Research on Measuring AI Welfare

Scientists are developing new methods to assess the 'welfare' of language models, exploring concepts like AI happiness. This research aims to evaluate AI systems beyond performance metrics.

Attorney General Bondi clashes with Blumenthal in Senate hearing

October 08, 2025 由 AI 报道

Attorney General Pam Bondi faced sharp questioning during a Senate Judiciary Committee hearing on October 7, 2025, leading to a heated exchange with Senator Richard Blumenthal over ethics and past controversies. Bondi defended her integrity by referencing Blumenthal's 2010 military service scandal. The confrontation highlighted tensions around her Justice Department leadership.

Global AI Treaties Discussed in Summit

September 15, 2025 由 AI 报道

An international summit has addressed global AI treaties, focusing on ethical standards and digital deception prevention. Participants from various nations discussed frameworks for responsible AI development.

OpenAI's Sora 2 enables AI videos of deceased celebrities

OpenAI's newly launched Sora 2 video generator blocks depictions of living public figures but allows creations featuring deceased ones, leading to widespread social media examples. This loophole has upset families, including Zelda Williams, who pleaded for an end to AI videos of her late father Robin Williams. The company defends the policy, citing consent issues for the dead.