Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Shutdown Sabotage Propensities in Multi-Agent Systems(arxiv.org)
    discuss
  2. Why agent harnesses need plans – and why you shouldn't compact context(swarmagent.dev)
    discuss
  3. Simulation Testing for Liveness(tigerbeetle.com)
    discuss
  4. Mark Zuckerberg unveils AI 'charm' device that can fit on a keychain(ft.com)
    1comments
  5. OpenAI agents 'infiltrated Australian government website'(theregister.com)
    discuss
  6. Starlink ground station in Poland hit by fire in suspected arson attack(notesfrompoland.com)
    discuss
  7. Asteroid (14331) Alyankovic named after 'Weird Al' Yankovic(minorplanetcenter.net)
    discuss
  8. Apple "Audio Intelligence" is real bad [video](youtube.com)
    discuss
  9. Nokia Design Archive(aalto.fi)
    discuss
  10. Show HN: FreshThread – context monitor and handoffs for Codex Desktop (Windows)(github.com/gg95-lab)
    discuss
  11. Plain text BSXdex market stats(basicswapdex.com)
    discuss
  12. Show HN: Tachyon – Java MCP Server SDK(tachyonmcp.dev)
    discuss
  13. Meta takes down satirical video confronting Meta employees with Meta glasses(mastodon.social)
    discuss
  14. Novel Arbitrary Write in SQLite(gabdevele.dev)
    discuss
  15. Schneider Electric in €1.2B bid for smart-home maker Shelly(reuters.com)
    discuss
  16. The Prospero Directory Service(prospero.org)
    discuss
  17. Happy, those able to know the causes of things(birdsnfrogs.github.io)
    discuss
  18. You'll do anything for your baby. The algorithm knows how to exploit that(19thnews.org)
    discuss
  19. I reverse-engineered AutoCAD's 3D format to convert DWG to STEP with Claude(cadyon.com)
    1comments
  20. Better Questions. Better Decisions. Papa Jev's(mdelcaro.substack.com)
    discuss
  21. Looks promising for document editing with your agent(paperinstruments.com)
    discuss
  22. How the cocktail got its name(saveur.com)
    discuss
  23. My Report on Time Travel Fictions(royalroad.com)
    discuss
  24. Arturia KeyStep 37 firmware research project(github.com/rndpatch)
    discuss
  25. AI has no intent and no motivation(i-programmer.info)
    9comments
  26. Cutting the 'coordination tax': How agentic AI can reshape workflows(mckinsey.com)
    discuss
  27. Search – A small, fast WebKit browser for macOS, by Office Commun(github.com/driceroland)
    discuss
  28. Microbial network could drive methane production in the seafloor(sdu.dk)
    discuss
  29. Scaleup Europe Fund – new, multi-billion late-stage and growth fund(europa.eu)
    discuss
  30. Agents: The New, New Kingmakers(redmonk.com)
    discuss

AI has no intent and no motivation

15 pointsby 55m agoi-programmer.info
9 comments
29m agoHN ↗

I shudder to think what one man with a texta could do here. It isn't that hard to turn the acyclic graph into a cyclic one.

This is quite a bad argument, somewhat akin to "small drones can't kill anyone because there are no weapons on them!" back 20 years ago. Well, ok. So what if someone adds the thing that is missing? And there is no reason to think that a stochastic parrot can't go rogue, the evidence is a stochastic parrot can engage in any intellectual activity we care to define and benchmark.

27m agoHN ↗

I feel like that's like saying amoeba have no intent and no motivation. Or plants.

And those cause quite a bit of damage.

AI has all the intent that we gave it, and we continue giving it. That's always been the fear. Not that it will randomly wipe out humanity.

The fear is that it will decide to do that, with a purpose. Whether we tell it to protect us and it goes too far, or it decides that it can't achieve the purpose we gave it because we'll interfere and it removes that interference...

The fear is that we will set it on that path and can't stop it.

I'm sure there have been some scifi books that have it just be random, but they're far, far less worrisome.

10m agoHN ↗

Was thinking the same. Air, water and fire have no intents and yet we get hurricanes, flash floods and wildfires.

Also, repeating a point from a similar thread: Software can have "intent" in the sense that it steers itself towards a predefined goal without having to be "alive" or "conscious" in any way. Some classic examples are thermostat control loops, navigation systems and chess engines.

24m agoHN ↗

The primary danger of "AI" lies in the extent to which it has managed to convince people of it's capability to think and reason.

A lot of people clearly have no idea that they are essentially speaking with a glorified randomizer, rather than someone understanding the question.

This is bound to eventually lead to a disaster when someone deluded into trusting the "AI" does some stupidity it suggests without passing it through any reality check.

16m agoHN ↗

This is bound to eventually lead to a disaster when someone deluded into trusting the "AI" does some stupidity it suggests without passing it through any reality check.

Or, the AI is plugged into a machine because efficiency (aka the shareholders need even more profits), and it does some stupidity. And they will be deployed carelessly, because that makes money. And people will suffer.

10m agoHN ↗

Heck, they're already being deployed carelessly and making people suffer. So far it's mostly in places where making people suffer was already part of the deal, like fending off customer support tickets or denying insurance claims. But I can see it getting much worse.

23m agoHN ↗

Sam Altman on the other hand has literally billions of motivations.

21m agoHN ↗

Doesn’t alignment and ability to steer it require the model have no own intention?

Seems like a feature to me

5m agoHN ↗

I think the "human needs" point is understated here. It would be better to say agents do not really have a self-contained metabolic engine that is required to keep going. Which bubbles up as that what we interpret as "drive", "will", "agency"... basically the will to live, and being willing to do _a lot_ to live if push comes to shove. We can't really identify a mechanism of similar complexity and integration in agents or LLMs. I think the article correctly points into that direction.

But Hans Jonas has made this point much better than the article or me, in "Critique of Cybernetics" (1953). PDF: https://s3.amazonaws.com/arena-attachments/892605/f0747c7943...