← sources

Massive AI bills, for what?

source: reddit (reddit:ExperiencedDevs)published 2026-10-14 22:50ingested 2026-10-15 02:38

retained text

Reposting. Last post was deleted because I didn’t have a user flair.

I work at a non faang big tech company. Like everybody else we have a way to track ai usage/cost. The monthly trend seems to be half in the low three figures while the other half are in the low four figures with very few people in between. The higher group spends about 7 - 8x more. I fall into the former group. but there doesn’t seem to be any difference in productivity between the two groups. And in fact engineers from the lower group routinely seem more productive than the higher spenders. I have two terminals working on tasks continuously through most of the day and can’t break 4 figures. Having more than 2 - 3 concurrent tasks is just too much for me to keep up with. For people that fall into a higher spending bracket what is it about your workflow that’s causing this? And has anyone else noticed a similar trend? For people in management, what is your perspective on employees spending thousands a month without very visible returns on productivity? And conversely how do you view the engineers not spending as much?

Edit: when I say productivity I’m not referring to metrics like pr count or loc. I think those are terrible ways to measure impact. More along the lines of how well are people pushing projects forward, collaborating, how well do they understand the system etc

document fetched 2026-10-15 00:27 · canonical https://reddit.com/r/ExperiencedDevs/comments/1x65khe/massive_ai_bills_for_what