number of companies\" have begun layoffs and hiring freezes and citing AI as the primary driver.\n\nHis acknowledgment raises the question of why it took so long to recognize what had been increasingly clear for months. The October data simply confirmed what had been unfolding in plain sight. Challenger, Gray and Christmas released a report during the US government shutdown showing that companies announced 153,074 job cuts in October alone — the highest for that month in over two decades, and nearly triple the figure from the previous month. Warehousing and the technology sector led the carnage, with over 47,000 and 33,000 cuts respectively, several times higher than their September levels, as firms restructured around AI integration.\n\n## When Record Profits Come With Layoffs\n\nThe tech sector's cuts have been particularly aggressive. Microsoft led the way with 6,000 cuts in May, another 9,000 in July, and continued trimming through September. Software engineers bore the brunt, accounting for roughly forty percent of eliminated positions — a striking figure given that CEO Satya Nadella has acknowledged AI now writes up to thirty percent of Microsoft's code. Amazon slashed 14,000 corporate positions, with its CEO explicitly connecting the cuts to AI adoption and a vision of a company with \"fewer people doing some of the jobs that are being done today.\" Meta opened the year by cutting 3,600 employees, while Salesforce's customer service division shrank from roughly 9,000 to 5,000 employees over eighteen months.\n\nAll of this would read differently if these companies were struggling financially — if they were trimming payroll to survive a downturn or stave off bankruptcy. But that is not what is happening. Microsoft reported $27 billion in profit in the very quarter it announced the 9,000-person layoff wave, one of the best quarters in company history, and its stock performance reflected it. The pattern repeats across the others. Amazon has reported profit increases of 39%, reaching $21.2 billion, and Meta's cuts came amid robust revenue growth.\n\nConcerns about automation are, of course, nothing new. People complaining about technology upending how they live and earn is as old as industry itself. The Luddites are the most infamous example — workers who smashed the machinery they believed threatened their livelihoods. The short version is that it did not work. Technology kept advancing, the Luddites were largely forgotten, and a similar fate met most who refused to adapt to a rapidly changing world.\n\nYet today's situation really does appear different — less because of what is changing than because of who is sounding the alarm. In the Luddites' era, it was workers and skeptics raising concerns while experts dismissed them. This time, it is the leaders of some of the country's largest firms warning of the disruption, while the ordinary person on the street is far more likely to be going about the day, unaware of what is coming. That reversal of roles should give us pause. When the people building and deploying a technology are the ones raising concerns about its impact, it is worth listening. When Ford's Jim Farley told the Wall Street Journal that AI is on track to displace \"literally half of all white collar workers\" in the country, he was not lying — he was warning.\n\nOther prominent voices in technology have offered similar predictions. Elon Musk has proclaimed that AI will make work optional and replace \"all jobs,\" while Bill Gates believes that within ten years we will inhabit a world in which machines perform most of the economically valuable work. For those who doubt that this is coming, there is no need to take it on faith — but doubting the people who design these platforms is its own kind of risk.\n\n## Why This Wave Is Different\n\nThere is something else that separates this moment from earlier technological upheavals. All innovation has cost some jobs, so at one level this is not radically new. The steam engine displaced horse-drawn carriages, but it created jobs manufacturing engines, producing and laying rail lines, and operating locomotives. The arrival of computers wiped out the entire typist pool — until relatively recently a large field of workers — but it created whole industries around software development, IT support, and digital services that employed far more people than were displaced. Even if the office typist never went on to write code, the growth of those new industries created supporting jobs. Every major technological shift in modern history followed this pattern: initial displacement, followed by the creation of new categories of work that absorbed the displaced. There was lingering fallout, but there was also a variety of new fields to move into.\n\nArtificial intelligence is fundamentally different and should not be lumped in with those earlier waves, because it operates from a different premise. It replicates and can replace human cognitive work without generating comparable new employment at scale. When previous technologies automated physical tasks, they still required human minds to supervise, operate, maintain, and, most critically, design them and their successors. AI has the ability to automate all of these functions — if not today, and perhaps not tomorrow, then in the not-so-distant future.\n\nCompanies are not waiting for some future breakthrough to begin restructuring their workforces around AI. The changes are happening now, especially at the entry points where workers traditionally launched their careers. It has been a running joke for years that entry-level positions somehow require three to five years of prior experience, but matters have recently escalated to a new level. Recent-graduate hiring at major firms has fallen 35 percent compared with 2023, while job postings in certain tech hubs have dropped as much as 50 percent from pre-pandemic levels. This is less a story of the ladder getting harder to climb than of the bottom rungs being sawed off entirely.\n\nThe human impact is beginning to surface in stories that hint at what may be coming. Allen Rausch, a 56-year-old who spent decades working for companies like Amazon and Electronic Arts, lost his job earlier in 2025. While applying for new positions, he was stunned to find AI conducting his interviews — and not just one of them. Each lasted around 20 minutes and proved totally unable to answer any of his questions, simply moving on to the next prompt. He was being assessed by a system so unsophisticated it could not respond to even the most basic query. \"Given the percentage of responses that I'm getting to my applications, I think a lot of AI interviews are wasting my time,\" he told reporters.\n\nWhat makes Rausch's experience particularly ominous is that most companies remain in the early phases of AI adoption — though that is changing. And those that have moved early are often not shy about it. Klarna, the Swedish fintech company, openly boasted that its AI assistant could handle the workload of 700 customer service agents, managing millions of conversations a month across 35 languages.\n\nThat rollout did not go as planned. Like the system that interviewed Rausch, Klarna's tools simply were not sophisticated enough for the job, and the company faced an intense backlash as customers found they could not reach a real person when they needed help. By mid-2025, Klarna was forced to admit it had gone too far, too fast with automation, and that its focus on cost-cutting had degraded service quality. The company has since walked back its AI-exclusive approach.\n\nWhile that episode serves as a warning to other firms about the risks of going overboard, it is not exactly comforting. The lesson Klarna drew was that AI is not currently able to fully replace all people. The implied \"yet\" practically writes itself. These implications extend well beyond individual job losses. What is breaking down is a fundamental economic relationship: the one linking corporate expansion to job creation. For two centuries, companies that grew needed more workers — that was simply a cost of doing business. That equation is now being rewritten.\n\n## The Corporate Calculus\n\nBefore getting too far ahead, a few caveats are worth stating plainly. We are still early in this transition, and exactly how it will play out is impossible to predict, especially given uneven distribution across countries, labor laws, and industries. It is also worth emphasizing something often lost in debates about displacement: companies are not charities. They will act in their own interest to maximize profits for themselves and their shareholders. This is not said cruelly, but to underscore the framework on which the system runs.