Ask HN: Corporate Disconnect Between "Tokenmaxxing" and Token Optimization

About 6 months ago I joined a new team within a top ten F500 company. My new boss strictly mandated AI use with the key principle being: "You shouldn't be manually writing any code".

Since then its been all agents, skills, MCP, harnesses, custom in-house frameworks, and running Opus 4.7 high non-stop.

Now at a company level, there are "encouraged attendance" workshops getting scheduled to learn how to optimize one's token use now that API pricing is becoming the norm at an enterprise level.

The direction I got from my direct leadership was very direct. AI / agents lead everything and the expectation is the team moves as quickly as it has. But the truth is most engineers candidly acknowledge: we don't fully understand anything. Especially because agents are also churning out the content of architectural documentation and user story requirements and acceptance criteria.

I feel like this is a situation where I am directly responsible for the non-deterministic output of these tools. The solution I get from my supervisor for any problem literally boils down to "You need to use an agent, skill, etc.".

Is anyone else going through this tug of war? How is it going?

4 points | by mc-0 8 hours ago

3 comments

  • hiroto_lemon 6 hours ago
    What made accountability tractable for me was treating agent output as untrusted input — the invariants I own (cost caps, tests, contracts) get enforced out-of-band, so the non-determinism stays bounded.
  • Sasisundar09 3 hours ago
    We're building MCP server reliability monitoring — has anyone been burned by silent failures? Would love to talk.vouqis.tech
  • throwaw12 7 hours ago
    > Is anyone else going through this

    I thought this became the norm already, isn't the whole industry working in the same way at the moment?

    [-]
    • apothegm 7 hours ago
      Only in the bleeding edge/kool-aid drinking bubble. I’m in a company that’s taking a gradual approach and supporting us with budget when we want to explore ways it might make us more efficient, but not setting any mandates. And we still have the expectation of understanding the output as if we’d written it.