Anthropic researcher Jacob Coxon's warnings about self-improving AI as an existential threat appeared on CNN and Fox News, sparking public debate. His concerns, which have now reached mainstream media, have drawn attention from several US politicians and popular figures like Joe Rogan.
Other safety researchers at Anthropic and OpenAI, including Paul Christiano, supported Coxon's views, warning that without better oversight, humanity could permanently lose control over a superintelligence. Christiano recently joined the OpenAI Foundation board, though without voting rights, to help shape the company's safety practices.
"The idea that AI could wipe out humanity has officially gone mainstream," said Matthias Bastian, a contributor to theDecoder. This statement reflects the growing public awareness and concern about AI safety, despite the lack of concrete evidence for the worst-case scenarios.
The announcement follows a surge in public interest in AI safety, with warnings from researchers gaining traction in both media and political circles. This shift highlights the increasing role of AI safety in public discourse, even as experts remain divided on the likelihood of the most extreme outcomes.
Anthropic did not specify the exact measures being taken to address these concerns, and the full extinction scenario remains an extreme and contested position. It is also unclear whether anything resembling a super AI could emerge from today's technology, leaving the situation open to interpretation.
Source: thedecoder