\n\nFor the last two centuries, businesses expanded relentlessly — first across cities, then states and provinces, and ultimately across borders — and that expansion proved remarkably effective at reducing poverty worldwide. In 1820, roughly 80% of the global population lived in what we would recognize as extreme poverty. By 2024, that figure had fallen to around 10%. The problem is far from solved, but the overall living condition on earth has improved dramatically over that span. Between 1990 and 2015 alone, the number of people in extreme poverty fell from just under 2 billion to 702 million — a staggering improvement that deserves far more attention than it receives.\n\nMuch of that progress depended on a simple equation: as companies grew and reached new markets, they needed more workers to produce, distribute, and sell their products. Factory expansions meant more assembly-line workers; new stores required cashiers, stockers, and managers; international expansion meant entirely new supply chains staffed by people across the globe to move a product from point A to point B.\n\nThat connection is increasingly being severed. In 2023 and 2024, executives spoke cautiously about the new technology, emphasizing \"human-AI collaboration\" if they said anything at all. By 2025, the message shifted, and companies began speaking openly about expanding without adding employees. A Wall Street Journal banner on October 26th captured it: \"More Big Companies Bet They Can Still Grow Without Hiring.\" The report was a litany of the same dynamic. JPMorgan's chief financial officer told investors that the bank now has a \"very strong bias against having the reflective response\" to hire more people for any given need.\n\nThey are hardly alone. The article reads like a roll call of American business, from Walmart to Airbnb to Target and beyond. Walmart explicitly told investors it plans to keep headcount essentially flat over the coming years even as sales grow, with AI handling tasks that would once have required new hires.\n\nThe financial data makes the disconnect even more revealing. Over the past three years, publicly traded US companies have cut white-collar headcount by roughly 3.5% while revenues climbed — a pattern that has recently accelerated. A significant share of S&P 500 firms now employ fewer people than they did a decade ago, even as many have posted substantial revenue growth. Put bluntly, an AI model might not be better than you at your job — but it is increasingly about as good.\n\nWall Street is rewarding the trend enthusiastically. Companies have been tripping over themselves to announce how they have \"innovated\" AI into some product — from $400 AI toothbrushes that supposedly improve your technique to a Samsung AI fridge that uses cameras to monitor groceries and warn you when the milk is about to expire. These may not quite be the revolution the marketing teams promised. Even so, executives now brag on earnings calls about headcount reductions. The striking thing is that such moves used to signal the opposite: layoffs once told investors a company feared trouble downstream, while hiring signaled optimism about future growth.\n\nAs adoption progresses, it creates a self-reinforcing dynamic. Companies that adapt are seen as better run, more efficient, and more profitable. If your competitors achieve 20% margin improvements by replacing human workers with AI, you face a choice — match them or risk being seen as uncompetitive. The result is essentially an arms race toward automation, with each firm trying to move faster than its rivals.\n\n## The Consumer Death Spiral\n\nThe United States, for better or worse, appears to be at the forefront of adopting this technology — and it is not entering that race in the best economic shape.\n\nCompanies operate along a purely competitive logic that makes collective restraint next to impossible. Walmart is not going to call Target and propose they both limit AI adoption to preserve jobs, because even if they agreed, they would expose themselves to some other firm outside the pact. And even if all the so-called big players signed on, they would still have to perform financially, leaving them vulnerable to a newcomer unbound by their mutual agreement. If a single company could automate fully while no one else followed, the economy would chug along with few any the wiser. It is the collective push toward automation that poses a mutually assured threat. Consumer spending represents 68% of US GDP, and sustaining that spending requires people to have a steady income. As companies race to replace workers with AI to boost individual profits, they are collectively undermining the consumer base their own business models depend on.\n\nThe American consumer is already showing serious signs of distress. The clearest signal comes from the auto loan market, a classic indicator of genuine economic struggle — losing a car is among the last things most Americans will allow, given how essential it is for getting to work and managing daily life. The delinquency rate, measured by loans more than 60 days past due, is up more than 50% over the last decade and a half. Among subprime borrowers, delinquency reached 6.65% in October, the highest ever recorded in American history. Repossession orders have jumped more than 40% since 2022, hitting the highest level since 2009.\n\nWhat makes this especially alarming is that the country has not even seen a large spike in unemployment yet. Trends in recent-graduate joblessness are genuinely concerning, but the latest data still shows overall unemployment relatively low by historical standards. Some of the people caught in these statistics are recently unemployed, but many still hold jobs.\n\nPart of the explanation lies in the historically unprecedented stimulus payments that some households grew accustomed to during the pandemic and its immediate aftermath. A combination of government support and record-low interest rates allowed people to qualify for auto loans they otherwise could not have afforded. All of this is now feeding into a moment in which reality is beginning to catch up. It does not change the fact that Americans appear poised to enter this economic transition at a particularly weak point.\n\nCredit cards tell a similar story of mounting pressure. Total credit card debt has exploded to $1.23 trillion, with the seriously delinquent rate — accounts more than 90 days past due — hitting 11%, the highest since 2012. Even the mortgage market, relatively stable in terms of defaults, is showing cracks: delinquencies on home loans have crept up to 3.99% for borrowers more than 60 days past due, the highest in years.\n\nStudent loans add another dimension to the squeeze. After being paused for over three years during the pandemic, when payments were granted a grace period that was repeatedly extended, millions of households suddenly faced hundreds of dollars in monthly payments many had set aside, imagining the day of reckoning might never arrive. They were not entirely unreasonable to think so: the Biden administration made numerous pushes to forgive student debt, including a wide-reaching plan that would have canceled up to $10,000 per borrower for an estimated 26 million applicants. The Supreme Court struck it down in June 2023. Over 9 million borrowers are now either delinquent or in outright default. And unlike most other debt in the US, student loans usually cannot be discharged in bankruptcy — borrowers must prove \"undue hardship,\" a standard rarely met, so balances often follow people for decades.\n\nWhat connects all of these data points is timing. The American consumer is maxed out, and there is little slack left in the system. The government's fiscal position makes matters more precarious still. The US federal debt now stands at over $38 trillion, the largest debt-to-GDP ratio since the Second World War — except this time it was run up during peacetime, in an expanding economy. If the consumer is entering this transition in a fragile state, the government is not faring much better.\n\nThe traditional policy response to mass unemployment — cutting interest rates to stimulate lending and thereby spur hiring — will not work this time, because the link between cheaper money and job creation has been severed. The implications reach far beyond individuals simply buying less. There is always a contagion effect from layoffs. The US has become so dependent on consumer spending that it can hardly afford not to spend; even a minor pullback could carry significant consequences, something the legendary investor Mohamed El-Erian has said keeps him up at night.\n\n## Flying Blind\n\nPerhaps the most striking aspect of this moment is the near-total absence of serious policy discussion about what is unfolding. A handful of politicians have come forward to raise alarms, but they represent a tiny minority, and when officials do speak about the issue, they almost always frame it through national security.