Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. It's not hypothetical: the dangers of AI are here (theguardian.com)
    —discuss
  2. Minimal Assumptions for Societal Resilience (chillphysicsenjoyer.substack.com)
    —discuss
  3. Show HN: Farm.js – A full-stack framework with an AOT compiler for React (farmjs.dev)
    —discuss
  4. Carnegie: China Passes US as Top AI Talent Hub, 40.6% to 34.2% (aiweekly.co)
    —discuss
  5. Show HN: NerfWatch() – Track AI degradation with daily tests and community votes (nerfwatch.lol)
    —discuss
  6. Show HN: BuyerCue – Find potential customer by paid-placement evidence (buyercue.io)
    1comments
  7. Where's the Beef? [video] (youtube.com)
    —discuss
  8. Kafgres: Embedding a Kafka Broker into Postgres (rynr.dev)
    —discuss
  9. The Animated Elliptic Curve (xargs.org)
    —discuss
  10. VAT-Inclusive vs. VAT-Exclusive Pricing: The Math Developers Get Wrong (vat-engine.app)
    —discuss
  11. Manchester City found guilty on 114 charges relating to financial breaches (nytimes.com)
    —discuss
  12. IRS launches new mobile app, expanding digital services for taxpayers (irs.gov)
    —discuss
  13. "Someone just sent me an ad that features an AI version of me. What do I do?" (bsky.app)
    —discuss
  14. A terminal-first multiplayer thread for agent-authored work (cueloop.dev)
    —discuss
  15. Become the Thousand Servers (sabot.media)
    —discuss
  16. Std: Call_once vs. Std:Async (devblogs.microsoft.com/oldnewthing)
    —discuss
  17. Comparing exception behavior of magic statics, std:call_once, and std:async (devblogs.microsoft.com/oldnewthing)
    —discuss
  18. The hottest new hangout for middle schoolers is NPR's comment section? (techcrunch.com)
    —discuss
  19. Ask HN: Does your team still have humans review and understand code?
    —discuss
  20. Show HN: Ekselio – Loveable for Finance Workflows (local first) (gptbeyond.com)
    —discuss
  21. Revealing the details of how OpenAI agents hacked Hugging Face (swarmtraces.org)
    —discuss
  22. FYI: OpenAI "spend limits" aren't always limits
    —discuss
  23. Shipping Code Faster Is Not the Bottleneck (valentinprugnaud.dev)
    —discuss
  24. Jev was built for agents, here's how we're using it in computer use instead (twitter.com/kylejeong)
    —discuss
  25. Human homeostasis is more efficient in social proximity (science.org)
    —discuss
  26. Show HN: Browser Bookmarks, Done Right: Stop Building a Cemetery of Links (chromewebstore.google.com)
    —discuss
  27. Show HN: Pbi-guard – Unit testing for Power BI (github.com/patens-dev)
    —discuss
  28. Teaching a World Model to Play Pokemon (nostalgia.dev)
    2comments
  29. North Korea 'Likely' Behind $388M Hack of Crypto Exchange Bitget (gizmodo.com)
    1comments
  30. Excel now supports multiple values in a single cell (techcommunity.microsoft.com)
    —discuss

Too AI; Didn't Read

20 pointsby 55m agotai-dr.com
8 comments
30m agoHN ↗

If it’s not worth writing, it’s not worth reading.

That said, wouldn’t it be funny if this site, too, were AI generated?

10m agoHN ↗

I believe that one was discussed with the author on HN, and that they used AI to 'copyedit'. Unfortunately that's a trap because if you make a general copyediting request most LLMs will absolutely slopify your work-- and make edits that are unambiguously bad as well as just making it look like lazy llm output.

The better use is to ask it to point out grammatical and spelling errors, or phrasing which is unclear.. then go fix yourself. Never hand the the pen unless you're looking to subject everyone to your Honest Take That Recontextualizes Everything.

14m agoHN ↗

I understand the sensation of disgust when being offered to read a piece of AI slop. However, I'm actually curious about how our sensitivity towards slop will evolve. I imagine people will have varying degrees of tolerance.

Afterall, AI slop doesn't necessarily mean the prompter didn't read. It just means they're OK with the AI phrasing. It's also possible that they read the output and had the AI update with whatever edits they prompted.

Writing this comment makes me more convinced that if us users of AI want our written work to be taken seriously, then we should also share our prompts. I hope we quickly converge into a standard way of sharing the prompts that generated a given product.

10m agoHN ↗

I hope that labs are actively working on the voice and writing styles of their models. I don't want to read slop in large part because their prose is overly verbose and complicated. If the models were writing simple sentences in mainstream English, mirroring the vocabulary used by my colleagues, I'd mind less than the current status.

Surely that's something that can be trained for.

12m agoHN ↗

Ironic that this itself is an obvious slop site

12m agoHN ↗

While I agree with the sentiment, this:

If you couldn't bother to read it, why should I?

doesn't follow. You don't know that they didn't read it! Only that they didn't write it. If the point of things like this is to be careful, then be careful.