Caveman: Why use many token when few token do trick • Hacker News Board

Re: Caveman: Why use many token when few token do trick

#311

Post by TiredOfLife » Sun, Apr 05, 2026, 9:35 PM UTC

altmanaltman wrote: post #235

teekert wrote: post #10

Idk I try talk like cavemen to claude. Claude seems answer less good. We have more misunderstandings. Feel like sometimes need more words in total to explain previous instructions. Also less context is more damage if typo. Who agrees? Could be just feeling I have. I often ad fluff. Feels like better result from LLM. Me think LLM also get less thinking and less info from own previous replies if talk like caveman.

Why say more word when less word do. Save time. Sea world.

You mean see the world or Sea World?

Re: Caveman: Why use many token when few token do trick

#312

Post by maxbeech » Sun, Apr 05, 2026, 9:49 PM UTC

[dead]

Re: Caveman: Why use many token when few token do trick

#313

Post by tonymet » Sun, Apr 05, 2026, 9:53 PM UTC

me ChatGPT like caveman always. Typing also faster.

Re: Caveman: Why use many token when few token do trick

#314

Post by Demiurg082 » Sun, Apr 05, 2026, 10:01 PM UTC

bitexploder wrote: post #182

Demiurg082 wrote: post #179

Earlier quoted context omitted.

CoT token are usually controled via 'extended thinking' or 'adapted thinking'. CoT tokens are usually not affected by the system prompt. There is an effort parameter, though, which states to have an effect on accuracy for over all token consumption. https://platform.claude.com/docs/en/build-with-claude/extend...

This helps, but the original prompt is still there. The system prompt is still influencing these thinking blocks. They just don’t end up clogging up your context. The system prompt sits at the very top of the context hierarchy. Even with isolated "thinking" blocks, the reasoning tokens are still autoregressively conditioned on the system instructions. If the system prompt forces "caveman speak" the model's attention…

I get your point but it seems that extended thinking is based on a hidden system prompt that is not so much affected by the style the user defines. Probably it's a bit in between.

https://docs.aws.amazon.com/bedrock/latest/userguide/claude-...

Re: Caveman: Why use many token when few token do trick

#315

Post by ggm » Sun, Apr 05, 2026, 10:07 PM UTC

F u cn Rd ths u cld wrk scrtry 'cpt w tk l thr jbs

Re: Caveman: Why use many token when few token do trick

#316

Post by testycool » Sun, Apr 05, 2026, 10:07 PM UTC

vurudlxtyt wrote: post #214

Grug brained developer meets AI tooling ( https://grugbrain.dev )

+1 Have used Grug as example for years to have LLM explain things to me.

Re: Caveman: Why use many token when few token do trick

#317

Post by nayroclade » Sun, Apr 05, 2026, 10:12 PM UTC

konaraddi wrote: post #277

nayroclade wrote: post #33

Cute idea, but you're never gonna blow your token budget on output. Input tokens are the bottleneck, because the agent's ingesting swathes of skills, directory trees, code files, tool outputs, etc. The output is generally a few hundred lines of code and a bit of natural language explanation.

In single-turn use, yeah, but across dozens of turns there's probably value in optimizing the output. Btw your point lands just as well without "Cute idea, but" https://odap.knrdd.com/patterns/condescending-reveal

I didn't mean it as condescending. I meant it literally is cute: A neat idea that is quite cool in its execution.

Re: Caveman: Why use many token when few token do trick

#318

Post by larodi » Sun, Apr 05, 2026, 10:15 PM UTC

sgbeal wrote: post #292

JBrussee-2 wrote: post #208

Author here. A few people are arguing against a stronger claim than the repo is meant to make. As well, this was very much intended to be a joke and not research level commentary. This skill is not intended to reduce hidden reasoning / thinking tokens. Anthropic’s own docs suggest more thinking budget can improve performance, so I would not claim otherwise. What it targets is the visible completion: less preamble, le…

> There is actual research suggesting concise prompting can reduce response length substantially without always wrecking quality, Anecdote: i discussed that with an LLM once and it explained to me that LLMs tend to respond to terse questions with terse answers because that's what humans (i.e. their training data) tend to do. Similarly, it explained to me that polite requests tend to lead to LLM responses with _more_…

this continual down-voting is not a personal thing for sure. perhaps there are crawlers that pretend to be more humane, or fully automated llm commenters which also randomly downvote.

Re: Caveman: Why use many token when few token do trick

#319

Post by fissible » Sun, Apr 05, 2026, 10:17 PM UTC

I have always been annoyed at the verbosity of ChatGPT and (to a lesser degree) Claude. I am aware of the long-term costs associated with trading that bloated context back and forth all the time.

Re: Caveman: Why use many token when few token do trick

#320

Post by codethief » Sun, Apr 05, 2026, 10:38 PM UTC

sumeno wrote: post #307

sgbeal wrote: post #292

Earlier quoted context omitted.

> There is actual research suggesting concise prompting can reduce response length substantially without always wrecking quality, Anecdote: i discussed that with an LLM once and it explained to me that LLMs tend to respond to terse questions with terse answers because that's what humans (i.e. their training data) tend to do. Similarly, it explained to me that polite requests tend to lead to LLM responses with _more_…

LLMs don't understand what they are doing, they can't explain it to you, it's just creating a reasonable sounding response

But that response is grounded in the training data they've seen, so it's not entirely unreasonable to think their answer might provide actual insights, not just statistical parroting.