Reporting archive
Evidence on AI and work
-
IBM’s 270,000 workforce tilts toward Red Hat and AI
IBM’s “rebalance” keeps headcount flat while quietly trading legacy roles for AI‑first work to buy margin—and time—on the P&L.
-
Amazon’s 14,000 cuts and the alibi of AI
When a pink slip says AI did it, is that technology at work—or a convenient alibi for old-fashioned cost cutting?
-
LinkedIn just defaulted your profile into its hiring models
LinkedIn quietly flipped default AI training on for public profiles across Europe, Canada, and Hong Kong—tightening the hiring funnel, amplifying lock‑in, and raising hard questions about consent, audits, and bias.
-
At Microsoft, one hire orchestrates many Copilot agents
Nadella just rewrote what a ‘hire’ means at Microsoft: add people only where a human-plus-agents system can deliver more than a pre‑AI peer.
-
Since ChatGPT S&P up 70%, openings down 30%, point to the Fed
The viral ‘scariest chart’ looks like robots versus workers, but the timestamps and sector fingerprints point to tight money—not automation—as the real job killer.
-
Powell flags job creation near zero as AI freezes hiring
Powell just legitimized AI-driven hiring freezes, warning job creation is “pretty close to zero” even as capex roars—reshaping incentives, labor metrics, and the path of rates.
-
Only 2.5% of Remote Labor Index work won acceptance
A benchmark that pays for deliverables found today’s agents can only bill for a sliver of real work—about 2.5%—resetting the automation narrative.
-
AI makes procurement the org designer, CFOs bank gains
AI just turned “do more with less” into headcount policy—and the new org architects sit in procurement.
-
Only 7% tell staff where AI-saved hours should go
AI is gifting teams 1.5 hours a day—and most companies are squandering it in meetings, handoffs, and Slack.
-
Amazon ties 30,000 corporate cuts to AI agents
Amazon just turned AI from a productivity slide into a headcount decision—and every board now has cover.