Snap sent the email at 9 a.m. Within an hour, about 1,000 people no longer worked there. Evan Spiegel, in a companywide memo, called it a "crucible moment." The stock went up 7%.

That is the entire story in three sentences. Everything else is dressing.

The cut is 16% of the global workforce. Add the 300-plus open roles Snap just closed and you are looking at roughly a quarter of the company's planned headcount erased in a single morning. Spiegel told staff that artificial intelligence offers "a new way of working" that lets smaller teams do what bigger ones used to. In a more revealing line, he said AI is already generating more than 65% of Snap's new code.

Think about what that means. Two out of every three commits shipping at Snap right now were written by a model. The humans who used to do that work are, in many cases, the ones packing up today. The company expects to save more than $500 million a year by the second half of 2026. Severance is four months plus healthcare and career transition support, which is above industry standard and does not undo anyone's mortgage.

Wall Street loved it. Shares jumped on the open, because that is what markets do when management finally cuts. Every other consumer tech company with soft ad revenue and a bloated headcount is watching this tape. Meta did a version of it. Google did a version of it. Salesforce did a version of it. The playbook is now explicit: ship AI into the toolchain, measure code output, cut the people whose work the model duplicates, and tell investors you are "AI-native."

Snap's total headcount was around 6,400 before this morning. Cutting 1,000 takes them closer to 5,400, which is roughly where they were in 2019, the last time the company was considered small and scrappy. Management is not trying to hide that trajectory. The press release describes flatter teams, fewer layers, and faster decision-making — the same three phrases every AI-era layoff announcement uses. What is different this time is that the efficiency number has a decimal place. Sixty-five percent of code written by a machine is not a vision. It is a current measurement.

And it's worth naming who actually takes the hit here. Snap employs disproportionately young engineers in Los Angeles and New York on visas or first-job contracts. A $500 million cost-saving number in a press release is 1,000 individual stories about rent, parents, spouses with no backup income, and the weird humiliation of explaining to your LinkedIn network that you are "exploring new opportunities" the same week your employer's stock had its best day of the year.

I'll be blunt. "Crucible moment" is CEO-speak for "we finally had cover to do the layoff we wanted to do anyway." Snap has been underwater on ad revenue for years. The company has missed earnings more times than I can count. If AI hadn't given Spiegel a narrative, this layoff would have been framed as a restructuring, a reorg, a focus on core verticals. Instead it's a productivity story, and the stock rewards him for it.

Here's what actually bugs me. Every CEO using the AI excuse this quarter is doing two things at once. They are telling investors: our product is so good we need fewer people. And they are telling the public: don't worry, AI won't take your job, it will augment you. Both things cannot be true. Snap just demonstrated which one these executives actually believe. The memo did not say "we are augmenting 1,000 people." It said those 1,000 people are gone.

The 65% code-generation number is the part nobody wants to sit with. If that figure is real, and is sustained, and spreads, then software engineering as a career category compresses faster than any previous white-collar profession has compressed. Not in ten years. In two. The next company to do this won't need a crucible moment to justify it. The playbook will just be the playbook.

Author: Yahor Kamarou (Mark) / www.humai.blog / 18 Apr 2026