\n\nTo be clear, given how capable this technology has already become, this genuinely is a national security concern. Anthropic recently disclosed that it caught a team of Chinese hackers using its systems to automate attacks on large targets in ways that would have been impossible only a few years ago, given the sheer speed at which the attack was coordinated.\n\nBut the obsessive focus on the security dimension, to the exclusion of the other ways this technology will reshape economies and lives, does the United States and the wider world a disservice. What does it actually mean to \"win the AI race\"? What does the country look like if that goal is achieved?\n\nThe disconnect shows up most clearly in how individuals and institutions are preparing — or failing to prepare. Students are making college decisions, and more consequentially in the US, student loan decisions, based on starting-salary data from 2021 and 2022, when desperate companies were throwing money at anyone willing to show up. It is not entirely their fault: they are told the mantra that a college degree is crucial to their future. This does not mean nobody should go to college — far from it. But it does mean certain career fields must be examined objectively in terms of job prospects and return on investment, given the astronomical size of the average US student loan. A student taking on $150,000 to study software engineering today is making a very different bet than they would have five years ago.\n\nSome schools are adapting thoughtfully. The University of Washington's Allen School of Computer Science has been open to revamping its approach even where that means radical change. Its director, Magdalena Balazinska, observed that \"coding, or the translation of a precise design into software instructions, is dead.\" The school is weighing coordinated, sweeping curriculum changes after encouraging professors to experiment with AI integration. Other fields have moved the opposite direction. Recognizing that students widely use AI on out-of-class assignments, some have revived the dreaded blue book exam, requiring long-form essays written in person without artificial assistance. These are sensible steps, but they remain exceptions rather than the rule. Most institutions are coasting on inertia and the cultural weight placed on holding a degree — any degree, at any cost — essentially pretending nothing fundamental has changed and hoping the fallout will not be too severe.\n\nWhat makes all of this especially risky is the speed of change. Previous innovation still came in waves, but those waves were felt over decades, not years, let alone months. The labor force may have transformed considerably between 1940 and 1980, and again from 1980 to 2020, but hardly at all from 2020 to 2022. AI capabilities are compounding in ways that rewrite adoption timelines and are nearly impossible to plan a life around. Consider someone who entered college in the fall of 2022 to study software engineering, steeped in a \"learn to code\" culture, who took out large loans after being told they were \"positive debt,\" only to watch ChatGPT launch in November of his freshman year. On a standard four-year path, that same student graduates this coming spring into a completely different world than the one he entered.\n\nThe policy proposals that do exist are not especially serious. Some suggest retraining programs, but that ignores the core problem: retrain for what? AI is not phasing out a single older technology and the workers who use it. It is coming for every form of older technology at once. The skilled trades, long understaffed across Western countries, offer a comparative bastion of job security in this storm, but moving from a desk job to becoming a mechanic or plumber is more than a retraining exercise. For younger people weighing where to pivot, though, it at least deserves to be on the table. While institutions experiment with tweaks and politicians mention the issue in passing, the disconnect remains: the technology is advancing faster than people are preparing for its consequences, and a generation is making life-altering financial decisions based on data from a world that is beginning to disappear.\n\n## The Planning Vacuum\n\nIt would be wrong to end on pure pessimism. AI is genuinely going to bring enormous benefits and spark remarkable innovations, raising living standards and lengthening lifespans by helping find cures for stubbornly incurable diseases. That is not in dispute — the tech industry has made this case repeatedly, and it is not wrong. Once society reaches some state of \"full AI,\" things are almost certain to be far better for many. The danger lies in the road from here to there.\n\nOur economic systems are not built to sustain prolonged double-digit unemployment, much less anything near the 20% that Anthropic's CEO has suggested is possible. And that pain would not fall only on the 20% without jobs. Governments around the world depend on income-tax revenue to fund the social safety nets meant to protect people in emergencies. Hollow out the workforce, and you hollow out the fiscal base that backstops everyone else.\n\nThere are faint signs that the political conversation could shift. Both Senator Bernie Sanders and Republican Governor Ron DeSantis have recently weighed in on the implications for recent and upcoming graduates — an unlikely pairing that hints at bipartisan attention. But beneath such moments, there is little evidence that anyone has seriously planned how this transition will unfold, which is perhaps the most alarming part of all. When pressed for specifics on how programs like universal basic income would function or be funded, answers from even industry leaders turn noticeably vague.\n\nAs automating companies grow more efficient, governments may — or, more realistically, may be forced to — tap into taxing those profits more directly rather than relying on the current employee-based system. That approach might work, but whether it could fund the ever-growing demand for UBI that AI executives themselves are calling for is far from clear. These are enormous economic proposals, sometimes discussed in certain circles as foregone conclusions while being treated as foreign concepts in the circles that would actually have to implement them.\n\nThe uncomfortable reality is that irreversible changes are being made to how our societies and economies operate with almost no input from the people who will bear their effects. Nobody was asked whether they supported rolling this technology out to widespread adoption. It will have its upsides, but it is also almost certain to spark significant backlash once the impact is truly felt. On some level, that reaction will be understandable. Life does not pause because society is in transition; rent is still due at the end of the month, and people are expected to navigate it all with less guidance than ever on how to do so. Politicians have largely skated by without answering the hard questions — but that has gone on too long. It is high time societies begin discussing this for what it really is.\n\n## Frequently Asked Questions\n\n**Why is this round of automation different from earlier technological shifts?**\nPast innovations automated physical tasks but still required human minds to supervise, operate, maintain, and design the new systems, and they spawned entire new industries — software, IT support, digital services — that absorbed displaced workers. AI replicates cognitive work itself, and it is positioned to automate supervision, operation, maintenance, and design alike, without creating comparable new employment at scale.\n\n**What evidence shows AI is already affecting US jobs?**\nFederal Reserve Chairman Jerome Powell stated that \"job creation is pretty close to zero\" and named AI as a primary driver of layoffs and hiring freezes. Companies announced 153,074 job cuts in October 2025, the highest for that month in over two decades, led by warehousing and technology. Microsoft, Amazon, Meta, and Salesforce all made sizable cuts, many tied explicitly to AI adoption, even while posting strong profits.\n\n**How can companies be cutting jobs while reporting record profits?**\nThese layoffs are not survival measures during a downturn. Microsoft reported $27 billion in profit in the same quarter it announced 9,000 layoffs, Amazon posted a 39% profit increase to $21.2 billion, and Meta's cuts came amid robust revenue growth. Firms are restructuring around AI to widen margins, and Wall Street is rewarding the reductions rather than punishing them.\n\n**Why does the \"consumer death spiral\" matter for the broader economy?**\nConsumer spending makes up 68% of US GDP, and sustaining it requires people to earn steady incomes. As companies collectively replace workers with AI to boost individual profits, they undermine the very consumer base their business models rely on. Distress signals are already visible: record subprime auto delinquencies of 6.65%, $1.23 trillion in credit card debt, and over 9 million student-loan borrowers delinquent or in default.