Tao Lin, who works on safeguards for Anthropic’s Claude models, said in a Palisade Research interview that “pretty much all humans on earth will become permanently unemployable”, in his view (Palisade Research video).
He gave timelines of “something like 1.3 years” to broadly superhuman AI, but said no when asked if near-universal unemployment comes “within the next 15 months”. The clip spread on Anthropic’s IPO day.
Business Pill · Guardrails Explained
An evergreen Business Engineer explainer on AI guardrails. It teaches the general idea only and does not discuss Lin or Anthropic.
The key insight: Lin, an Anthropic safeguards researcher, said in a Palisade interview he expects broadly superhuman AI in about 1.3 years and eventual near-universal unemployment, but not within 15 months; Anthropic’s January 2026 Economic Index showed 52% augmentation vs 45% automation.
What Lin Said
Lin introduced himself in the video: “I work at Anthropic. I’ve worked there for about a year and a half”. He said he now works on safeguards, which blocks requests such as help making bioweapons.
Asked what people most need to understand, he said: “superhuman AI is going to come here very quickly”. He added: “I have timelines of something like 1.3 years to getting broadly superhuman AI”.
On jobs, he described a sequence: “everyone losing their current jobs”, moving “to a different occupation briefly”, then being “fully unemployable and needing to rely on the state”.
Asked if that is very likely, he said “pretty much all humans on earth” “will need to strictly rely on government welfare to sustain themselves”, per our transcript of the video.

Why He Does Not Expect It Within 15 Months
Asked whether near-universal unemployment comes “within the next 15 months”, Lin said no. He said he expects “a large gap” between AI being able to replace people and actually doing so.
His reason was physical: “there aren’t enough robots yet”. He said the world would first need to get good at building robots and then build “billions of robots”.
What He Wants AI Companies to Do
Lin said “AI companies should dramatically slow down their rate of research” to “something like one quarter of its current rate”. Asked if people should be angry at him, he said: “people should be angry at me and everyone else developing AI”.
We found no Anthropic statement on the interview, and Lin did not cite company data for his forecasts in the excerpts we transcribed.
What Anthropic’s Own Data Shows Today
Anthropic’s Economic Index report of 15 January 2026 measured how people use Claude. “The share of conversations classified as augmented jumped 5pp to 52%”, it said (Anthropic).
The same report said “the share deemed automated fell 4pp to 45%”, on samples from 13 to 20 November 2025. That is a snapshot of usage, not a forecast of jobs.
Why It Matters on Debut Day
The interview was published on 15 November. It spread widely on 19 November, the day Anthropic began trading after pricing its IPO at $83 (our live blog).
It joins early investor Dustin Moskovitz telling the New York Times labs should slow down (our piece), on the same day public investors bought in.
The Structural Read
As we read the full transcript, the viral line leaves out his ‘No’ on a 15-month horizon; his case rests on robots, not software alone.
As we read the two together, Lin’s forecast is about what AI could do; Anthropic’s Economic Index is about how people use Claude today.
Tao Lin, Anthropic safeguards researcher, to Palisade Research
“I think that it’s very likely that pretty much all humans on earth will become permanently unemployable.”
Three Implications
FOR ANTHROPIC A staff researcher’s public forecast circulated widely on its first trading day.
FOR WORKERS Lin expects a large gap between AI capability and job replacement, tied to building robots.
FOR THE INDUSTRY He called for research to slow to about a quarter of its current rate.
What Is Not Established
Lin spoke for himself in a video by Palisade Research, which says it wants to help the world understand AI. He did not present his forecasts as Anthropic’s position, and the quotes are from our machine transcript.
The Bottom Line
An Anthropic safeguards researcher expects broadly superhuman AI in about 1.3 years and, eventually, near-universal unemployment, but not within 15 months. Anthropic’s January 2026 Economic Index showed more augmentation (52%) than automation (45%).
97,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. Lin’s words are from our machine transcript of Palisade Research’s published 63-minute video; we checked each quote against the audio timestamps. Usage shares are from Anthropic’s Economic Index. Lin’s forecasts are his own views. We found no Anthropic statement on the interview.
Sources: Palisade Research, interview with Tao Lin (YouTube), 15 Nov 2026 · Anthropic Economic Index, January 2026 report · unusual_whales on X, 19 Nov 2026 · Anthropic pricing free writing prospectus, 18 Nov 2026









