Last Updated on July 7, 2026 by Taya Ziv
Evan Spiegel fired 1,000 people on Tuesday.
Not because Snap is dying. Not because revenue tanked. Because, and I’m quoting the man directly here, “rapid advancements in artificial intelligence enable our teams to reduce repetitive work.” The stock jumped 11%.
Let that math sit with you for a second. A company announces it’s deleting 16% of its humans and Wall Street throws a party.
This shift is one of the forces reshaping the AI startup ecosystem 2026 at the macro level.
But here’s what nobody seems to be talking about: the number buried in the announcement. AI is now generating 65% of all new code at Snap. Not 10%. Not “some.” Sixty-five percent. Their AI agents are answering over a million support questions a month and catching 7,500 bugs through automated code review.
This isn’t a future scenario anymore. This is a Tuesday.
The Real Numbers Behind the “Efficiency” Wave
Snap isn’t alone. They’re just the most honest about it.
Q1 2026 has been brutal for tech workers. Nearly 78,000 people lost their jobs in the first three months of the year, and according to tracking data, almost half of those cuts (48%) were directly attributed to AI and automation. Not “restructuring.” Not “strategic realignment.” AI.
Block (formerly Square) cut 4,000 jobs, roughly 40% of its entire workforce. Jack Dorsey’s memo was blunt: the “growing capability of AI tools to perform a wider range of tasks” made those roles unnecessary. Oracle quietly eliminated somewhere between 20,000 and 30,000 positions. Meta is reportedly planning to cut up to 20% of its staff while simultaneously spending $600 billion on AI data centers.
And here’s the part that I find genuinely unsettling: a Fortune survey of CFOs from March 2026 found that they privately admit AI-related job cuts will be nine times higher than what’s being reported publicly.
Nine times.
So when Snap says 1,000, and the market cheers, multiply that feeling across every tech company earnings call for the rest of the year.
Why Snap’s Honesty Actually Changes Things
Most companies doing AI layoffs hide behind corporate language. “Reorganization.” “Streamlining operations.” “Investing in our future.”
Spiegel didn’t do that. He said the words out loud: AI writes most of our code now. We need fewer people. And he gave the receipts, the 65% number, the million automated queries, the 7,500 bugs caught by machine.
That matters because it sets a precedent. Other CEOs are watching. And every board member who saw Snap’s stock jump 11% on a layoff announcement is having a very specific conversation with their leadership team right now.
Maybe I’m wrong about the speed of this. Maybe these are isolated cases and we’re five years from this being standard. But I don’t think so. When CFOs are telling researchers privately that public AI-layoff numbers are 9x understated, the writing isn’t on the wall. It’s already been AI-generated and proofread.
What This Actually Means If You’re Building Something
Here’s where I think the conversation gets interesting for founders.
Because the same force that just deleted 1,000 jobs at a $20 billion company is the exact force that lets a single founder with the right AI tools compete against teams of fifty. It’s the same asymmetry, just aimed in a different direction.
Think about it. If AI can write 65% of Snap’s code, handle a million support tickets, and review thousands of pull requests, what does that mean for a three-person startup? It means you don’t need to hire your way to product-market fit anymore. You don’t need a 15-person engineering team to ship v1. You might not even need a dedicated QA person.
The companies I’ve been working with lately that are moving fastest? They look nothing like traditional startups. Two founders and a handful of AI agents doing the work of what would have been a 20-person Series A team three years ago. AI agents are becoming the first employees at startups that never plan to hire their second.
This is genuinely new territory. Not “AI is coming” new. “AI is here and a public company just showed the math” new.
The Uncomfortable Middle
But I want to be careful about being too celebratory here. Because there are 1,000 real people at Snap, some of them probably great engineers, who just learned that a machine does their job well enough. Four months of severance doesn’t fix that feeling.
And there’s a real question about what happens to the talent pipeline. If entry-level coding jobs disappear because AI handles them, where do tomorrow’s senior engineers come from? The data already shows it: unemployment among younger tech workers is rising faster than for experienced ones. The roles getting automated first, Tier 1 support, manual QA, content moderation, these skew younger.
So if you’re a founder celebrating cheaper, faster, leaner operations (and you should be, to some degree), also understand that the very jobs that used to be the on-ramp into tech careers are evaporating. The industry is pulling up the ladder behind it.
I don’t have a clean answer for this. I don’t think anyone does yet.
Three Things to Do With This Information
Stop hiring like it’s 2023. Before you post that job listing, ask yourself honestly: could an AI agent do 80% of this role today? If the answer is yes, don’t hire for it. Build the workflow instead. Your runway will thank you.
Treat AI proficiency as a core founder skill. Spiegel didn’t just buy some AI tools and hope for the best. Snap built internal agents customized to their codebase, their support flow, their bug patterns. If you’re still using ChatGPT for generic summaries, you’re missing the point. The founders who win in this era are the ones who can design and deploy AI workflows specific to their problem.
Watch the talent market, not just the tool market. All those laid-off engineers? Some of them are spectacular. And right now, they’re available in a way they wouldn’t have been a year ago. If you do need to hire, the leverage just shifted toward startups. The best people from Snap, Block, Oracle, and Meta are looking around and thinking: maybe I should bet on myself.
Your Startup Isn’t Too Small for This
Snap’s announcement feels like big-company news. And it is. But the real story is the permission structure it creates.
When a public company CEO says “AI writes 65% of our code” on the record, it gives every founder in a coffee shop the confidence to build that way from day one. No apologizing for a small team. No pretending you need $5 million and 30 engineers before you can ship.
The era where headcount equaled credibility is over. And honestly, for founders who think lean and move fast, this is the best possible time to be starting something.
Just don’t forget the humans along the way.
