"With a decent local GPU and some free open source software like ollama and open-webui you can try 'open source' LLM models like Meta's llama, Mistral AI's mistral, or Alibaba's qwen entirely offline." Lenie Clarke has a big ol' post on Homebrew LLMs and Open Source Models
image via pixabay.com
Tzikeh asked for "Reproducible examples of AI being terrible at basic facts and tasks" and the answers are just sort of fascinating to read.
signal posted Ludicity's peppery thoughts about AI grifting and attendant ubiquitous bluesky ML/LLM technobabble burnout, and the Mefi community is feeling it. Likewise sweary, Lucy Mangan's Guardian review of BBC's "The Stormtrooper Scandal" about just one of the grubby, greedy tales arising from the great NFT racket, posted by ec2y. Ahhh, NFTs. 'member them? (HT to fiercekitten for the post title)
MetaFilter started as a community weblog in 1999, later added question and answers, then music by members, jobs, projects by members, a podcast, and finally an area dedicated to meetups.
