Silicon Valley wants you to clap for artificial intelligence. It also wants you not to notice the pink slips piling up behind the curtain.
That’s the central contradiction flagged by Jean-Pierre Robin, a columnist atLe Figaro: the same executives selling AI as the future are now stuck managing the job cuts that future is already producing. And the tension isn’t theoretical, it’s showing up in real headcount, real careers, and a labor market that’s getting reshuffled in plain sight heading into 2025.
AI is replacing work people swore “couldn’t be automated”
We’re past the era when automation meant robots on factory floors. The new wave is software eating white-collar tasks, basic coding, customer support, routine data analysis, work that used to be a safe bet for college grads and mid-career professionals.
And then came large language models, tools like ChatGPT that put “good enough” writing, summarizing, troubleshooting, and drafting into the hands of anyone with Wi‑Fi. That’s the accelerant. Companies don’t need a science-fiction breakthrough to cut roles; they just need a spreadsheet that shows a team of 12 can become a team of 7 with the right AI workflow.
The result: AI isn’t just changing how work gets done. It’s changing how many people get to do it.
Silicon Valley’s awkward two-step: build faster, lay off quieter
Big Tech is trapped between two forces that don’t play nice together.
First: the arms race. If U.S. firms slow down, they risk losing ground to international competitors, Robin points directly to China as the pressure point. So executives pour money into AI to stay ahead, because nobody gets a bonus for “responsible pacing.”
Second: the blowback. Layoffs aren’t just a PR headache anymore. They’re political fuel. When high-profile companies cut thousands of jobs while bragging about productivity gains from AI, it doesn’t exactly scream “shared prosperity.”
This isn’t only about quarterly earnings. It’s about what role these companies think they play in American life, just profit machines, or institutions with some obligation to the communities and workers they’ve relied on for decades.
New jobs, sure, but not necessarily for the people getting pushed out
Yes, AI creates jobs. You need people to build the systems, maintain them, monitor them, and keep them from doing dumb, or dangerous, things. Roles tied to AI development, oversight, and operations are growing.
But here’s the part executives tend to mumble: those new jobs don’t automatically match the old ones. The worker who gets displaced from a support role or a junior programming track doesn’t instantly become an AI auditor or model engineer. The skills gap is real, and the timeline is brutal.
Companies are experimenting with retraining, internal programs, university partnerships, funding career switches. Robin’s point is that it’s still not enough relative to the scale of the shift. Training a few thousand workers sounds generous until you realize how many roles can be compressed when AI becomes the default assistant across departments.
And that’s where this stops being a “tech story” and starts being an American social contract story: if the winners keep winning and the displaced get a Coursera link and a pat on the head, the backlash won’t be subtle.
