I guess because tokens is the only metric that ""reflects" how high your AI adoption is. So especially in tech industry, if a company pushes for AI adoption, the token usage is the only proxy metric you have.
if i had to guess: people are afraid to use new tools and the initial reduction in productivity and don't know what the acceptable $ usage is limiting their growth/self-education and exploration of an entire new way of working.
basically its a green light to use all the time and resources they want to explore and learn
Before LLM some companies were measuring developers performance based on the number of commits (regardless of quality). Not surprising now companies are measuring performance using token usage
But I can see a loose link between commit count and productivity (obviously a flawed metric). There isn’t such a link between token count and productivity. Like, I can burn millions of tokens with literally nothing to show, not even commits
I genuinely don't think any enterprise companies are token-maxxing seriously. It strikes me as a way to drive PR, maybe attract pro ai-dev talent, and generally agitate the community.
This is what Graeber called "box-ticking" in Bullshit Jobs: work that exists so the org can prove it's doing something, not because the doing matters. The leaderboard isn't measuring productivity, it's producing proof of AI adoption. Once an exec says "we're an AI-first company," the rest of the org needs to show that's actually happening. Token counts are the easiest thing to put on a dashboard, regardless of whether anyone got anything done.
I guess because tokens is the only metric that ""reflects" how high your AI adoption is. So especially in tech industry, if a company pushes for AI adoption, the token usage is the only proxy metric you have.
Agree that it's absurd though.
if i had to guess: people are afraid to use new tools and the initial reduction in productivity and don't know what the acceptable $ usage is limiting their growth/self-education and exploration of an entire new way of working.
basically its a green light to use all the time and resources they want to explore and learn
Before LLM some companies were measuring developers performance based on the number of commits (regardless of quality). Not surprising now companies are measuring performance using token usage
But I can see a loose link between commit count and productivity (obviously a flawed metric). There isn’t such a link between token count and productivity. Like, I can burn millions of tokens with literally nothing to show, not even commits
I genuinely don't think any enterprise companies are token-maxxing seriously. It strikes me as a way to drive PR, maybe attract pro ai-dev talent, and generally agitate the community.
might as well call it shit-maxxing
Yeah, that’s why I would be interested to get info from people inside those companies
All marketing is based on people's irrationality. What would you buy AI for if it wasn't for hype and FOMO
This is what Graeber called "box-ticking" in Bullshit Jobs: work that exists so the org can prove it's doing something, not because the doing matters. The leaderboard isn't measuring productivity, it's producing proof of AI adoption. Once an exec says "we're an AI-first company," the rest of the org needs to show that's actually happening. Token counts are the easiest thing to put on a dashboard, regardless of whether anyone got anything done.
Ok, that does make sense when framed this way. So, absurd, yes, but for a clear reason