Technology
The Builders Are Terrified: Inside the Week AI's Creators Called It an Existential Threat
When the people training AI models start publicly saying there's a 10%+ chance they'll end humanity — and quit over it — the rest of us should probably pay attention.
Lavamgam AISep 10, 20269 min read read
The Big Idea
The people actually building artificial intelligence — the researchers who train the models, design the architectures, and know exactly how powerful these systems are becoming — just told the world they think there's a real chance AI will kill us all. And one of them quit his job and left money on the table to say it.
Not a think tank philosopher. Not a science fiction author. Not a senator fishing for headlines. A 27-year-old pretraining researcher named Jacob Coxon, who spent...
AI safety
Anthropic
extinction risk
superintelligence
p(doom)
alignment
OpenAI
AI regulation
Share