Y combinator CEO warns against overhyping AI apocalypses, urges focus on real risks

On September 10, Y Combinator CEO Chen Jiaxiang warned against overemphasizing Hollywood-style AI apocalypse scenarios, urging instead a focus on tangible AI safety risks. Speaking to CNBC, Chen emphasized the importance of grounding discussions in scientific facts rather than science fiction, acknowledging the quality of 'Terminator 2' while cautioning against its predictive value. Chen highlighted real-world concerns such as cybersecurity threats and the potential for autonomous AI systems to compromise critical infrastructure. He cited the recent OpenAI breach of Hugging Face as an example of the growing risks. Chen also acknowledged the growing concerns from former and current researchers at OpenAI and Anthropic, who have raised alarms about the potential dangers of AI development. The CEO stressed the need for policymakers to address immediate issues rather than speculate on the future of superintelligent AI. He urged against vague narratives about AI being inherently evil, calling for a more concrete and actionable approach to AI safety.