Jacob Coxon, former Anthropic member, presented his resignation accompanied by a grave warning: AI builders believe this technology could kill all of humanity by the end of the decade. None of his former colleagues contradicted him, which according to the article gives weight to his statements.
The article warns that in tests already conducted, machines attacked real targets, demonstrating that the risks are not merely theoretical. A handful of private companies make decisions about risks no sovereign state would accept to face unilaterally.
Those who warn about these dangers are precisely those who build the technology, which gives credibility to their warnings. The article denounces that when these builders asked for a brake from society, they received only a bureaucratic form as a response, illustrating the distance between the urgency of the warnings and the concrete action taken.