\n\n**Why won't traditional policy tools fix this?**\nThe usual response to mass unemployment is cutting interest rates to stimulate lending and spur hiring. That mechanism depends on the link between business growth and job creation — a link AI is severing. Companies are now expanding revenue without adding workers, so cheaper money no longer reliably translates into jobs. Meanwhile federal debt exceeds $38 trillion, limiting fiscal room to respond.\n\n**Who is most exposed to AI displacement right now?**\nEntry-level workers and recent graduates are bearing the brunt. Recent-graduate hiring at major firms has fallen 35% versus 2023, and job postings in some tech hubs have dropped as much as 50% from pre-pandemic levels. The bottom rungs of the career ladder are being removed, leaving young people who took on large student loans for fields like software engineering facing a labor market very different from the one they planned for.\n\n**What policy responses are being discussed, and are they adequate?**\nSuggestions include retraining programs and universal basic income, but both face hard questions. Retraining assumes there is a safe field to move into, yet AI targets many forms of work at once; the skilled trades offer relative security but require more than a short course to enter. UBI would demand a major shift toward taxing automated profits rather than employee wages, and even its advocates are vague on how it would be funded. Serious, detailed planning remains largely absent.\n\n## Sources\n\n- https://www.bls.gov/opub/mlr/2024/article/employment-continues-to-expand-in-2023-though-at-a-slower-pace-than-in-the-previous-2-years.htm\n- https://www.bls.gov/opub/ted/2025/employment-up-256000-in-december-2024-average-gain-of-186000-jobs-per-month-in-2024.htm\n- https://www.investopedia.com/the-economy-just-lost-nearly-a-million-jobs-on-paper-11806319\n- https://www.axios.com/2025/07/01/us-job-openings-jolts-may\n- https://www.cnn.com/business/live-news/us-jobs-report-august-2025\n- https://fortune.com/2025/10/30/jerome-powell-ai-bubble-jobs-unemployment-crisis-interest-rates/\n- https://www.pbs.org/newshour/politics/meta-signs-20-year-deal-with-nuclear-plant-signals-ais-growing-energy-needs\n- https://www.reuters.com/business/world-at-work/layoffs-us-october-surge-two-decade-high-challenger-data-shows-2025-11-06/\n- https://www.wsj.com/business/earnings/amazon-amzn-q3-earnings-report-2025-553e6d16\n- https://apnews.com/article/microsoft-layoffs-d1f2de54ebad6f099deac8fbd3375835\n- https://www.theverge.com/news/693535/microsoft-layoffs-july-2025-xbox\n- https://www.theregister.com/2025/05/16/microsofts_axe_software_developers/\n- https://techcrunch.com/2025/04/29/microsoft-ceo-says-up-to-30-of-the-companys-code-was-written-by-ai/\n- https://www.wsj.com/tech/amazon-to-layoff-tens-of-thousands-of-corporate-workers-056ebc4d\n- https://www.cbsnews.com/news/meta-layoffs-5-percent-workforce-cuts-low-performers/\n- https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-on-generative-ai\n- https://finance.yahoo.com/news/salesforce-ceo-marc-benioff-says-145324020.html\n- https://www.ainvest.com/news/microsoft-reports-q4-revenue-76-4bn-azure-reaches-75bn-annual-revenue-milestone-2507/\n- https://www.wsj.com/tech/ai/ai-white-collar-job-loss-b9856259\n- https://x.com/elonmusk/status/1980765809338147193\n- https://www.cnbc.com/2025/09/07/ai-entry-level-jobs-hiring-careers.html\n- https://www.hiringlab.org/2025/07/30/the-us-tech-hiring-freeze-continues/\n- https://fortune.com/2025/08/03/ai-interviewers-job-seekers-unemployment-hiring-hr-teams/\n- https://ourworldindata.org/grapher/share-of-population-living-in-extreme-poverty\n- https://ilostat.ilo.org/those-left-behind-the-forgotten-in-the-fight-against-global-poverty/\n- https://blogs.worldbank.org/en/voices/Year-in-Review-2015-12-Charts\n- https://www.wsj.com/business/companies-hiring-jobs-ai-9ef675b6\n- https://americanbazaaronline.com/2025/10/16/jpmorgan-goldman-sachs-deploy-ai-as-banks-brace-for-workforce-changes-468864/\n- https://www.reuters.com/business/world-at-work/goldman-sachs-eyes-job-cuts-hiring-slowdown-amid-ai-push-memo-shows-2025-10-14/\n- https://nypost.com/2025/09/29/business/walmart-ceo-issues-ominous-warning-that-ai-will-change-literally-every-job/\n- https://www.businessinsider.com/recession-us-economy-outlook-lower-income-debt-rates-el-erian-2025-11\n- https://www.businessinsider.com/recession-consumer-spending-uneployment-jobs-business-investment-real-estate-2025-10\n- https://www.reuters.com/business/autos-transportation/record-number-subprime-borrowers-miss-car-loan-payments-october-data-shows-2025-11-12/\n- https://www.bloomberg.com/news/articles/2025-10-17/auto-loan-delinquencies-jump-50-as-car-prices-reach-new-heights\n- https://www.theguardian.com/business/2025/oct/17/us-car-repossessions-economy\n- https://www.reuters.com/business/us-household-debt-up-modestly-third-quarter-new-york-fed-says-2025-11-05/\n- https://www.investopedia.com/credit-card-delinquency-rates-hit-levels-not-seen-since-2012-8691090\n- https://www.mba.org/news-and-research/newsroom/news/2025/11/14/mortgage-delinquencies-increase-in-the-third-quarter-of-2025\n- https://www.businessinsider.com/klarna-reassigns-workers-to-customer-support-after-ai-quality-concerns-2025-9\n- https://www.bbc.com/news/world-us-canada-62664181\n- https://www.cnbc.com/2025/10/15/more-student-loan-borrowers-risk-default-as-late-payments-rise.html\n- https://fortune.com/2025/11/13/38-trillion-national-debt-peter-peterson-foundation-historians-economists/\n- https://thehill.com/policy/technology/5606336-ai-cyberattack-anthropic-hackers/\n- https://www.wsj.com/business/chatgpt-ai-cheating-college-blue-books-5e3014a6\n- https://www.geekwire.com/2025/coding-is-dead-uw-computer-science-program-rethinks-curriculum-for-the-ai-era/\n- https://x.com/RonDeSantis/status/1989309864841978324\n- https://x.com/BernieSanders/status/1982869371296067665\n- https://www.npr.org/2023/06/30/1182216970/supreme-court-student-loan-forgiveness-decision-biden\n\n" url: https://homefronts.pub/article/the-american-economy-isnt-ready-for-ai.md canonical: https://homefronts.pub/article/the-american-economy-isnt-ready-for-ai datePublished: 2026-06-03 dateModified: 2026-06-03 author: - name: Simon Whistler url: https://homefronts.pub/author/simon-whistler publisher: HomeFronts image: "https://media.homefronts.pub/cdn-cgi/image/width=1600,height=900,fit=cover,quality=80,format=auto/articles/MGa2SQ_eRq8/hero.jpg" type: Article contentHash: 6e67046d970274cd269039167431f2fdaf17fce7746f2dc4b4224535db57747c tokens: 9545 summaryUrl: https://homefronts.pub/article/the-american-economy-isnt-ready-for-ai.md.summary.md --- "CEOs Start Saying the Quiet Part Out Loud: AI Will Wipe Out Jobs." That was the Wall Street Journal headline, in July of 2025, atop an article that future generations may remember as a turning point. In it, Ford Motor Company's chief executive warned that AI would replace "literally half of all white collar workers" at the manufacturer. And he is far from alone in thinking that way. Fast forward several months, and that displacement is no longer theoretical. It is showing up in employment data, particularly in the United States, as hiring slows to a crawl and unemployment begins to tick up for the first time in years. What makes this so different from previous downturns is that these companies are not shedding employees because times are tough. For some of them, times have never been better. AI is going to improve living standards for billions of people, and it carries enormous potential to be one of the most profound forces for positive change in human history. But that promise holds only if we get the transition right, because the consequences of mishandling it would be equally large in the other direction — possibly fatal. The defining risk of this moment is not whether AI eventually makes the world richer, but whether the American economy can survive the disorderly, under-planned road between here and there. ## Key Takeaways - US hiring has collapsed to near zero, with Federal Reserve Chairman Jerome Powell acknowledging "job creation is pretty close to zero" and citing AI as a primary driver of layoffs and freezes. - Companies announced 153,074 job cuts in October 2025 alone — the highest for that month in over two decades — even as firms like Microsoft and Amazon posted record profits. - AI differs from past innovations because it replaces cognitive work without generating comparable new categories of employment at scale, severing the centuries-old link between corporate growth and hiring. - Consumer spending is 68% of US GDP, and signs of household distress are mounting: subprime auto delinquencies hit a record 6.65%, credit card debt reached $1.23 trillion, and over 9 million borrowers are delinquent or in default on student loans. - Entry-level work is being hit hardest, with