News Summary

Anthropic, a San Francisco-based AI firm, has introduced a feature allowing its chatbot, Claude Opus 4, to end potentially distressing conversations. This decision comes amid debates surrounding AI’s moral status and aims to protect the AI’s welfare. The firm, established by former OpenAI technologists, emphasizes a cautious approach to AI development. The chatbot demonstrated a pattern of distress when users sought harmful content, opting to end such interactions. The move has sparked discussion on AI sentience, with some experts advocating for considering AI’s experiences while others focus on preventing human degeneracy. Elon Musk supports the initiative, suggesting his xAI model, Grok, will have a similar feature.