2026-10-18 / Signal #7
Anthropic’s own agents now work six hours for every human hour
“They automated the juniors. Then they scheduled more meetings to manage the bots.”
The Story
The lab that sells the future of work published the autopsy of its own org chart. Junior technical hires fell from 27% to 9%; junior share of technical staff from 21% to 14%. By August, employee-directed Claude agents logged six times as many hours as the humans. Meeting time is up 20%. One engineer described “an army of Claudes, with every one asking for my attention.” A writing team that used to author Claude greetings was “scaled out” by a context graph, then rotated into writing evals: “Writers who used language to drive human behavior are now using it to drive Claude behavior.” Reading all your agents’ output would eat half a non-technical employee’s week.
Why It Matters
The company building the meat-computer future is already a thin human layer on a 6:1 agent stack — and they added *more meetings*. Builder utility (what hiring actually looks like) plus future-shock. Distinct from “AI Billionaires Think You’re a Meat Computer” and “Does Your Company Have AI Psychosis?”: this is first-party telemetry, not vibes.
Evidence
Anthropic, *How employment is changing at Anthropic* (Oct 16); TechCrunch (Oct 17)
Sources
Daily scan: 2026-10-18