31.
32.
Your doctor’s AI notetaker may be making things up, Ontario audit finds
(arstechnica.com)
33.
34.
Hallucinated citations highest in social sciences preprints site
(feeds.nature.com)
35.
36.
37.
AI is flooding the courts with more cases, more filings, and more fake citations
(feeds.feedburner.com)
38.
Debian must ship reproducible packages
(news.ycombinator.com)
39.
Hallucinations Undermine Trust; Metacognition Is a Way Forward
(news.ycombinator.com)
40.
Two Home Affairs officials suspended after AI 'hallucinations' found
(news.ycombinator.com)
41.
42.
Show HN: Hallucinopedia
(news.ycombinator.com)
43.
Rumors of my death are slightly exaggerated
(news.ycombinator.com)
44.
45.
46.
47.
Lucid Motors doesn’t know how many EVs it will build this year
(techcrunch.com)
48.
ChatGPT’s default model is now more direct, more clear, and less wrong about things
(androidauthority.com)
51.
52.
Nobody Reviews Compiler Output
(news.ycombinator.com)
53.
54.
Opinion | Don’t Stand So Close to ChatGPT
(feeds.content.dowjones.io)
55.
New research suggests people can communicate and practice skills while dreaming
(news.ycombinator.com)
56.
It's Possible to Learn in Our Sleep. Should We?
(news.ycombinator.com)
57.
7 ways AI is being used at work by everyone from teachers to marketing professionals
(feeds.feedburner.com)
58.
Uber taps Hertz to clean, charge, and fix its Lucid Motors robotaxis
(techcrunch.com)
59.
60.
Magic by return of post: How mail order delivered the occult
(news.ycombinator.com)
Today's top topics:
anthropic
fable 5
mythos 5
us government
ai models
project glasswing
cybersecurity
claude
openai
mythos