Everyone's training AI to refuse. We post-trained a model to break in. Anthropic and OpenAI's publicly available models are explicitly guard-railed so that they refuse offensive tasks.
Source: [Hacker News](https://www.argusred.com/cli)
📰 Read Full Story
This is an aggregated headline summary. For the complete report, visit the original publisher.
Continue Reading at Hacker News ↗