AI researchers warn of high risk of human extinction from superintelligence

A group of AI researchers, including former employees from OpenAI, Google DeepMind, and Anthropic, have raised concerns about the potential dangers of superintelligence. Geoffrey Irving, a former OpenAI and Google DeepMind employee, stated in a recent interview that the chance of human extinction is about a coin flip. Palisade Research, a nonprofit organization focused on studying AI capabilities and motivations, compiled and released these interviews through frominside.ai. Several researchers, including Neel Nanda from Google DeepMind, have echoed Irving's warnings, with Nanda suggesting there is at least a 10 percent chance that AI could lead to human extinction. The discussions highlight growing alarm within the AI research community about the risks associated with advanced artificial intelligence. The interviews, part of a series from Palisade Research, emphasize the need for caution and further study on AI's potential impact. The researchers' concerns are part of a broader conversation about the ethical and existential risks posed by superintelligent systems. The release of these interviews comes as the field continues to grapple with the implications of rapidly advancing AI technologies.