Who even counts tokens? Who counts the number of servers? Who counts the seconds per request? Who counts anything anymore?

If it doesn’t matter how much you pay Anthropic or OpenAI for tokens, if it doesn’t matter how many instances are running on Amazon, if it doesn’t matter that an Azure function cold-starts for 10 seconds every single time — then the programmer probably doesn’t particularly matter either.

Replaced one with another. The other with a third. Fired the third. Hired a freelancer with a Max subscription to Claude Code. It doesn’t matter whether the freelancer holds one job or ten. Doesn’t matter if they’re frontend or backend. Doesn’t matter the language, the framework. Only the subscription matters.

What’s left when nothing matters?

What remains important in a time when nothing matters, nothing is needed, everything has died, everything is obsolete?

People often say in chats that what matters for a programmer is some kind of fundamental knowledge. But nobody can explain what that actually is. Then comes the flood of links, or the flame war.

What is this “fundamental knowledge”? This “programmer’s foundation”? This “what every programmer should know”? Some mysterious abstraction that programmers pronounce mysteriously and then walk off into the sunset — from the micro-kitchen back to the computer — while influencers record five-hour videos about the fundamentals.

Then a real number showed up

Uber introduced a limit: $1,500 on Cursor and Claude Code, per engineer, per month.

Tokenmaxxing reached such a peak that companies started introducing limits. But which limits exactly, and how to justify the token spend — that’s the interesting part.

Add up the two tools, multiply by 12, and you get $36k a year. The average developer salary at Uber in the US is about $330k a year. So the company is willing to pay roughly 11% on top, for AI agents.

The numbers aren’t exact. The numbers are relative. But the trend is what’s interesting — how much a company is willing to spend at all — because the previous figure was “as much as you like.”

Well, now there’s a figure. Multiply your annual salary by 0.11 and do the math. You get a number. Not a simple number.

So: are you ready to pay that much for AI agents?

And do you bring the product that much value yourself?