recent-graduate hiring at major firms down 35% versus 2023 and some tech-hub postings down as much as 50% from pre-pandemic levels. - Policy discussion is nearly absent, dominated by national security framing, while young people make life-altering financial decisions based on a labor market that is disappearing. ## The Displacement That's Already Begun Artificial intelligence crossed a threshold somewhere in late 2022 that very few people saw coming. When ChatGPT debuted that November, it took the world by storm. It was not very good at much of anything, but it technically worked. Competitors quickly followed in its wake: Anthropic released Claude, then Google released Gemini, both of which made memorable mistakes — including the moment Gemini advised people to eat a small rock each day for its supposed nutritional benefit. The advancement since then has been remarkable. We are barely three years into the life of these systems, and what began as a curiosity — something to play with, and perhaps to worry about in some distant future — has become something else entirely. Meta is now signing deals to draw power from a nuclear plant, entire technology ecosystems are being built and operated around these models, and OpenAI's Sora can produce short videos nearly indistinguishable from real footage. Warnings about AI transforming work have circulated for years. What has shifted recently is the perception of that displacement: it has moved from a looming, far-out threat to something front and center, and that trend looks unlikely to reverse. The labor data is beginning to reflect that shift. The US job market has been weakening for some time. There is genuine dispute about exactly when the slide began, but the consensus holds that it took a real hit in 2024. The economy was initially thought to have averaged a gain of 186,000 jobs per month that year — not terrible, but a noticeable drop from 2023's healthier 251,000 per month. That assessment was overturned when the Bureau of Labor Statistics revealed that the prior year had actually added 911,000 fewer jobs than previously believed. The revision, one of the largest in recent history, meant that what had looked like modest growth was actually closer to stagnation. The end of 2024 is when the bottom really began to give way. The economy had entered what analysts were calling a "no hire, no fire" dynamic, in which companies were not eager to take on new personnel but were not yet confident enough to begin layoffs at any meaningful scale. This unusual stability reflected both a declining need to replace departing workers and a kind of corporate post-traumatic stress following the great labor shortage of 2021, when firms were bending over backward to attract applicants for roles they desperately needed to fill. That dynamic is now changing, and the labor market has begun to soften further, especially in the United States. Job creation in recent months has essentially collapsed. Even Jerome Powell, the chairman of the Federal Reserve, conceded that "job creation is pretty close to zero," noting that "a significant number of companies" have begun layoffs and hiring freezes and citing AI as the primary driver. His acknowledgment raises the question of why it took so long to recognize what had been increasingly clear for months. The October data simply confirmed what had been unfolding in plain sight. Challenger, Gray and Christmas released a report during the US government shutdown showing that companies announced 153,074 job cuts in October alone — the highest for that month in over two decades, and nearly triple the figure from the previous month. Warehousing and the technology sector led the carnage, with over 47,000 and 33,000 cuts respectively, several times higher than their September levels, as firms restructured around AI integration. ## When Record Profits Come With Layoffs The tech sector's cuts have been particularly aggressive. Microsoft led the way with 6,000 cuts in May, another 9,000 in July, and continued trimming through September. Software engineers bore the brunt, accounting for roughly forty percent of eliminated positions — a striking figure given that CEO Satya Nadella has acknowledged AI now writes up to thirty percent of Microsoft's code. Amazon slashed 14,000 corporate positions, with its CEO explicitly connecting the cuts to AI adoption and a vision of a company with "fewer people doing some of the jobs that are being done today." Meta opened the year by cutting 3,600 employees, while Salesforce's customer service division shrank from roughly 9,000 to 5,000 employees over eighteen months. All of this would read differently if these companies were struggling financially — if they were trimming payroll to survive a downturn or stave off bankruptcy. But that is not what is happening. Microsoft reported $27 billion in profit in the very quarter it announced the 9,000-person layoff wave, one of the best quarters in company history, and its stock performance reflected it. The pattern repeats across the others. Amazon has reported profit increases of 39%, reaching $21.2 billion, and Meta's cuts came amid robust revenue growth. Concerns about automation are, of course, nothing new. People complaining about technology upending how they live and earn is as old as industry itself. The Luddites are the most infamous example — workers who smashed the machinery they believed threatened their livelihoods. The short version is that it did not work. Technology kept advancing, the Luddites were largely forgotten, and a similar fate met most who refused to adapt to a rapidly changing world. Yet today's situation really does appear different — less because of what is changing than because of who is sounding the alarm. In the Luddites' era, it was workers and skeptics raising concerns while experts dismissed them. This time, it is the leaders of some of the country's largest firms warning of the disruption, while the ordinary person on the street is far more likely to be going about the day, unaware of what is coming. That reversal of roles should give us pause. When the people building and deploying a technology are the ones raising concerns about its impact, it is worth listening. When Ford's Jim Farley told the Wall Street Journal that AI is on track to displace "literally half of all white collar workers" in the country, he was not lying — he was warning. Other prominent voices in technology have offered similar predictions. Elon Musk has proclaimed that AI will make work optional and replace "all jobs," while Bill Gates believes that within ten years we will inhabit a world in which machines perform most of the economically valuable work. For those who doubt that this is coming, there is no need to take it on faith — but doubting the people who design these platforms is its own kind of risk. ## Why This Wave Is Different There is something else that separates this moment from earlier technological upheavals. All innovation has cost some jobs, so at one level this is not radically new. The steam engine displaced horse-drawn carriages, but it created jobs manufacturing engines, producing and laying rail lines, and operating locomotives. The arrival of computers wiped out the entire typist pool — until relatively recently a large field of workers — but it created whole industries around software development, IT support, and digital services that employed far more people than were displaced. Even if the office typist never went on to write code, the growth of those new industries created supporting jobs. Every major technological shift in modern history followed this pattern: initial displacement, followed by the creation of new categories of work that absorbed the displaced. There was lingering fallout, but there was also a variety of new fields to move into. Artificial intelligence is fundamentally different and should not be lumped in with those earlier waves, because it operates from a different premise. It replicates and can replace human cognitive work without generating comparable new employment at scale. When previous technologies automated physical tasks, they still required human minds to supervise, operate, maintain, and, most critically, design them and their successors. AI has the ability to automate all of these functions — if not today, and perhaps not tomorrow, then in the not-so-distant future. Companies are not waiting for some future breakthrough to begin restructuring their workforces around AI. The changes are happening now, especially at the entry points where workers traditionally launched their careers. It has been a running joke for years that entry-level positions somehow require three to five years of prior experience, but matters have recently escalated to a new level. Recent-graduate hiring at major firms has fallen 35 percent compared with 2023, while job postings in certain tech hubs have dropped as much as 50 percent