Tau² benchmark: How a prompt rewrite boosted GPT-5-mini by 22%(quesma.com)

180 points by blndrt 1 year ago | 58 comments

Classifying customer messages with LLMs vs traditional ML(trygloo.com)

237 points by hellovai 3 years ago | 102 comments

Llama 2 is about as factually accurate as GPT-4 for summaries and is 30X cheaper(anyscale.com)

71 points by pallas_athena 3 years ago | 19 comments

UK Government’s ‘AI Skills Hub’ was delivered by PwC for £4.1M(mahadk.com)

329 points by JustSkyfall 9 months ago | 97 comments

Blog post criticizing the UK Government's AI Skills Hub, a site delivered by PwC under a £4.1 million contract, describing it as essentially a bookmarks site.

xAI is looking more like a datacentre REIT than a frontier lab(martinalderson.com)

658 points by martinald 4 months ago | 507 comments

Blog post by Martin Alderson arguing xAI's capacity deals with Anthropic and Google make it resemble a datacentre REIT more than a frontier AI lab.

Malleable Software(mdubakov.me)

90 points by tablet 1 year ago | 79 comments