<\/p>\n
Anthropic’s warnings about the rapidly improving pace of AI development have drawn reactions from a range of people, from Donald Trump’s former crypto czar David Sacks to 2012 Republican presidential nominee Mitt Romney.<\/p>\n<\/div>\n
Notably, the company in a blog post earlier in the week warned that AI systems could eventually reach a stage where they are capable of designing their own successors, a concept known as recursive self-improvement.<\/p>\n<\/div>\n
The blog, written by Anthropic co-founder Jack Clark<\/a> and the head of its research institute, Marina Favaro, notes that once AI systems reach this stage, it increases the \u2018risks of humans losing control over AI systems\u2019.<\/p>\n<\/div>\n “If it were possible to effectively slow the development of this technology to give ourselves more time to deal with its immense implications, we think that would likely be a good thing,” the company said.<\/p>\n<\/div>\n Donald Trump’s former adviser David Sacks, in a post on X (formerly Twitter), criticised Anthropic for trying to get its AI lab nationalised.<\/p>\n<\/div>\n “You compare it to nukes\u2026 threaten half of white-collar jobs\u2026 warn recursive self-improvement could end humanity\u2026 then race ahead anyway. In other words, you want the government to save us from\u2026 you,” wrote Sacks.<\/p>\n<\/div>\n Meanwhile, 2012 Republican presidential nominee Mitt Romney wrote on X, \u201cOur highest and most urgent national priority should be AI safeguards. The risks of AI weapons, pathogens, mass unemployment, surveillance, and even extinction must not continue to be largely ignored.\u201d<\/p>\n<\/div>\n Andrew B. Hall, a Stanford professor of political economy who previously advised Meta Platforms, said that Anthropic’s proposal wasn’t \u2018far-fetched\u2019.<\/p>\n<\/div>\n “This would have seemed totally non-credible even pretty recently, but between what we’ve seen with the EO, Glasswing, and OpenAI<\/a>‘s proposal for beefed-up model review yesterday, it no longer seems so far-fetched,” Hall wrote on X.<\/p>\n<\/div>\n Hall added that while he remains sceptical, he is interested in seeing whether the mechanisms Anthropic, Demis Hassabis and others propose can coordinate a possible future slowdown.<\/p>\n<\/div>\nHow is the tech world reacting to Anthropic’s warnings?<\/h2>\n