Measuring AI work by token use can make costs run away
Using as a can push AI costs out of control. If people are rewarded for using more tokens, they may focus on increasing AI usage instead of producing better work.
A better metric is . The real question should be what was finished or shipped, not how much AI processing was consumed.
Key points
- can be a bad or pay metric.
- Rewarding usage can encourage higher AI bills without better output.
- are a better measure than raw AI consumption.
- Cost matters, but value should be judged by finished work.