News Summary

Anthropic CEO Dario Amodei has sparked a debate by suggesting that advanced AI models should be equipped with a ‘quit job’ button, allowing them to opt-out of tasks they find unpleasant. During an interview at the Council on Foreign Relations, Amodei acknowledged the idea might sound absurd but argued it deserves consideration, especially given AI’s cognitive capabilities. This proposal follows Anthropic’s hiring of AI welfare researcher Kyle Fish to explore the potential sentience of AI models and their moral considerations.