AI scientist

A new discussion is circulating on whether artificial intelligence has cracked one of mathematics' most difficult open problems. The claim is being debated by scientists and enthusiasts, with Neil deGrasse Tyson's StarTalk programme weighing in on what the reported breakthrough would mean for maths and for the future of AI research. Details of the specific problem and the claimed proof remain limited in the coverage.

Anthropic has published a research essay titled 'Claude-Shaped Science', exploring how its AI assistant Claude is influencing the way scientific work is conducted and shaped. The piece is drawing attention on Hacker News, where readers are debating what it means for research when AI tools increasingly mediate how scientists write, think and run their work.

The Bulletin of the Atomic Scientists argues that apocalyptic claims about artificial intelligence wiping out humanity rest on weak science, and that the companies spreading these warnings benefit from them. By framing existential risk as the main AI danger, the argument goes, firms divert attention from present harms, regulation, and accountability, while positioning themselves as the only ones capable of building safe AI.

The Bulletin of the Atomic Scientists, long known for its Doomsday Clock warning about nuclear risk, has published an argument that AI doomerism amounts to bad science. The piece pushes back on claims that artificial intelligence poses existential threats, framing such fears as unsupported by evidence. It has drawn attention in discussions about whether AI alarmism distracts from more concrete technological and nuclear dangers.