Key Points
- Meta is installing tracking software on U.S. employees’ computers to help train AI agents for white-collar work, while reportedly preparing for a new round of layoffs in May with more cuts expected later in the year.
- The AI-driven layoff wave is accelerating, with Microsoft offering its first voluntary buyouts, Snap cutting 1,000 workers, and Oracle eliminating up to 30,000 roles.
- The U.S. white-collar sector has now contracted for three consecutive years, aligning with the “ripping point” scenario Whitney Tilson has warned about and suggesting a narrowing window for investors to position ahead of further disruption.
If you work at Facebook parent company Meta Platforms (META), you may soon be training your replacement, whether you like it or not.
Meta is installing new tracking software across its employees’ computers that will capture every mouse movement, every click, and every keystroke. And for good measure, periodic screenshots of everything on their screen.
Reuters broke the news, citing internal memos that it was part of Meta’s efforts to build “AI agents that can perform work tasks autonomously.”
“The vision we are building towards is one where our agents primarily do the work and our role is to direct, review and help them improve,” [Meta Chief Technology Officer Andrew] Bosworth said. The aim, he added, was for agents to “automatically see where we felt the need to intervene so they can be better next time.”
The company is building AI agents to handle everyday computer tasks for humans. The problem is that the models are clumsy at basic computer-use behaviors. They struggle with the small stuff: filling out the right form fields, navigating menus, and picking options from a list. To fix that, Meta needs real-world examples of how humans actually do this work.
So, Meta’s solution is to turn each of its employees into training data for the AI that will replace their jobs.
After all, Meta is expected to lay off roughly 10% of its global workforce starting next month, according to Reuters, with further job cuts expected later in the year.
As I wrote last month when the 16,000-employee layoff rumors first started hitting the Street:
A Meta spokesperson called the report “speculative reporting about theoretical approaches.”
That’s a whole lot of syllables to avoid saying, “Yep, that’s the plan, but we don’t want to talk about it yet.”
If the cuts do happen at that scale, it would be Meta’s largest restructuring since the “year of efficiency” in late 2022 and early 2023, when the company eliminated more than 21,000 jobs in two waves.
Meta has also created a new engineering team focused on building AI agents that can perform the majority of the work required to ship products and infrastructure inside the company. And it has collapsed several distinct job titles into a single generic label: “AI builder.”
If you’re a white-collar worker in the tech industry, that ominous ticking sound you hear is counting down whatever time you have left. Tick tock, tick tock…
Meta Employees to Train AI, Much Like Musk’s Macrohard
We wrote about this exact playbook a month ago in our piece on Elon Musk’s “Macrohard” AI system.
Musk’s project pairs xAI’s Grok with computer-vision hardware from Tesla (TSLA) to monitor a computer screen in real time, track every keystroke and mouse movement, and eventually do the user’s entire job. Musk himself described it as an AI capable of “emulating the function of entire companies.”
As we noted at the time:
Soon, you may have an “always-on” AI at your desk, watching what you do, and learning to do it without you.
That sounds dystopian… or maybe just far-fetched.
But it has happened before with American workers training their offshore replacements in the late 1990s… That trend sped up following the dot-com crash in the early 2000s as companies faced intense pressure to survive by cutting costs.
This time, instead of being replaced with a cheaper worker in a foreign country, it’s an AI in the cloud.
Meta’s version is even simpler than Musk’s and, arguably, more efficient. It just turned its own 80,000-employee workforce into the training set.
The employees show up and do their jobs as the AI learns. All it takes is one quiet piece of software running in the background of every corporate laptop to hand Meta’s Superintelligence Labs the raw material it needs to put those same employees out of work.
Snap Just Fired 1,000 People… Its Stock Jumped 7%
Meta isn’t alone, of course. Every major tech company is heading in this same direction…
Last week, social media company Snap (SNAP) announced it was cutting around a thousand employees, or roughly 16% of its workforce, and leaving 300 open roles unfilled.
Snap CEO Evan Spiegel blamed AI in a letter to staff:
“Last fall, I described Snap as facing a crucible moment, requiring a new way of working that is faster and more efficient, while pivoting towards profitable growth,” Spiegel wrote.
“We believe that rapid advancements in artificial intelligence enable our teams to reduce repetitive work, increase velocity, and better support our community, partners, and advertisers,” Spiegel added.
AI agents are already generating 65% of Snapchat’s new code, and the company expects to save more than $500 million thanks to the layoffs. Wall Street loved the news, with SNAP stock jumping roughly 7% on the announcement.
That reaction was the same one we flagged back in our February piece on the AI “doom loop” for American jobs. Then, Block (XYZ) had told its 10,000 employees that 40% of them were no longer needed. Some 4,000 jobs, gone…
“Intelligence tools have changed what it means to build and run a company,” [Block CEO Jack] Dorsey wrote. “A significantly smaller team, using the tools we’re building, can do more and do it better.”
That’s the polite version. The blunt version is that Block’s CEO believes artificial intelligence (“AI”) can do most of their jobs, and for far cheaper.
