AI researchers warn that superintelligence poses existential risks

A new series of interviews conducted by Palisade Research features insights from former and current employees of major AI firms, including OpenAI, Google DeepMind, and Anthropic. The participants, who are experts in the field, express significant concern regarding the rapid development of artificial intelligence. Notably, former researcher Geoffrey Irving suggested that the probability of human extinction due to superintelligence could be as high as 50 percent. These interviews highlight a growing divide within the tech industry, where internal researchers are increasingly vocal about the potential dangers of their own work. The project aims to shed light on the internal safety culture and the existential risks perceived by those building these systems. As the industry races toward more powerful models, these warnings serve as a critical reminder of the ongoing debate surrounding AI safety, alignment, and the long-term implications of human-level machine intelligence.