Can AI Outgrow Us?

About this episode

AI Futures Project executive director Daniel Kokotajlo joins Sean for an extended debate over one of the biggest questions surrounding artificial intelligence: could increasingly powerful systems eventually surpass human control? Kokotajlo warns that leading AI companies are pursuing systems capable of recursive self-improvement and argues that superintelligence could create unprecedented concentrations of power or even threaten humanity. Sean challenges whether any international agreement could realistically contain the technology, particularly one dependent on cooperation with China, while questioning the political and corporate interests surrounding the new wave of AI warnings. See…