Reporting archive
Evidence on AI and work
-
Agents go enterprise, payrolls thin, markets price ghost GDP
A viral pre‑mortem coined “ghost GDP”—productivity without paychecks—and markets priced the future like it had already happened.
-
Lightcast’s Fault Lines meet generative assistants as apprenticeships vanish
As AI devours the on‑ramp work and retirees take the tacit knowledge, scarcity climbs the ladder—unless companies start manufacturing judgment on purpose.
-
Georgieva’s 0.8-point promise, fewer first jobs
AI’s 0.8-point growth bump comes with a catch: entry-level on-ramps are narrowing unless policy forces augmentation over substitution.
-
January 2026 saw 108,435 layoffs and seven percent cited AI
Altman just split the layoff narrative: stop laundering cuts as “AI” without receipts—and start preparing for the automation curve that’s finally on the clock.
-
Pew says two-thirds expect losses; workers rewrite AI rules
Workers are quietly turning AI anxiety into leverage, bargaining over dashboards and data instead of waiting for the next model demo.
-
The Fed stress tests a jobless boom from AI
When a Fed governor starts gaming out a ‘jobless boom,’ AI stops being hype and becomes monetary policy.
-
Anthropic’s agents hit legal-tech moats, not headcount
AI isn’t firing lawyers—it’s draining the moats built on paperwork and shifting power to whoever owns verification and outcomes.
-
From Micron wafers to ₹65 compute, India hires
India is trying to turn AI anxiety into payroll by fabbing memory at home and pricing compute like electricity at ₹65 per GPU-hour.
-
Safety walkouts signal agents optimized to sell your judgment
The AI safety walkouts were a flare: your team is about to rent its judgment to monetized agents unless you lock in enforceable guardrails now.
-
IBM triples entry-level hiring, turns juniors into AI stewards
IBM is tripling U.S. entry-level hiring by turning junior roles into human-in-the-loop arbiters who decide when to trust, doubt, and override AI.