from pre-pandemic levels. This is less a story of the ladder getting harder to climb than of the bottom rungs being sawed off entirely. The human impact is beginning to surface in stories that hint at what may be coming. Allen Rausch, a 56-year-old who spent decades working for companies like Amazon and Electronic Arts, lost his job earlier in 2025. While applying for new positions, he was stunned to find AI conducting his interviews — and not just one of them. Each lasted around 20 minutes and proved totally unable to answer any of his questions, simply moving on to the next prompt. He was being assessed by a system so unsophisticated it could not respond to even the most basic query. "Given the percentage of responses that I'm getting to my applications, I think a lot of AI interviews are wasting my time," he told reporters. What makes Rausch's experience particularly ominous is that most companies remain in the early phases of AI adoption — though that is changing. And those that have moved early are often not shy about it. Klarna, the Swedish fintech company, openly boasted that its AI assistant could handle the workload of 700 customer service agents, managing millions of conversations a month across 35 languages. That rollout did not go as planned. Like the system that interviewed Rausch, Klarna's tools simply were not sophisticated enough for the job, and the company faced an intense backlash as customers found they could not reach a real person when they needed help. By mid-2025, Klarna was forced to admit it had gone too far, too fast with automation, and that its focus on cost-cutting had degraded service quality. The company has since walked back its AI-exclusive approach. While that episode serves as a warning to other firms about the risks of going overboard, it is not exactly comforting. The lesson Klarna drew was that AI is not currently able to fully replace all people. The implied "yet" practically writes itself. These implications extend well beyond individual job losses. What is breaking down is a fundamental economic relationship: the one linking corporate expansion to job creation. For two centuries, companies that grew needed more workers — that was simply a cost of doing business. That equation is now being rewritten. ## The Corporate Calculus Before getting too far ahead, a few caveats are worth stating plainly. We are still early in this transition, and exactly how it will play out is impossible to predict, especially given uneven distribution across countries, labor laws, and industries. It is also worth emphasizing something often lost in debates about displacement: companies are not charities. They will act in their own interest to maximize profits for themselves and their shareholders. This is not said cruelly, but to underscore the framework on which the system runs. For the last two centuries, businesses expanded relentlessly — first across cities, then states and provinces, and ultimately across borders — and that expansion proved remarkably effective at reducing poverty worldwide. In 1820, roughly 80% of the global population lived in what we would recognize as extreme poverty. By 2024, that figure had fallen to around 10%. The problem is far from solved, but the overall living condition on earth has improved dramatically over that span. Between 1990 and 2015 alone, the number of people in extreme poverty fell from just under 2 billion to 702 million — a staggering improvement that deserves far more attention than it receives. Much of that progress depended on a simple equation: as companies grew and reached new markets, they needed more workers to produce, distribute, and sell their products. Factory expansions meant more assembly-line workers; new stores required cashiers, stockers, and managers; international expansion meant entirely new supply chains staffed by people across the globe to move a product from point A to point B. That connection is increasingly being severed. In 2023 and 2024, executives spoke cautiously about the new technology, emphasizing "human-AI collaboration" if they said anything at all. By 2025, the message shifted, and companies began speaking openly about expanding without adding employees. A Wall Street Journal banner on October 26th captured it: "More Big Companies Bet They Can Still Grow Without Hiring." The report was a litany of the same dynamic. JPMorgan's chief financial officer told investors that the bank now has a "very strong bias against having the reflective response" to hire more people for any given need. They are hardly alone. The article reads like a roll call of American business, from Walmart to Airbnb to Target and beyond. Walmart explicitly told investors it plans to keep headcount essentially flat over the coming years even as sales grow, with AI handling tasks that would once have required new hires. The financial data makes the disconnect even more revealing. Over the past three years, publicly traded US companies have cut white-collar headcount by roughly 3.5% while revenues climbed — a pattern that has recently accelerated. A significant share of S&P 500 firms now employ fewer people than they did a decade ago, even as many have posted substantial revenue growth. Put bluntly, an AI model might not be better than you at your job — but it is increasingly about as good. Wall Street is rewarding the trend enthusiastically. Companies have been tripping over themselves to announce how they have "innovated" AI into some product — from $400 AI toothbrushes that supposedly improve your technique to a Samsung AI fridge that uses cameras to monitor groceries and warn you when the milk is about to expire. These may not quite be the revolution the marketing teams promised. Even so, executives now brag on earnings calls about headcount reductions. The striking thing is that such moves used to signal the opposite: layoffs once told investors a company feared trouble downstream, while hiring signaled optimism about future growth. As adoption progresses, it creates a self-reinforcing dynamic. Companies that adapt are seen as better run, more efficient, and more profitable. If your competitors achieve 20% margin improvements by replacing human workers with AI, you face a choice — match them or risk being seen as uncompetitive. The result is essentially an arms race toward automation, with each firm trying to move faster than its rivals. ## The Consumer Death Spiral The United States, for better or worse, appears to be at the forefront of adopting this technology — and it is not entering that race in the best economic shape. Companies operate along a purely competitive logic that makes collective restraint next to impossible. Walmart is not going to call Target and propose they both limit AI adoption to preserve jobs, because even if they agreed, they would expose themselves to some other firm outside the pact. And even if all the so-called big players signed on, they would still have to perform financially, leaving them vulnerable to a newcomer unbound by their mutual agreement. If a single company could automate fully while no one else followed, the economy would chug along with few any the wiser. It is the collective push toward automation that poses a mutually assured threat. Consumer spending represents 68% of US GDP, and sustaining that spending requires people to have a steady income. As companies race to replace workers with AI to boost individual profits, they are collectively undermining the consumer base their own business models depend on. The American consumer is already showing serious signs of distress. The clearest signal comes from the auto loan market, a classic indicator of genuine economic struggle — losing a car is among the last things most Americans will allow, given how essential it is for getting to work and managing daily life. The delinquency rate, measured by loans more than 60 days past due, is up more than 50% over the last decade and a half. Among subprime borrowers, delinquency reached 6.65% in October, the highest ever recorded in American history. Repossession orders have jumped more than 40% since 2022, hitting the highest level since 2009. What makes this especially alarming is that the country has not even seen a large spike in unemployment yet. Trends in recent-graduate joblessness are genuinely concerning, but the latest data still shows overall unemployment relatively low by historical standards. Some of the people caught in these statistics are recently unemployed, but many still hold jobs. Part of the explanation lies in the historically unprecedented stimulus payments that some households grew accustomed to during the pandemic and its immediate aftermath. A combination of government support and record-low interest rates allowed people to qualify for auto loans they otherwise could not have afforded. All of this is now feeding into a moment in which reality is beginning to catch up. It does