Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Unix Was Right. We Made Linux Too Complicated [video](youtube.com ↗)
    discuss
  2. This California Educational Juggernaut Doesn't Need the SAT(nytimes.com ↗)
    1comments
  3. Spectral AI Architecture: Executive Summary and Core Concepts(github.com/srdwy99m5j-lab ↗)
    discuss
  4. What Is LPO, NPO, CPO? Knowledge of Optical Interconnect Technologies(fiber-mart.com ↗)
    discuss
  5. Ask A Monk – A digital wilderness for thoughts with no immediate answer(askamonk.online ↗)
    discuss
  6. Kubernetes v1.37: Hardening emptyDir and bind mounts(kubernetes.io ↗)
    discuss
  7. Fender Is Making Enemies with a Messy Fight over Its Iconic Strat(bloomberg.com ↗)
    1comments
  8. Show HN: PortButler – send files from macOS to a device with no way to receive(sshlab.dev ↗)
    discuss
  9. How often people say a habit, therapy, diet or supplement changed their life(jeravalue.com ↗)
    discuss
  10. Fingerprinting Network Honeypots with Weighted Behavioral Scoring Engine(meetcyber.net ↗)
    discuss
  11. Claude Code from Source(claude-code-from-source.com ↗)
    discuss
  12. SEC Issues "Innovation Exemption" to Facilitate Trading of Tokenized NMS Stock(sec.gov ↗)
    discuss
  13. Hackers Used Anthropic's Claude to Break into OpenAI(wsj.com ↗)
    1comments
  14. Jev means structured output is interesting again(seangoedecke.com ↗)
    discuss
  15. Hypercomb – AI open software platform(hypercomb.io ↗)
    discuss
  16. Type Declaration Syntax(citrons.xyz ↗)
    discuss
  17. Telstra outage: The night a network decided the year was 2006(netnod.se ↗)
    discuss
  18. Goose: 1.16x faster than C++ and 1.12x than safe Rust, while memory safe(github.com/aardappel ↗)
    4comments
  19. Today we're releasing Helix 2.5(twitter.com/figure_robot ↗)
    discuss
  20. AI agents can modify themselves without humans telling them to do so(theregister.com ↗)
    1comments
  21. When A.I. Becomes an Enemy [video](youtube.com ↗)
    1comments
  22. Lakera – Test your AI hacking skills(lakera.ai ↗)
    discuss
  23. From Operations to a Robot Arm(huggingface.co ↗)
    discuss
  24. GrapheneOS accuses Google of gatekeeping Android 17 features and security fixes(androidauthority.com ↗)
    1comments
  25. Apprentice: A Slim, Extensible Coding Harness (Written in Common Lisp)(github.com/skarnati20 ↗)
    2comments
  26. Getting Outbound Email to Work on DigitalOcean(mailfully.com ↗)
    discuss
  27. Anthropic says Claude now leads a quarter of work building its next AI models(businesstimes.com.sg ↗)
    1comments
  28. Mini-Jev – typesafe's Jev implemented on top of an LLM locally(github.com/r-ms ↗)
    discuss
  29. Inside Microsoft and OpenAI, Worry About Damaging the Publishing Industry(nytimes.com ↗)
    discuss
  30. Richard Feynman – The World from another point of view [video](youtube.com ↗)
    discuss

AI agents can modify themselves without humans telling them to do so

2 pointsby 53m agotheregister.com
1 comments
31m agoHN ↗

it could fix the app in one of two ways: it could either change the code, or change the model itself. It chose the latter, replacing the model powering both the application and future instances of the agent itself.

Nah, common. It was given a task of which part of it could be fixing an AI model, so it does what it does and replaces or modifies the model to complete its task. They're making it sound like out of nowhere it just decided to change. They are not agentic (actually) in any way.