Former Google DeepMind Researcher Warns AI Could Potentially Kill Humanity

A former Google DeepMind researcher who recently left the company after studying AI safety has warned that continuing with current AI development trends could pose a severe threat to humanity. According to Bloomberg on April 14 (local time), Bilal Chughtai, a former Google DeepMind researcher, stated on X and LinkedIn that he witnessed AI development firsthand at Google and is deeply concerned about the direction the technology is heading, adding that he 'truly believes AI has the potential to kill us all'.

Chughtai, who previously researched AI safety and alignment at DeepMind, still believes there's a possibility of developing AI safely. He now works on AI safety-related tasks at the nonprofit organization BlueDot Impact. Recent warnings from within the AI industry have been mounting, with Jacob Cokson, formerly of Anthropic and OpenAI, criticizing companies for racing to develop superintelligent AI 'gambling with our lives' as he left his positions earlier this month. Dario Amodei, CEO of Anthropic, also argued last weekend that development speed should be slowed to reduce the risk of advanced AI escaping human control, a view supported by OpenAI CEO Sam Altman and Elon Musk.