OpenAI取消了最新模型的发布，Anthropic据称强调了其技术带来的生存性风险；在一次重要的白宫会议前夕，安全担忧进一步升级。两家公司的负责人将与美国总统唐纳德·特朗普，以及Meta和Nvidia的掌门人共进午餐，AI监管将成为首要议题。
OpenAI scrapped its latest model release and Anthropic reportedly highlighted its technology’s existential risks, ratcheting up safety concerns ahead of a major White House meeting. AI regulation will top the agenda when leaders from the two firms lunch with US President Donald Trump alongside the bosses of Meta and Nvidia.
然而，AI实验室的最新举措——OpenAI称此次更新未达到安全门槛，Anthropic则在上市前文件中提出相关担忧——似乎不太可能改变白宫的看法：特朗普拒绝接受对AI加以限制，认为这只会帮助中国。此外，顶尖研究人员在最新一份白皮书中警告，随着更多突破即将到来，一场“智能爆炸”可能很快发生。——普拉尚特·拉奥
The latest moves by the AI labs — OpenAI said its update did not meet safety thresholds, and Anthropic flagged concerns in pre-IPO documents — appear unlikely to change views in the White House, however: Trump has refused to countenance AI curbs, arguing they would only help China. And further advances appear on the horizon, as leading researchers warned in a new white paper of an impending “intelligence explosion.”— Prashant Rao