OpenAI is testing another new way to expose the complicated processes at work inside large language models. Researchers at the company can make an LLM produce what they call a confession, in which the model explains how it carried out a task and (most of the time) owns up to any bad behavior. Figuring out…
Daniel Liss, co-founder of the social network Dispo and the dating app Teaser AI, is convinced he’s onto the next big thing: steelmaking. It all started, incongruously, with a few op-eds he wrote for TechCrunch about anti-trust enforcement in social media. The commentaries apparently caught the attention of some folks in Washington D.C., Liss told […]