UK launches AI Safety Institute to set global standards for responsible AI development
• UK's new AI Safety Institute aims to set global standards for AI testing rather than try to test all models itself • Institute assisting UK government and companies like NHS and Home Office on AI safety • CEO Marc Warner warns institute could get overwhelmed trying to vet all new AI models • Instead it should focus on setting testing standards other countries and companies can follow • Last week institute said it will prioritize most advanced AI systems as it lacks capacity to evaluate all models