Wall Street loved it… Block shares soared more than 20% in after-hours trading.
A few weeks before the Snap news, Oracle (ORCL) had delivered the single largest AI-attributed layoff in tech history.
Multiple layoff trackers put the cut at 20,000 to 30,000 employees with the express purpose of Oracle freeing up to $10 billion in capital, most of which is earmarked for AI data-center investment.
And on April 23, Microsoft (MSFT) announced its first-ever voluntary buyout offer for U.S. employees nearing retirement. Roughly 7% of the company’s 125,000 U.S. employees are eligible.
As CNBC reported:
Microsoft has been ramping up capital spending on data centers to supply cloud clients with computing power that can handle generative AI models. Technology peers such as Alphabet and Amazon are doing the same. Meanwhile, software stocks are getting hammered as coding tools from Anthropic and others threaten to disrupt established companies.
Across the board, this is a widespread, deliberate reallocation of capital from human salaries to AI chips and computing power. And it’s not slowing down.
According to Layoffs.fyi, a prominent tech and startup layoff tracker, more than 80,000 tech employees have been axed so far this year… And we expect far more to come.
Uber’s CEO Said What Others Won’t
Against that backdrop, one CEO’s recent admissions stand out.
Uber (UBER) CEO Dara Khosrowshahi sat down with host Steven Bartlett on The Diary of a CEO podcast earlier this year and shared something few others will admit.
He shared that he often hears private conversations among fellow executives about the “sheer amount of disruption” they expect from AI, then watches those same people appear on CNBC or at Davos to tell the public that everything will work out with jobs.
His own AI forecast is unsparing. He estimates that AI will ultimately be able to handle 70% to 80% of all human work – much sooner than most folks expect.
As he predicts:
Probably in the next 10 years for intellectual jobs… 15-20 years for physical jobs because physical AI is harder. Robots, cars, et cetera, more capital-heavy [work] takes longer [since you] have to deal with the physical world. So the changes in society are going to be giant. But my view is you can’t slow down the rate of change.
And if you’re a part of that change, at least you can have some say as to how that change imprints on society and imprints on the real world.
When asked what happens to Uber’s 9.5 million drivers and couriers once autonomous vehicles dominate the network, he answered with three words: “I don’t know.”
Whitney’s “Ripping Point” Prediction Playing Out in Real Time
All of this feeds directly into something my colleague and Wall Street veteran Whitney Tilson has been warning about for weeks.
Whitney calls it America’s “ripping point.” Not a tipping point. A genuine tear across the economic fabric of our nation.
On one side, a small group of Americans is getting dramatically richer from AI-driven productivity gains and the soaring stocks of AI infrastructure companies.
On the other, a growing pool of white-collar workers is watching their careers quietly disappear into a chatbot… or, in Meta’s case, helping build their replacement with every keystroke and mouse click they make.
We believe we are in the earliest stages of what we’re calling an AI doom loop. That is, all the job losses that are happening now – with the economy still humming, consumer spending mostly holding up, and stocks hitting all-time highs – will accelerate.
As we wrote when we initially posed our AI doom-loop thought experiment:
Even after a stock market crash, the data centers don’t disappear. The chips don’t stop working. The trained models don’t forget what they’ve learned.
Companies won’t stop using AI when their stock prices fall. They’ll use it more aggressively… because now they’re desperate to cut costs.
That means when the stock market crashes, practically every job that can be replaced by AI will be replaced by AI.
Right now, the top 10% of American households account for roughly half of U.S. consumer spending. These are the high-paying, white-collar jobs that are at risk.
Already, white-collar employment has been contracting for three straight years, even as AI models keep getting cheaper and more capable.
You’ll see another tech layoff in the headlines soon. And another after that. You’ll see hundreds more as Whitney’s ripping point prediction plays out.
His advice, and ours, remains the same. Stop fearing AI. Start using it.
The same technology that has compressed a century of productivity gains into a single decade is also, for the first time, putting real analytical firepower in the hands of individual investors. Whitney’s “N.E.W. System” – short for New Engine of Wealth – is built specifically for this moment, using AI to comb through financial data on thousands of companies and find the ones most likely to outperform.
But whatever you do – no matter whether you’re one of Whitney’s paid subscribers or just a curious investor stopping by – please consider our core advice: Pay off debt. Save rather than spend. And invest in strong businesses and assets that can survive a crisis.
Because a crisis is coming.
In the coming years, we will see a rise in political leaders who promise to stop AI job losses through any means necessary. We will see the sabotage of systems training AI to replace workers, exactly as Meta is rolling out right now. We will see more targeted violence against AI executives and employees. And we will see data centers burned.
I’m not only worried for older folks like me. I have three beautiful daughters. And it feels like their future gets more unstable with every passing day.
It’s not just because of the extreme political division or violent protests that seem to be a regular part of our lives now. I mean the deeper change that we can all feel.
We’re living through a permanent reset in our economy. It’s making the world – the one our children are going to inherit – deeply unstable.
You still have time to prepare… But you may not have long.
