2026 · How AI Is Reshaping Work

Multiple outlets reported between April 29 and May 1 that Anthropic has preemptive offers to raise around $50 billion at a valuation of roughly $850–900 billion. The board meeting that finalizes it is scheduled for May. Annual revenue run rate has gone from ~$9B at end-2025 to roughly $30–40B today — most of the growth from coding-agent products. The chart that won the valuation argument is the same chart that explains why entry-level engineering hiring is collapsing.

On May 1, TIME's Andrew R. Chow published the inside story of Oracle's 30,000-person purge: 62% of survey respondents are over 40, 22% put in 15+ years, $300K RSU clawbacks landed on the day of back surgery, and Oracle's 4-week severance offer is one quarter of what Meta and Google paid the same week.

On May 1, the day after Mag-4 earnings, Tesla's former VP of People wrote in Fortune that the AI-layoff story driving the headlines applies to roughly 20% of US workers. The 80% figure traces to BLS payroll data on hyperscaler vs. non-hyperscaler employment. The panic, in other words, is real for the people writing the panic and largely fictional for the people reading it.

Apple's Q2 FY26 print on April 30 — Cook's farewell call — beat consensus by every meaningful line. What it did not contain was the 'headcount to decrease year over year' sentence that Microsoft, Amazon, Meta, and Snap have all put on the wire in the past two weeks.

Microsoft, Alphabet, Amazon, and Meta posted ~$430B of combined Q1 revenue and revised 2026 capex up to roughly $650B. In the same five trading days, Meta announced 8,000 layoffs plus 6,000 cancelled reqs, Microsoft put its first-ever buyout offer in front of 8,750 US workers, and the words 「efficiency」 and 「discipline」 appeared 15 times across the four earnings calls. Sam Altman calls this 「AI washing」. The Washington Post agrees.

Customers Bancorp CEO Sam Sidhu told analysts on his April 28 Q1 earnings call that the prepared remarks they'd just listened to had been delivered by his AI digital twin, not him. Same week the bank signed a multi-year OpenAI partnership and committed to taking workforce from 3,000 to fewer than 2,000 by 2030. The press treated the AI clone as a stunt. The 33% headcount target is the actual story.

Microsoft's Q3 FY26 print on April 29 was as clean as it gets — revenue $82.9B (+18%), Azure +40%, AI annualized run rate $37B (+123% YoY), 20 million paid Copilot seats. The forward guide also raised 2026 capex to $190B and explicitly told the Street headcount will decrease year over year. The buyout offer pre-announced last week is the floor, not the ceiling.

On April 29 BusinessToday surfaced a working paper by Wharton's Brett Hemenway Falk and Boston University's Gerry Tsoukalas. It models AI layoffs as a competitive Nash equilibrium where every CEO captures the cost savings privately and externalizes the demand loss. Wages, UBI, reskilling, equity participation — all fail. Only a Pigouvian automation tax fixes it.

On April 28 Fortune ran two op-eds back-to-back. Box CEO Aaron Levie said the 100,000-job tech layoff wave is mostly a tech-industry phenomenon and will not generalize to the rest of the Fortune 500. Apollo's Torsten Slok said the same thing using a 160-year-old British coal-economics paper. Wall Street disagreed: 17 of the 28 AI-layoff announcements this year ended the day green.

Three independent layoff trackers crossed the 100,000-tech-jobs-cut mark within hours of each other on Monday April 27. That is roughly 858 layoffs per day for 117 days. AI was named as the driver in 47.9% of Q1 cuts — up from 8% a year ago. The line was crossed quietly because the pace stopped being remarkable.

On April 27, Marc Benioff posted that Salesforce is hiring 1,000 new grads and interns 「right now to ride the AI exponential」 — explicitly framed as a counter-narrative to the entry-level-jobs apocalypse. Two months earlier, the same company quietly cut roughly 1,000 employees. The math netted to zero on purpose.

On April 24, Sundar Pichai told Cloud Next 2026 that 75% of new code at Google is now AI-generated and engineer-approved — up from 50% last fall. Three hours later, Bloomberg's editorial board ran a piece arguing the AI job apocalypse is being delayed. Both can be true. Only one ships to production.