The guys building AI think it might kill us all, and one of them just walked out and said it in plain English. Jacob Coxon spent three years doing pretraining at OpenAI and Anthropic. Helped build GPT-4o. Eight months ago he joined Anthropic because it was supposed to be the safe one. Today he quit the whole industry and told the WSJ that by the end of next year things could already be out of control. This is a guy who trained the actual models telling you what people inside these labs…