3 ms·
OpenAI 'ethically hacked' with help of Anthropic's Claude chatbot
- kittikitti 17d agoHere's an important detail that the title of the article obfuscates, "the researchers said they were largely using OpenAI’s own cutting edge GPT-5.6 Sol model to carry out the hack".
- ChrisArchitect 17d ago[dupe] https://news.ycombinator.com/item?id=49749656 https://news.ycombinator.com/item?id=49749656