OpenAI CEO Agrees With AI Safety Concerns But Criticizes Pause Letter
Sam Altman states he agrees with the need for AI safety precautions but argues the recent open letter calling for a six-month research halt lacks necessary technical nuance.
OpenAI CEO Sam Altman expresses partial agreement with an open letter signed by Elon Musk and Steve Wozniak that calls for a six-month pause on advanced AI research. However, Altman argues during an MIT event that the letter misses crucial technical nuances about exactly where a pause is necessary. He emphasizes that while caution is important, the letter does not represent the optimal way to address these complex safety issues.
The open letter, which gathers over 25,000 signatures since March, demands a halt to training AI models more powerful than GPT-4. Signatories urge AI labs to use this timeout to develop shared safety protocols overseen by independent experts. Altman confirms his belief that the safety bar must increase as AI capabilities become more serious, but he notes that the letter completely leaves out other projects OpenAI is developing on top of GPT-4 that present important safety concerns.
Questions surrounding safe and ethical AI use currently dominate discussions at the White House, on Capitol Hill, and in corporate boardrooms across America. Altman openly acknowledges that AI technology makes him a little bit scared, reflecting a broader industry anxiety. The rapid deployment of ChatGPT, along with competing chatbots like Microsoft's Bing and Google's Bard, continues to drive a massive investment frenzy and an international AI arms race.