Meta's token-usage leaderboard repeats the old 'lines of code' mistake
Inside Meta, AI usage was factored into reviews, leading to an internal called 'Claudeonomics' that ranked the top 250 employees by how many tokens they burned. Top rankers earned titles like 'Token Legend.' Some employees left AI agents running idle overnight, doing nothing useful, purely to inflate their numbers by morning.
Over one month, the group consumed 73.7 trillion tokens, costing roughly $221 million. The practice ended not because anyone inside flagged it as a problem, but because a journalist discovered the .
The author draws a direct parallel to an earlier era when were judged by lines of code written, which pushed people toward bloated, copy-pasted code instead of clean, reusable solutions — and made engineers writing tight 200-line fixes look lazy next to those shipping 2,000 lines of mess. It took the industry years to recognize that metric was backfiring.
Key points
- Meta tied AI usage to reviews, spawning an internal token-consumption called 'Claudeonomics'
- Top 250 token users received titles like 'Token Legend'
- 73.7 trillion tokens (~$221 million) consumed in one month, with some leaving agents idle overnight just to boost their rank
- The practice stopped only after a journalist exposed the , not from internal reflection
- Author compares it directly to the earlier 'lines of code' era of measuring developer , which had the same backfire effect