All Topics
All Topics
Technology
Technology
Design
Design
Programming
Programming
Science
Science
News
News
Gaming
Gaming
Entertainment
Entertainment
Business
Business
Finance
Finance
Sports
Sports
Health
Health
Food
Food
Travel
Travel
Art
Art
Music
Music
Books
Books
Education
Education
Politics
Politics
Personal
Personal
No algorithm. No AI slop. No ads. Just RSS. Pro-human. Indie writers. Real journalism. Open web. Chronological. Hand toasted.

Why measuring developer productivity by AI token usage is a flawed metric

By

Alonso del Arte

10d ago· 7 min readenOpinion

Summary

The article criticizes the trend of managers using AI token usage as a metric to measure software developer productivity. It argues that this is a flawed approach that leads developers to game the system through practices like "tokenmaxxing" — artificially inflating AI token consumption. The piece draws parallels to other bad productivity metrics that have historically been used in software development, and contends that pushing AI tools on developers and measuring their output through token counts is counterproductive and misunderstands how real development work gets done.

Key quotes

· 3 pulled
When a dev's job depends on a bad metric, he's got no choice but to game it.
Something about A.I. makes pointy-haired bosses focus on the worst metrics for measuring the productivity of software developers.
If the pointy-haired boss is tracking A.I. token usage, the devs are going to engage in what is called 'tokenmaxxing.'
Snippet from the RSS feed
Something about A.I. makes pointy-haired bosses focus on the worst metrics for measuring the productivity of software developers. Some of these bad metrics have been around for a long time. Others…

You might also wanna read