Adam Mosseri predicted that AI tokens for one engineer could soon cost as much as their salary, pushing Meta to consider per-engineer spending caps. The conversation explores token-spending leaderboards, internal competition, and how monitoring AI spend could hurt morale in ways companies have learned from before.
Meta's Adam Mosseri says AI token budgets could soon be capped per engineer. So in a recent interview, Instagram head Adam said he can see a time in the future, perhaps only a year or two, when putting limits on Meta employees' AI token spend will become necessary. Quote, I think that you can imagine at least in a year or two that the burn rate of a strong engineer might be the same as their salary or their cost of employment, and in that world you're going to probably need to put in some caps. That's really interesting thinking about a cap per engineer and then subsequently like what does a leaderboard look like for that? And are we just getting into a commoditization race internally at these companies? And I wonder, you know, I know a lot of Meta engineers, they were really depressed when Zuck rolled out, hey, we're going to monitor everything you do and that's our IP. I can only see the sentiment in the engineering world, they're going even lower with these types of statements coming out. What do you think about this type of statement from from Meta's Adam Mosseri here running the engineering team. Token caps per engineer.
I think it's pretty predictable where we're at. The same organization that had token maxing leaderboard is now like, wait, we have to put constraints on this. We're in a learning phase right now, right? And so the amount of tokens that you're spending is really a justification for how do we stay at the front of the pack and how do we know where to invest? But there's not an ROI case for it right now. So a lot of the companies are just saying, let's use this token investment for training. And so why wouldn't you regulate it like any other training budget that you have per employee?
