Why use many token when few token do trick.

It is fun, and it costs you nothing measurable in quality.

82 paired coding tasks. 8.5% fewer output tokens; no detectable quality difference. Total cost varied.

I don’t know about token savings, but I find the “caveman style” much easier to read and understand than typical LLM-slop.

Turns out, you can understand the caveman output just fine.

Mass fun. Starred.

seeing something framed by a caveman in a couple of occasions peeled back a layer i didnt see before.

it has already made a lot of developers significantly less annoyed.

Reporting on the wider caveman-style movement, including this Skill.

So it’s a prompt to turn Jarvis into Hulk!

honestly it will make my day not to have to read through entire essays about some trivial solution.

A senior OpenAI employee has contributed code to the project, simply called ‘caveman.’

From the article’s introduction. Full reporting requires a subscription.

Me think this good idea. Regular language unnecessary complex. Distract meaning.

this gave me a good laugh + some solid ideas.

The Elastic-Caveman initiative has successfully optimized AI response size without compromising quality or functionality.

Elastic’s own variant: 63.6% fewer output tokens across eight live MCP scenarios.

But combining this with caveman? Gold!

If you’re not cavemaxxing you’re falling behind.

if we allow midterms, it’s hard to beat Caveman.

From an opinion column on the surreal economics of AI.

if goal make code, few word better. if goal make insight, more word better. depend on task.

kevin would be proud

Great idea

me like that

I like