Dossier
LessWrong
Coverage of LessWrong in the Nexus archive.
- Did the Pope use AI to write about the dangers of AI?
An analysis suggests parts of Pope Leo XIV's encyclical 'Magnifica Humanitas' may be AI-generated, with 40-100% of certain paragraphs flagged by AI detector Pangram. The document shows traits like high use of 'genuinely,' similar to Anthropic's Claude, and 62% of the first chapter was identified as AI-written.
- The effects of caffeine consumption do not decay with a ~5 hour half-life
The article challenges the common belief that caffeine's effects decay with a 5-hour half-life, suggesting the decay rate may vary. It references a study and discussion on platforms like LessWrong and Hacker News.