Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Xiaomi MiMo v2.6(xiaomi.com)
    137comments
  2. The NASA/ESA Mars Sample Return mission has been canceled(science.org)
    154comments
  3. Transformers Explained Visually(poloclub.github.io)
    16comments
  4. What Sun got wrong(dtrace.org)
    249comments
  5. Attention is all you have(alicegg.tech)
    148comments
  6. Divide by Depth for Instant 3D(gabrieloc.com)
    4comments
  7. Why does mathmain need an encrypted loader?(safedep.io)
    25comments
  8. The Advisory Group on Mathematics and Artificial Intelligence(terrytao.wordpress.com)
    27comments
  9. Apple Copland D11E4 Booting in the Browser(pagetable.com)
    15comments
  10. Grok 4.7(x.ai)
    364comments
  11. In Search of a Compositional Theory of Self-Stabilization(muratbuffalo.blogspot.com)
    3comments
  12. Roboharm: Do frontier robot policies refuse unsafe instructions?(robocurve.org)
    10comments
  13. CBP suspends all personal prescription importation Oct 22(personalimportation.org)
    26comments
  14. US halts flights at busy East Coast airports, says fiber line cut(reuters.com)
    94comments
  15. AI coding has made CI a bottleneck, so we reworked ours to keep up(linear.app)
    71comments
  16. Turn off and restrict access to Apple Intelligence features on Mac(support.apple.com)
    128comments
  17. Kev: Tiny Jev-like family of decision models built on top of Qwen3.5(github.com/jaredpalmer)
    169comments
  18. TXR: An Original, New Programming Language for Convenient Data Munging(nongnu.org)
    discuss
  19. Frontier AI on Your Own Hardware(timdettmers.com)
    26comments
  20. Python Workers are now generally available(cloudflare.com)
    26comments
  21. This Digital Radio Gets Messages to the World’s Remotest Locations(ieee.org)
    33comments
  22. Fable 5 – Median thinking declined in August(twitter.com/lon)
    205comments
  23. Show HN: Foremerge – Catch intent conflicts between parallel coding agents(github.com/naw103)
    discuss
  24. Avoiding the babbling-idiot failure in a time-triggered communication system(ieee.org)
    6comments
  25. How do Traffic Signals Work (2019)(practical.engineering)
    33comments
  26. Noodle Gallery- Open-source, self-hosted alternative to Google Photos and Immich(digitalescapetools.com)
    34comments
  27. A restored PDP-11/83 serving this page on 211BSD Unix(pdp1173.com)
    32comments
  28. M5 Ultra Mac Studio Review(macstories.net)
    209comments
  29. Heretic removes restrictions from language models(heretic-project.org)
    94comments
  30. macOS 27: Workaround to avoid downloading AI models and save storage(reddit.com)
    89comments

Artificial intelligence is ineffective and potentially harmful for fact checking

1 pointsby 3y agoarxiv.org
1 comments
3y agoHN ↗

What about systems like pdfgpt and knowledge_gpt; that, in returning citations from the trained texts, might be useful for legal discovery with admissibility?

pdfgpt: https://github.com/bhaskatripathi/pdfGPT

knowledge_gpt: https://github.com/mmz-001/knowledge_gpt

Are LLM tools better or worse than e.g. meilisearch or elasticsearch for searching with snippets over a set of document resources?

How does search compare to generating things with citations?