On June 29, a report landed that ruins a good doomscroll. Companies that spend the
most on AI are not firing people. They are hiring them.
A
joint report from Ramp and Revelio Labs, which between them track enterprise AI spend and the workforce records of nearly 22,000 companies, found the heaviest AI spenders grew headcount 10.2%. The label they used is “high-intensity adopters,” meaning firms dropping about $30 per employee per month on AI in their first three months. Those are the companies supposedly automating everyone into the unemployment line. They added the most people.
The part that should stop the scroll cold: entry-level headcount at tech-forward firms rose 12%. The exact roles everyone swears are first to die grew faster than the average. The machine that was supposed to eat the junior analyst is sitting at a company that just hired three more of them.
Now the honest asterisk, because we don’t do cheerleading here.
That data skews. It leans hard toward VC-backed, tech-forward, knowledge-work firms that were already growing like weeds. When a fast-growing software company spends big on AI and also hires like crazy, it’s genuinely hard to say the AI caused the hiring or just showed up to the party. The report’s own authors say it plainly: the paper does not show AI universally creates jobs, but it does counter claims that AI will lead to broad job losses.
And the counter-number is real. Goldman Sachs found AI has erased roughly 16,000 net jobs a month over the past year, with early-career workers taking the worst of it. Both things are true at once. AI is cutting jobs in some places and fueling hiring sprees in others. The story was never “AI creates jobs” or “AI destroys jobs.” It was always which room you’re standing in.
The companies that swung too hard are quietly swinging back
Here is the plot twist nobody put in the recruiting deck. Ford just
rehired 350 veteran engineers, the kind the company reportedly calls “gray beards,” after its automated quality systems produced disappointing results. Some were former employees. Some got poached back from suppliers. Their whole job now is hunting for failure points before a bad part ever reaches the plant floor, the exact instinct the AI didn’t have.
Ford’s VP of vehicle hardware engineering, Charles Poon, said the quiet part into a microphone: “Mistakenly we thought that by just introducing artificial intelligence and ingesting the design requirements that we had, that that would produce a high-quality product.”
Translated from corporate: we automated the judgment, the judgment turned out to matter, and now we’re paying to bring it back. That is a company discovering the value of experience the expensive way, in public, in a Bloomberg story.
Meanwhile, the robots are getting genuinely good
None of this means the pressure is fake. A Swiss startup called Flexion Robotics, founded by ex-Nvidia researchers, just showed off a system that teaches humanoid robots simple skills in simulation, opening doors, climbing stairs, carrying boxes, then uses a master AI to string those skills together on the fly. Most robot demos are puppet shows with a human working the strings offscreen. Flexion’s pitch is that its robots figure out
unfamiliar tasks on their own. Point one at a snack delivery and it just goes.
So the fear has a face, and it’s a modified humanoid competently walking a parcel across an office. The instinct to panic is not crazy. But look at what the panic is producing.
Some people are fleeing the desk entirely. A British writer just spent a day retraining as a bricklayer because an experienced self-employed brickie can clear £90,000 a year and no large language model is going to lay a straight course of brick anytime soon. Trade schools report people in their 50s and 60s showing up specifically because they feel an AI clock ticking on their office job. That’s a rational hedge. It’s also a full career U-turn triggered by a trend line, not by anyone actually losing their job yet.
Here’s what the bricklayer and Ford stories have in common. Both are the same realization arriving from opposite directions: the human judgment, the physical skill, the “I’ve seen this go wrong before” instinct, that’s the part AI is worst at and the part that keeps getting more valuable. Ford is buying it back. The career-switchers are running toward it. You don’t have to pick up a trowel to act on that.
