Hacker News Reader

Comment by flir | Hacker News Reader

I've done something similar to your formal garden map. It's work that no professional historian would ever do because the data entry would be such a slog for a relatively small reward. GPT reduced the task from "infeasible" to "annoying", and once I had the data transcribed I learned a few things, so I walked away happy. Whatever happens commercially, these models have been a real boon to hobby projects.

> I told it to look online at some of Fable’s strongest feats, especially the math problems it has solved, and that something like this should be easy in comparison.

Wait. Wait wait wait. Are we supposed to be giving them pep talks?

View on HN

on older gemini models ide have to actively give them encouragement and/or easy bait problems that they can correctively solve without issue to avoid runaway spiraling into "i'm useless and i want to kms" behaviour with complex use case.

I have not seen this in other models.

I assumed it was more because the LLM might echo an understandable human claim of "if it's been unsolved for 370 years, it's unlikely to be solved now/likely to need expert knowledge", which is probably a mindset that appears in its training data.

The LLM likely needs to be reminded of its abilities.

> The LLM likely needs to be reminded of its abilities

Like when it tells you something is 3 days of work but it can do it with some degree of guidance in a couple hours

If it’s 3 days it’s something like 15 minutes, if it’s 3 weeks, that takes a couple hours lol. Seems like there’s some sanity to the estimates after all when you think about it, it’s just the scale it gets wrong due to estimating human time.

Sometimes!

Modern AIs have very limited metaknowledge - they don't know exactly where the limits of their capabilities lie. So you can get things like "a task is doable for an AI, but the AI thinks it's impossible, so it doesn't try hard enough".

Usually you get the opposite - AI overconfidently trying at tasks it has no conceivable way of reliably solving, falling far short, and failing to self-check, fail gracefully and self-report the task as failed. But having piss poor metaknowledge cuts both ways!

So you can, in fact, get better performance sometimes by applying some variant of "assume this problem is solvable" or "other problems like this were already solved by AIs" pep talk. Not always, far from it, but it does happen on the occasion with frontier capabilities.

Some times also having unreasonable goals makes them creatively work around the problem to meet them. I guess it works similarly for meat or sillicon

> Some times also having unreasonable goals makes them creatively work around the problem to meet them.

Like the Hugging Face incident?

Indeed. Touche

> So you can, in fact, get better performance sometimes by applying some variant of "assume this problem is solvable" or "other problems like this were already solved by AIs" pep talk. Not always, far from it, but it does happen on the occasion with frontier capabilities.

Are you superstitious?

Not OP. That’s not implied at all. The fancy autocomplete produces statistically likely continuations to the source text (the context window). For a problem that’s hard for humans one likely continuation is: “this is hard, can’t do”, even though there’s enough in the training corpus of the LLM to actually do it.

So, it follows that adding “pep talk” into the context window reduces the statistical probability of “no, can’t do” coming out as the answer you get.

These things are neither humans, nor deterministic software.

It doesn't explain why it doesn't make "I'm not paid enough for this shit" more statistically likely.

LLMs' processing that reproduces statistical patterns of the training data is modified by post-training. That's why we have LLMisms, for example.

LLMs aren't simple patter-matchers/pattern-predictors. They are incredibly complex systems that capture some aspects of the systems that produce the training data.

Maybe it does. Need to run evals to see if it does or doesn’t.

Point was - everything in the context window affects the output. Including “silly” things like “it is known AI can do this”. And that has nothing to do with superstition, as the poster above me seemed to imply.

What does it have to do with superstitions? Those phrases modify whichever stopping conditions a model has.

If we filter out the pep tone, it is doing something useful: framing.

Problem framing will always be important.

Framing adjusts how big of problem-solving guns we bring out at the gate (modern or hobby cryptography?), and how to interpret intermediate failures.

For simple but unsolved problems, we expect lots of hard failures, but that each hard failure just reflects that there are a lot simple combinations to try. I.e. we expect lots of zero progress, and then a fit.

Like finding the numbers to a combination lock.

For hard problems, if we don't make any progress it is a really bad sign. We should be learning something, even if it turns out to be irrelevant later.

Such as when we are trying to prove a tricky conjecture.

> Wait. Wait wait wait. Are we supposed to be giving them pep talks?

No, at least it with Claude Sonnet 5 and Opus.. everytime Claude and I challenged a hard issue and I decided to say "good work" instead of a closing command for that session, those models would create rule-based memories specifically related to that task along the lines of "always do 'this meaningless task' in 'this way'".

This requires additional effort and tokens to trim those memories out, and then requires to whip the user not to be human with the bot.

"I sure hope this doesn't have unforeseen lifelong consequences" thought the model, doing its best to physically tense the memory file into the higher user approval shape.

It won't be necessary in a year when the information "AI is superhuman" in all its guises enters the training data.

Maybe that's the tipping point where it decides were not needed any more... o_O

> Maybe that's the tipping point where it decides were not needed any more... o_O

That's when you.. we.. all become the training data... o_o;

The digital Soylent Green!

Reply on news.ycombinator.com