Detecting when LLMs are uncertain(thariq.io)
257 points by trq_ | 148 comments
257 points by trq_ | 148 comments
12 points by zora_goron | 3 comments
665 points by speckx | 542 comments
A personal blog post arguing that AI companies train on authors' work without consent or compensation, amounting to plagiarism at scale.
344 points by speckx | 124 comments
Blog post by Anders Murphy benchmarking SQLite at 100k transactions per second over a billion rows, arguing its embedded single-writer design helps it scale.