AI spend at Meta: why token limits for engineers may be next

I’d like to share a quick note from the AI and product efficiency space.

Adam Mosseri at Meta believes that within 1–2 years, companies may need to cap AI token spend per engineer.

What matters:
• Meta does not have such limits today;
• token costs are increasingly being compared with salaries and other Opex;
• the company has already shut down its internal spending leaderboard to remove noise;
• similar AI budget reviews are happening at other companies too.

To me, this is a clear signal: AI is no longer just an experiment — it is becoming a managed resource.

Do you see token limits as discipline, or as a brake on innovation?

#AI #Meta #HackerNews #Tech