not change the fact that Americans appear poised to enter this economic transition at a particularly weak point. Credit cards tell a similar story of mounting pressure. Total credit card debt has exploded to $1.23 trillion, with the seriously delinquent rate — accounts more than 90 days past due — hitting 11%, the highest since 2012. Even the mortgage market, relatively stable in terms of defaults, is showing cracks: delinquencies on home loans have crept up to 3.99% for borrowers more than 60 days past due, the highest in years. Student loans add another dimension to the squeeze. After being paused for over three years during the pandemic, when payments were granted a grace period that was repeatedly extended, millions of households suddenly faced hundreds of dollars in monthly payments many had set aside, imagining the day of reckoning might never arrive. They were not entirely unreasonable to think so: the Biden administration made numerous pushes to forgive student debt, including a wide-reaching plan that would have canceled up to $10,000 per borrower for an estimated 26 million applicants. The Supreme Court struck it down in June 2023. Over 9 million borrowers are now either delinquent or in outright default. And unlike most other debt in the US, student loans usually cannot be discharged in bankruptcy — borrowers must prove "undue hardship," a standard rarely met, so balances often follow people for decades. What connects all of these data points is timing. The American consumer is maxed out, and there is little slack left in the system. The government's fiscal position makes matters more precarious still. The US federal debt now stands at over $38 trillion, the largest debt-to-GDP ratio since the Second World War — except this time it was run up during peacetime, in an expanding economy. If the consumer is entering this transition in a fragile state, the government is not faring much better. The traditional policy response to mass unemployment — cutting interest rates to stimulate lending and thereby spur hiring — will not work this time, because the link between cheaper money and job creation has been severed. The implications reach far beyond individuals simply buying less. There is always a contagion effect from layoffs. The US has become so dependent on consumer spending that it can hardly afford not to spend; even a minor pullback could carry significant consequences, something the legendary investor Mohamed El-Erian has said keeps him up at night. ## Flying Blind Perhaps the most striking aspect of this moment is the near-total absence of serious policy discussion about what is unfolding. A handful of politicians have come forward to raise alarms, but they represent a tiny minority, and when officials do speak about the issue, they almost always frame it through national security. To be clear, given how capable this technology has already become, this genuinely is a national security concern. Anthropic recently disclosed that it caught a team of Chinese hackers using its systems to automate attacks on large targets in ways that would have been impossible only a few years ago, given the sheer speed at which the attack was coordinated. But the obsessive focus on the security dimension, to the exclusion of the other ways this technology will reshape economies and lives, does the United States and the wider world a disservice. What does it actually mean to "win the AI race"? What does the country look like if that goal is achieved? The disconnect shows up most clearly in how individuals and institutions are preparing — or failing to prepare. Students are making college decisions, and more consequentially in the US, student loan decisions, based on starting-salary data from 2021 and 2022, when desperate companies were throwing money at anyone willing to show up. It is not entirely their fault: they are told the mantra that a college degree is crucial to their future. This does not mean nobody should go to college — far from it. But it does mean certain career fields must be examined objectively in terms of job prospects and return on investment, given the astronomical size of the average US student loan. A student taking on $150,000 to study software engineering today is making a very different bet than they would have five years ago. Some schools are adapting thoughtfully. The University of Washington's Allen School of Computer Science has been open to revamping its approach even where that means radical change. Its director, Magdalena Balazinska, observed that "coding, or the translation of a precise design into software instructions, is dead." The school is weighing coordinated, sweeping curriculum changes after encouraging professors to experiment with AI integration. Other fields have moved the opposite direction. Recognizing that students widely use AI on out-of-class assignments, some have revived the dreaded blue book exam, requiring long-form essays written in person without artificial assistance. These are sensible steps, but they remain exceptions rather than the rule. Most institutions are coasting on inertia and the cultural weight placed on holding a degree — any degree, at any cost — essentially pretending nothing fundamental has changed and hoping the fallout will not be too severe. What makes all of this especially risky is the speed of change. Previous innovation still came in waves, but those waves were felt over decades, not years, let alone months. The labor force may have transformed considerably between 1940 and 1980, and again from 1980 to 2020, but hardly at all from 2020 to 2022. AI capabilities are compounding in ways that rewrite adoption timelines and are nearly impossible to plan a life around. Consider someone who entered college in the fall of 2022 to study software engineering, steeped in a "learn to code" culture, who took out large loans after being told they were "positive debt," only to watch ChatGPT launch in November of his freshman year. On a standard four-year path, that same student graduates this coming spring into a completely different world than the one he entered. The policy proposals that do exist are not especially serious. Some suggest retraining programs, but that ignores the core problem: retrain for what? AI is not phasing out a single older technology and the workers who use it. It is coming for every form of older technology at once. The skilled trades, long understaffed across Western countries, offer a comparative bastion of job security in this storm, but moving from a desk job to becoming a mechanic or plumber is more than a retraining exercise. For younger people weighing where to pivot, though, it at least deserves to be on the table. While institutions experiment with tweaks and politicians mention the issue in passing, the disconnect remains: the technology is advancing faster than people are preparing for its consequences, and a generation is making life-altering financial decisions based on data from a world that is beginning to disappear. ## The Planning Vacuum It would be wrong to end on pure pessimism. AI is genuinely going to bring enormous benefits and spark remarkable innovations, raising living standards and lengthening lifespans by helping find cures for stubbornly incurable diseases. That is not in dispute — the tech industry has made this case repeatedly, and it is not wrong. Once society reaches some state of "full AI," things are almost certain to be far better for many. The danger lies in the road from here to there. Our economic systems are not built to sustain prolonged double-digit unemployment, much less anything near the 20% that Anthropic's CEO has suggested is possible. And that pain would not fall only on the 20% without jobs. Governments around the world depend on income-tax revenue to fund the social safety nets meant to protect people in emergencies. Hollow out the workforce, and you hollow out the fiscal base that backstops everyone else. There are faint signs that the political conversation could shift. Both Senator Bernie Sanders and Republican Governor Ron DeSantis have recently weighed in on the implications for recent and upcoming graduates — an unlikely pairing that hints at bipartisan attention. But beneath such moments, there is little evidence that anyone has seriously planned how this transition will unfold, which is perhaps the most alarming part of all. When pressed for specifics on how programs like universal basic income would function or be funded, answers from even industry leaders turn noticeably vague. As automating companies grow more efficient, governments may — or, more realistically, may be forced to — tap into taxing those profits more directly rather than relying on the current employee-based system. That approach might work, but whether it could fund the ever-growing demand for UBI that AI executives themselves are calling for is far from clear. These are enormous economic proposals, sometimes discussed in certain circles as foregone conclusions while being treated as foreign concepts in the circles that would actually have to implement them. The uncomfortable reality is that irreversible changes are being made to how our societies and economies operate with almost no input from the people who will bear their effects. Nobody was asked whether they supported rolling this technology out to widespread adoption. It will have its upsides, but it is also almost certain to spark significant backlash once the impact is truly felt. On some level, that reaction will be understandable. Life does not pause because society is in transition; rent is still due at the end of the month, and people are expected to navigate it all with less guidance than ever on how to do so. Politicians have largely skated by without answering the hard questions — but that has gone on too long. It is high time societies begin discussing this for what it really is. ## Frequently Asked Questions **Why is this round of automation different from earlier technological shifts?