AI's costs are going through the roof - so businesses are telling LLMs to talk like cavemen
Me think this hilarious sign of times.
I find this very funny:
“Companies are deliberately making their AI tools speak like cavemen in an attempt to stop burning through AI tokens and curb their massive expenditure on AI, 404 Media has found. The tool turns the usually verbose outpost of LLMs like Claude Code, Codex, or Gemini into a much more to the point answer. Think less “you’re right to push back, I was wrong,” and more “Hulk smash.””
If only we had other limited-vocabulary lexicons designed to talk to computers efficiently!
I think we’re circling a few different possibilities that may show up over the next few years:
- Literally LLM-specific “programming languages” that humans can use to talk to models more efficiently, of which Caveman is the hilarious first step
- A proprietary bytecode-like language for LLMs that makes interactions more efficient but also just happens to be owned by one of the major vendors and creates a hitherto-unobtainable moat for their business
- This all becomes moot when local models become viable for most businesses without insanely high hardware prices or configuration costs
- LLM costs eventually fall to a fraction of their existing level
But who knows? Maybe enterprise businesses will continue to talk in stilted caveman language to achieve their business goals forever.
