Comment by stingraycharles

Comment by stingraycharles 2 days ago

16 replies

View on Hacker News

> As a chatbot, it's the only one that seems to really relish calling you out on mistakes or nonsense, and it doesn't hesitate to be blunt with you.

My experience is that Sonnet 4.5 does this a lot as well, but this is more often than not due to a lack of full context, eg accusing the user of not doing X or Y when it just wasn’t told that was already done, and proceeding to apologize.

How is Kimi K2 in this regard?

Isn’t “instruction following” the most important thing you’d want out of a model in general, and a model pushing back more likely than not being wrong?

Kim_Bruning 2 days ago

> Isn’t “instruction following” the most important thing you’d want out of a model in general,

No. And for the same reason that pure "instruction following" in humans is considered a form of protest/sabotage.

https://en.wikipedia.org/wiki/Work-to-rule

Reply View 14 replies
  • stingraycharles 2 days ago

    I don’t understand the point you’re trying to make. LLMs are not humans.

    From my perspective, the whole problem with LLMs (at least for writing code) is that it shouldn’t assume anything, follow the instructions faithfully, and ask the user for clarification if there is ambiguity in the request.

    I find it extremely annoying when the model pushes back / disagrees, instead of asking for clarification. For this reason, I’m not a big fan of Sonnet 4.5.

    Reply View 12 replies
    • IgorPartola 2 days ago

      Full instruction following looks like monkey’s paw/malicious compliance. A good way to eliminate a bug from a codebase is to delete the codebase, that type of thing. You want the model to have enough creative freedom to solve the problem otherwise you are just coding using an imprecise language spec.

      I know what you mean: a lot of my prompts include “never use em-dashes” but all models forget this sooner or later. But in other circumstances I do want it to push back on something I am asking. “I can implement what you are asking but I just want to confirm that you are ok with this feature introducing an SQL injection attack into this API endpoint”

      Reply View 2 replies
      • stingraycharles 2 days ago

        My point is that it’s better that the model asks questions to better understand what’s going on before pushing back.

        Reply View 1 reply
        • IgorPartola 2 days ago

          Agreed. With Claude Code I will often specify the feature I want to develop, then tell it to summarize the plan for me, give me its opinion on the plan, and ask questions before it does anything. This works very well. Often times it actually catches some piece I didn’t consider and this almost always results in usable code or code that is close enough that Claude can fix after I review what it did and point out problems.

          Reply View 0 replies
    • Kim_Bruning 2 days ago

      I can't help you then. You can find a close analogue in the OSS/CIA Simple Sabotage Field Manual. [1]

      For that reason, I don't trust Agents (human or ai, secret or overt :-P) who don't push back.

      [1] https://www.cia.gov/static/5c875f3ec660e092cf893f60b4a288df/... esp. Section 5(11)(b)(14): "Apply all regulations to the last letter." - [as a form of sabotage]

      Reply View 2 replies
      • stingraycharles 2 days ago

        How is asking for clarification before pushing back a bad thing?

        Reply View 1 reply
        • Kim_Bruning 2 days ago

          Sounds like we're not too far apart then!

          Sometimes pushback is appropriate, sometimes clarification. The key thing is that one doesn't just blindly follow instructions; at least that's the thrust of it.

          Reply View 0 replies
    • InsideOutSanta 2 days ago

      I would assume that if the model made no assumptions, it would be unable to complete most requests given in natural language.

      Reply View 1 reply
      • stingraycharles 2 days ago

        Well yes, but asking the model to ask questions to resolve ambiguities is critical if you want to have any success in eg a coding assistant.

        There are shitloads of ambiguities. Most of the problems people have with LLMs is the implicit assumptions being made.

        Phrased differently, telling the model to ask questions before responding to resolve ambiguities is an extremely easy way to get a lot more success.

        Reply View 0 replies
    • simlevesque 2 days ago

      I think the opposite. I don't want to write down everything and I like when my agents take some initiative or come up with solutions I didn't think of.

      Reply View 0 replies
    • wat10000 2 days ago

      If I tell it to fetch the information using HTPP, I want it to ask if I meant HTTP, not go off and try to find a way to fetch the info using an old printing protocol from IBM.

      Reply View 0 replies
    • MangoToupe 2 days ago

      > and ask the user for clarification if there is ambiguity in the request.

      You'd just be endlessly talking to the chatbots. Humans are really bad at expressing ourselves precisely, which is why we have formal languages that preclude ambiguity.

      Reply View 0 replies
    • scotty79 2 days ago

      > is that it shouldn’t assume anything, follow the instructions faithfully, and ask the user for clarification if there is ambiguity in the request

      We already had those. They are called programming languages. And interacting with them used to be a very well paid job.

      Reply View 0 replies
  • SkyeCA 2 days ago

    It's still insanity to me that doing your job exactly as defined and not giving away extra work is considered a form of action.

    Everyone should be working-to-rule all the time.

    Reply View 0 replies
hugh-avherald a day ago

Only if you're really, really good at constructing precise instructions, at which point you don't really need a coding agent.

Reply View 0 replies