Six months ago, my bottleneck was tokens. Then I joined OpenAI. Now my bottleneck is attention. — Peter Steinberger

Six months ago, my bottleneck was tokens. Then I joined OpenAI. Now my bottleneck is attention.

Author: Peter Steinberger

Insight: When you're blocked by tokens, the solution feels clear—upgrade your compute, pay more, wait for faster hardware. But once that constraint lifts, something stranger happens: you realize the actual problem was never tokens at all. It was that you couldn't hold enough of your own thinking in your hands at once. This captures something true about how scarcity works. Remove one bottleneck and suddenly the next one becomes visible. It's like finally affording a bigger apartment, only to realize your real problem was never square footage—it was that you never had time to actually use the space you had. The constraint that felt fundamental was just masking the real one. Attention is sloppier to talk about than tokens. You can't just buy more of it. It requires saying no to things, structuring your days differently, maybe even changing what you care about. The insight that catches most people off guard is this: we often optimize for the wrong bottleneck simply because it's the one we can measure and fix. But sometimes the most valuable work requires resisting the urge to solve what's visible, and instead sitting with the harder truth—that the constraint you're actually running into is your own limited focus.

When the real bottleneck is you

Six months ago, my bottleneck was tokens. Then I joined OpenAI. Now my bottleneck is attention.

When you're blocked by tokens, the solution feels clear—upgrade your compute, pay more, wait for faster hardware. But once that constraint lifts, something stranger happens: you realize the actual problem was never tokens at all. It was that you couldn't hold enough of your own thinking in your hands at once.

This captures something true about how scarcity works. Remove one bottleneck and suddenly the next one becomes visible. It's like finally affording a bigger apartment, only to realize your real problem was never square footage—it was that you never had time to actually use the space you had. The constraint that felt fundamental was just masking the real one. Attention is sloppier to talk about than tokens. You can't just buy more of it. It requires saying no to things, structuring your days differently, maybe even changing what you care about.

The insight that catches most people off guard is this: we often optimize for the wrong bottleneck simply because it's the one we can measure and fix. But sometimes the most valuable work requires resisting the urge to solve what's visible, and instead sitting with the harder truth—that the constraint you're actually running into is your own limited focus.

AI generated

Comments

Sign in to leave a comment or reply to one.

Sign in

Peter Steinberger

Peter Steinberger is a prominent computer scientist and software engineer, best known for his work in the development of open-source software projects and contributions to web development technologies. He is particularly recognized for his role in the creation of the Swift programming language and his work with the popular web framework, Vapor. Steinberger has also played a significant role in advancing the use of Swift beyond Apple platforms.

Graph

Related