TL;DR
- Silicon Valley is the loudest voice saying AI is killing jobs... because software is one of the few places where AI has already delivered visible productivity gains. That's not the whole economy.
- A January 2026 Oxford Economics study found AI's impact on employment is "patchy at best." Wharton economists labeled a lot of what's being blamed on AI as "AI-washing" of job losses.
- The real driver: COVID-era overhiring, salaries inflated 30 to 40%, and companies discovering that framing cuts as "AI transformation" sends the stock price up.
- Tech layoffs appear to have bottomed. Companies are already rehiring roles they cut last year... sometimes calling back the same people.
- The right move isn't to wait and see. It's to learn the tools now, while the shift is still happening.
The conversation that started this episode came from a pretty honest place.
What do you do when you think your job might quietly disappear while everyone around you insists everything's fine?
Jackson's answer was exactly what a lot of people are feeling: wake up in the morning, tell yourself it's fine, get to work, wonder if you'll have a job in a year.
I get it. But here's what the data actually says... and it's a lot less apocalyptic than the headlines want you to believe.
The panic is running ahead of the evidence
Pearson CEO Omar Abbosh wrote in Fortune on April 6, 2026 that the AI job apocalypse is a Silicon Valley story. The broader labor market is telling a different one.
The key evidence: research published in January by Oxford Economics found the evidence of an AI-driven employment shakeup to be "patchy at best." Labor economists and AI experts at the Wharton School went further, calling much of what's being attributed to AI as "AI-washing" of job losses.
Meaning: the pink slips are real. The cause being cited is mostly not.
U.S. unemployment sits at 4.4% as of early 2026. That's not apocalypse territory. For context, EU-wide unemployment hit 11% overall in the 1990s during the process reengineering wave... and nobody called that a workforce extinction event.
The real cause: COVID bloat and a very convenient story
Here's the actual sequence of events.
During COVID, money got cheap. Companies hired like crazy, often at salaries 30 to 40% above what the same role would have commanded six months earlier. Headcount ballooned. Then interest rates climbed, global economic uncertainty picked up, and suddenly those payrolls looked unsustainable.
And then... AI showed up. Publicly. Loudly. In every earnings call.
Companies figured out something very useful: if you frame layoffs as "AI-driven transformation," shareholders get excited. The stock pops. The narrative shifts from "we hired too many people at inflated salaries during a low-rate era" to "we're leaning into the future."
Those are very different stories. Only one of them moves a stock price.
It was the perfect excuse to reduce headcount while being rewarded for it by the market. That's a business decision. It's not a signal about what AI is doing to the workforce at large.
Entry-level is getting squeezed. But this isn't new.
Yes, junior and entry-level roles are taking more pressure right now. AI handles simple tasks well (if you just need a button on a page, that's exactly what these tools are built for). Companies that hired aggressively during the tech boom hired a lot of people to fill basic-level task work. Now those roles are easier to compress.
But here's the thing: senior-over-junior preference isn't new. That tension existed long before anyone heard of ChatGPT. What changed is the economics got sharper and the justification got easier.
The people most at risk aren't people with real, growing skills. They're the ones who were hired to do the simplest slice of the work and haven't moved past it. That's a different problem than "AI is coming for everyone."
What's actually changing about how we work
The more accurate framing isn't "AI is replacing jobs." It's that AI is reshaping what work looks like.
Developers who would have spent a week on a feature are now expected to do it in a day. That's not job loss. That's productivity inflation... and with it, a raise in expectations. You still need the developer. You just need them to do more.
Something else is happening too. Non-technical employees are starting to build their own tools. We're seeing it firsthand: people in accounting spinning up AI to automate their own processes. Nobody asked them to. They just did it because the tools exist and the barriers are low now.
There's a whole category of internal tooling that IT and engineering teams were never going to get around to building for individual stakeholders. Now those stakeholders are building it themselves. That's a net gain for the business, not a threat to anyone's job.
Not sure which AI tools are actually worth your time? The AI tools checklist at Infacto is a practical starting point for figuring out what fits where.
The self-selection story no one's talking about
Did the last 18 months of "software dev is dead" headlines shake people out of computer science programs?
Probably some. But here's the flip side: the ones who stuck with it through all that noise probably saw a path forward for themselves. That's a kind of self-selection bias that tends to work in a field's favor. The graduates coming out now are the ones who were doing it because they wanted to... not just because CS is the highest-paid individual contributor role in most org charts.
Two tracks are forming: the deep computer science people who will be advancing models and systems, and the more practical practitioners who can install open-source tools and build solutions. Both are valuable. The line between "technical" and "non-technical" is blurring faster than most people realize, and that creates opportunity if you're paying attention.
What the rebound looks like
Jackson called his shot on this episode: he took a prediction market trade that 2026 won't have more tech layoffs than 2025. He thinks the cycle peaked in January and we're on the way back up.
The evidence supports that read. Companies are already rehiring roles they cut last year... often the same people, because those people know the product and it's cheaper to bring them back than to train someone new.
We saw it with Elon and X. We saw versions of it after some of the government cuts. The pattern is: cut aggressively, feel the constraints, bring back what you actually needed.
AI is also stabilizing at the user level. The jaw-dropping announcements are getting more technical (smaller chips, open-source model forks, on-premise infrastructure). The average user's experience of AI tools is going to normalize this year... which means the chaos period is winding down and the "this is just how we work now" period is beginning.
If you want to make sure you're on the right side of that shift, the strategy diagnosis quiz at Infacto can help you figure out where the real leverage is in your work or business right now.
What to do with all of this
The panic arrived early. The data is catching up slowly, and that's actually the good news.
AI is changing work. That part is real. But it's happening more gradually than the headlines suggest across most industries outside of software. The companies using "AI" to explain their layoffs are mostly cleaning up a hiring mess they made in 2021 and 2022.
Your job probably isn't quietly disappearing while everyone pretends it's fine.
But the landscape is shifting. The best thing you can do is learn what work looks like in it now... and build skills that actually work in that world.
The people who will be fine are the ones who started doing that already. It's not too late to be one of them.