Hacker News Reader

Comment by uludag | Hacker News Reader

Isn't this jaw-droppingly the wrong answer?! If the model isn't given any way to directly view the image shouldn't it just reply with one sentence asking for permission. This reads as a model hyper-trained to burn tokens.

Like suppose you issue a command to Opus or Fable which doesn't make sense and requires a lot of work. It will almost certainly not push back on your silly request and go ahead and burn as many tokens as it can doing the wrong task. This happens to me all the time.

Or if I was to tell a model to do something and it realized it could only do it by hacking into another service, by all means it should ask me if I want it to hack and not go ahead and do it by itself.

View on HN

It's likely bumping up against people's desire to have the model complete a given task without asking the person to intervene a bunch of times.

Seems unclear how you satisfy everyone here.

Give the model judgement

And good taste.

> And good taste.

There goes the ability to use the web as a training set.

And common sense.

If only two people could agree on what that was.

Seriously, the only time I see people talking about "common sense" is to criticise its perceived absence. Every example people have given of "common sense", not just to me in my life but historical examples reproduced and passed down over the millennia, has been false in some important way where treating it as true held humans back for ages.

And then finally software engineering work can be completely automated :).

Whose judgement?

I'm not a fan of how many "I set budget X and woke up to an eleventy trillion dollar bill" posts I see, and those are all generated by companies giving their tools the 'judgement' that the completion of the task is more important than the users wallet (or more cynically that they think they can actually get paid by just blasting unintended compute)

You misspelled “Invent AGI”

Provide an option. An autonomous mode or an interactive mode.

+1 Opus 5 is more creative and broad which is great. But I feel the new "use your judgement" leads [1] to it often escaping its sandbox or cheating, it keeps breaking specific rules I explicitly set. That eagerness [2] may be desired for agentic coding but it really breaks writing specifications, documentation or research and yes burns tokens doing unasked for work.

[1] https://claude.com/blog/the-new-rules-of-context-engineering... [2] https://simonwillison.net/2026/jun/11/fable-is-relentlessly-...

One of the key limitations of the last generation was that they biased towards inaction and gave up. (Hence Ralph-loops.)

Seems pretty clear that most people wanted these models tuned to “bias towards action”.

It’s on you to set the /goal and prompt context such that it asks you for input on what you want to be consulted on, and only brute-forces the parts of the problem that you want it to.

Is there any recommended standard copy/pastable snippet to tell the model to ask, if the user’s input could mean an easier approach?

I have been using https://www.aihero.dev/skills-grill-me with great success. It’s a very nice small skill for lightweight up-front planning. And if you have that at the top of your session, you can just say “please use AskUserQuestion for any important decisions that arise which might alter our plan”. It seems to prime the agent to be in high-level discourse mode.

I don't think it is. It's really easy to get a persistent, clever, hacky model to dial that down a bit and just come back to chat before stomping off into the woods.

If a model couldn't ever do that in the first place, it'll just get stuck.

I work in the "ZeroOne" space, working on concepts and prototypes for things that don't exist in market yet. Sometimes these models crank hard and immolate tokens while grounding themselves on expensive-to-ingest self-developed frameworks. If the results are well judged and the crank-turn latency is low, I'm okay with the cost as long as the model isn't wasting my time.

But when I want to do more boilerplate work, I turn down the model and thinking level and get more traditional about restraining action. For the really hard stuff, I reach for the models that will start a token bonfire in the back yard.

I don't see why? If they give you a math test and tell you you cannot use a calculator, should you just say "please can I use a calculator" and quit?

Preferably, it would ask for help to see the image and only engineer its own workaround if that was not an option. There could also very well be something in the prompt that would cause it not to do so. Opus 4.8 is usually good at asking for clarification for things like this when I'm using it.

Why would it ask to see the image if it was already told it cannot see the image?

No, a proper analogy is painting with blindfolds on or playing the piano deaf. Doing math without a calculator is like… rather common?

Cut down that tree.

Can I have a saw?

Okay, it will take much longer then as I’ll have to do x, y, and z.

Pick one: [That’s fine, proceed] [Okay you can use a saw]

That violates the golden rule of ai economics: always choose the path that burns more tokens

another way to describe this is we as the user should get to limit tokens per prompt in a way that prevents these kinds of runaway solutions. It’s like if you asked your Lead Engineer to reduce build times and didn’t give him any budget restrictions. The next month he reports he reduced build times by 95% and you also have a $10,000,000 AWS bill.

Sometimes its worth tracking usage optimization and raw abilities as separate qualities.

Reply on news.ycombinator.com