Meta is cutting 8,000 people on May 20 and explicitly blaming AI. Microsoft, Atlassian, Salesforce, Oracle, Block, Dell — same playbook, same press release, same line about "becoming a more AI-native organization."
Meanwhile, IBM tripled its entry-level hiring this year.
Same AI. Same labor market. Same quarterly earnings pressure. Two opposite bets. One of these companies is going to look very stupid in 24 months.
Read the Tape
Q1 2026 tech layoffs hit nearly 80,000 people. Roughly 47% were directly attributed to AI. The narrative writes itself: AI is here, juniors are obsolete, the leverage is in senior operators and the model itself.
That narrative is also wrong, or at least incomplete, and IBM's CEO said the part nobody else will say out loud: AI can do a lot of entry-level work. It still needs a human touch. So they tripled the juniors.
This is not sentimentality. IBM is not running a charity. They've been more aggressive about automation than almost any of their peers — they were quietly doing internal HR-replacement pilots when Meta was still arguing about the metaverse. If IBM is hiring more juniors, it's because their internal data is telling them something the layoff press releases are not.
What the Layoff Crowd is Missing
Here's the bet you make when you cut entry-level headcount and assume AI will fill the gap:
- The model can do 80% of the junior task. ✓ Mostly true today.
- A senior person can drive the model and replace the junior entirely. ✓ Sometimes true.
- That senior person will still be there in three years to do that. ❌ This is where it falls apart.
The bench you don't hire today is the senior bench you don't have in 2029. Every senior operator in your company started as a junior somewhere. AI didn't change that pipeline — it just made it look optional for 18 months.
If you cut the bottom of the funnel, the math works for two budget cycles. Then it doesn't.
The Other Thing IBM Sees
There's a second thing happening at IBM that the layoff playbook misses: a junior with AI is now a mid-level.
The traditional ratio was one senior to manage three or four juniors. With AI in the loop, a single junior can ship the output of what used to take three. So the right move isn't to fire the juniors — it's to hire more of them, give them the AI tools, and reduce the senior-to-junior leverage ratio.
That's a totally different operating model than "fire everyone, hire one expensive ML engineer." It's also the one that scales, because your real bottleneck is rarely senior IC throughput — it's how fast you can put problem-solvers in front of problems.
IBM is hiring juniors because, with AI, juniors are cheaper and more productive than they've been in a generation. Meta is firing them because the spreadsheet from McKinsey said to. Different bets, different time horizons, very different outcomes.
What This Means For Your Company
You're probably not Meta. You're probably not IBM either. You're probably running a 50-to-5,000-person company where this debate gets resolved one hiring decision at a time. So:
If your AI program plan assumed "we won't backfill the next attrition," ask why. It's the easy answer, but it's almost always the wrong one. The right answer is usually "we backfill, but we hire someone who can do 2x with AI in the loop."
Your junior hiring profile should change before your headcount does. The kid you hire in 2026 is not the kid you hired in 2022. They need to show up on day one knowing how to use the tools — Claude, Cursor, the agent framework you're standardizing on. If your interview loop isn't testing for AI-fluency at the junior level, you're hiring 2022 talent for a 2026 job.
Cap-out on consultants and start building the bench. Right now most companies are paying outside firms 3-5x what an in-house operator costs because they cut the in-house team that would have grown into that role. Bring it back inside. Your 2028 senior architect is somebody you should hire as a senior associate this quarter.
The pipeline is the leverage. Companies that win the next decade will not be the ones who fired the most. They'll be the ones who restructured the funnel — fewer people, more leverage per person, and a deliberate path from junior to senior that runs through AI tooling, not around it.
The Uncomfortable Question
Look at your own org chart. If you're cutting juniors right now because AI lets you, ask the question Meta isn't asking: what does my talent pipeline look like in 2029, and who's running my company in 2032?
If the honest answer is "we'll figure it out then" — that's exactly how IBM is going to eat your lunch.
VC5 Consulting builds recruiting engines and AI systems for growing companies. We help you hire the bench, not just the heroes.