AI researchers warn superintelligence poses existential risk

Geoffrey Irving, a former employee of OpenAI and Google DeepMind, stated that the chance of human extinction from artificial intelligence is about a coin flip. This assessment forms part of a broader warning issued by a group of AI researchers who described superintelligence as being exactly as dangerous as it sounds.

Palisade Research, a non-profit organization focused on studying AI capabilities and motivations, compiled and released a dozen interviews with these experts on the platform frominside.ai. The participants include current and former staff from OpenAI, Google, and Anthropic. Several of the interviewed researchers echoed Irving's concerns regarding the potential for AI to drive humanity to extinction.

Neel Nanda, a research scientist at Google DeepMind, provided a more specific probability estimate within the collection. He stated that there is at least a ten percent chance that artificial intelligence causes human extinction. The interviews serve as a documented record of these internal perspectives from key players in the field.