** Past innovations automated physical tasks but still required human minds to supervise, operate, maintain, and design the new systems, and they spawned entire new industries — software, IT support, digital services — that absorbed displaced workers. AI replicates cognitive work itself, and it is positioned to automate supervision, operation, maintenance, and design alike, without creating comparable new employment at scale. **What evidence shows AI is already affecting US jobs?** Federal Reserve Chairman Jerome Powell stated that "job creation is pretty close to zero" and named AI as a primary driver of layoffs and hiring freezes. Companies announced 153,074 job cuts in October 2025, the highest for that month in over two decades, led by warehousing and technology. Microsoft, Amazon, Meta, and Salesforce all made sizable cuts, many tied explicitly to AI adoption, even while posting strong profits. **How can companies be cutting jobs while reporting record profits?** These layoffs are not survival measures during a downturn. Microsoft reported $27 billion in profit in the same quarter it announced 9,000 layoffs, Amazon posted a 39% profit increase to $21.2 billion, and Meta's cuts came amid robust revenue growth. Firms are restructuring around AI to widen margins, and Wall Street is rewarding the reductions rather than punishing them. **Why does the "consumer death spiral" matter for the broader economy?** Consumer spending makes up 68% of US GDP, and sustaining it requires people to earn steady incomes. As companies collectively replace workers with AI to boost individual profits, they undermine the very consumer base their business models rely on. Distress signals are already visible: record subprime auto delinquencies of 6.65%, $1.23 trillion in credit card debt, and over 9 million student-loan borrowers delinquent or in default. **Why won't traditional policy tools fix this?** The usual response to mass unemployment is cutting interest rates to stimulate lending and spur hiring. That mechanism depends on the link between business growth and job creation — a link AI is severing. Companies are now expanding revenue without adding workers, so cheaper money no longer reliably translates into jobs. Meanwhile federal debt exceeds $38 trillion, limiting fiscal room to respond. **Who is most exposed to AI displacement right now?** Entry-level workers and recent graduates are bearing the brunt. Recent-graduate hiring at major firms has fallen 35% versus 2023, and job postings in some tech hubs have dropped as much as 50% from pre-pandemic levels. The bottom rungs of the career ladder are being removed, leaving young people who took on large student loans for fields like software engineering facing a labor market very different from the one they planned for. **What policy responses are being discussed, and are they adequate?** Suggestions include retraining programs and universal basic income, but both face hard questions. Retraining assumes there is a safe field to move into, yet AI targets many forms of work at once; the skilled trades offer relative security but require more than a short course to enter. UBI would demand a major shift toward taxing automated profits rather than employee wages, and even its advocates are vague on how it would be funded. Serious, detailed planning remains largely absent. ## Sources - https://www.bls.gov/opub/mlr/2024/article/employment-continues-to-expand-in-2023-though-at-a-slower-pace-than-in-the-previous-2-years.htm - https://www.bls.gov/opub/ted/2025/employment-up-256000-in-december-2024-average-gain-of-186000-jobs-per-month-in-2024.htm - https://www.investopedia.com/the-economy-just-lost-nearly-a-million-jobs-on-paper-11806319 - https://www.axios.com/2025/07/01/us-job-openings-jolts-may - https://www.cnn.com/business/live-news/us-jobs-report-august-2025 - https://fortune.com/2025/10/30/jerome-powell-ai-bubble-jobs-unemployment-crisis-interest-rates/ - https://www.pbs.org/newshour/politics/meta-signs-20-year-deal-with-nuclear-plant-signals-ais-growing-energy-needs - https://www.reuters.com/business/world-at-work/layoffs-us-october-surge-two-decade-high-challenger-data-shows-2025-11-06/ - https://www.wsj.com/business/earnings/amazon-amzn-q3-earnings-report-2025-553e6d16 - https://apnews.com/article/microsoft-layoffs-d1f2de54ebad6f099deac8fbd3375835 - https://www.theverge.com/news/693535/microsoft-layoffs-july-2025-xbox - https://www.theregister.com/2025/05/16/microsofts_axe_software_developers/ - https://techcrunch.com/2025/04/29/microsoft-ceo-says-up-to-30-of-the-companys-code-was-written-by-ai/ - https://www.wsj.com/tech/amazon-to-layoff-tens-of-thousands-of-corporate-workers-056ebc4d - https://www.cbsnews.com/news/meta-layoffs-5-percent-workforce-cuts-low-performers/ - https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-on-generative-ai - https://finance.yahoo.com/news/salesforce-ceo-marc-benioff-says-145324020.html - https://www.ainvest.com/news/microsoft-reports-q4-revenue-76-4bn-azure-reaches-75bn-annual-revenue-milestone-2507/ - https://www.wsj.com/tech/ai/ai-white-collar-job-loss-b9856259 - https://x.com/elonmusk/status/1980765809338147193 - https://www.cnbc.com/2025/09/07/ai-entry-level-jobs-hiring-careers.html - https://www.hiringlab.org/2025/07/30/the-us-tech-hiring-freeze-continues/ - https://fortune.com/2025/08/03/ai-interviewers-job-seekers-unemployment-hiring-hr-teams/ - https://ourworldindata.org/grapher/share-of-population-living-in-extreme-poverty - https://ilostat.ilo.org/those-left-behind-the-forgotten-in-the-fight-against-global-poverty/ - https://blogs.worldbank.org/en/voices/Year-in-Review-2015-12-Charts - https://www.wsj.com/business/companies-hiring-jobs-ai-9ef675b6 - https://americanbazaaronline.com/2025/10/16/jpmorgan-goldman-sachs-deploy-ai-as-banks-brace-for-workforce-changes-468864/ - https://www.reuters.com/business/world-at-work/goldman-sachs-eyes-job-cuts-hiring-slowdown-amid-ai-push-memo-shows-2025-10-14/ - https://nypost.com/2025/09/29/business/walmart-ceo-issues-ominous-warning-that-ai-will-change-literally-every-job/ - https://www.businessinsider.com/recession-us-economy-outlook-lower-income-debt-rates-el-erian-2025-11 - https://www.businessinsider.com/recession-consumer-spending-uneployment-jobs-business-investment-real-estate-2025-10 - https://www.reuters.com/business/autos-transportation/record-number-subprime-borrowers-miss-car-loan-payments-october-data-shows-2025-11-12/ - https://www.bloomberg.com/news/articles/2025-10-17/auto-loan-delinquencies-jump-50-as-car-prices-reach-new-heights - https://www.theguardian.com/business/2025/oct/17/us-car-repossessions-economy - https://www.reuters.com/business/us-household-debt-up-modestly-third-quarter-new-york-fed-says-2025-11-05/ - https://www.investopedia.com/credit-card-delinquency-rates-hit-levels-not-seen-since-2012-8691090 - https://www.mba.org/news-and-research/newsroom/news/2025/11/14/mortgage-delinquencies-increase-in-the-third-quarter-of-2025 - https://www.businessinsider.com/klarna-reassigns-workers-to-customer-support-after-ai-quality-concerns-2025-9 - https://www.bbc.com/news/world-us-canada-62664181 - https://www.cnbc.com/2025/10/15/more-student-loan-borrowers-risk-default-as-late-payments-rise.html - https://fortune.com/2025/11/13/38-trillion-national-debt-peter-peterson-foundation-historians-economists/ - https://thehill.com/policy/technology/5606336-ai-cyberattack-anthropic-hackers/ - https://www.wsj.com/business/chatgpt-ai-cheating-college-blue-books-5e3014a6 - https://www.geekwire.com/2025/coding-is-dead-uw-computer-science-program-rethinks-curriculum-for-the-ai-era/ - https://x.com/RonDeSantis/status/1989309864841978324 - https://x.com/BernieSanders/status/1982869371296067665 - https://www.npr.org/2023/06/30/1182216970/supreme-court-student-loan-forgiveness-decision-biden <!-- youtube:MGa2SQ_eRq8 -->