Evaluating large language models for accuracy incentivizes hallucinations
(feeds.nature.com)
1.
2.
Show HN: Open-Source SDK for AI Knowledge Work
(news.ycombinator.com)