01

You cannot freely experiment with agentic pipelines

A developer receives a specific subscription, not a company compute budget. They therefore explore the best way to fit a task into the tool they were given—not the best way to solve it.

Trying Claude instead of Codex means requesting another subscription, waiting for approval, and explaining why it is needed. If the experiment fails, switching back is just as awkward. Even within one subscription, the quota must be saved for current work: today’s task still has to be finished, and the search for a better process competes with it for the same tokens.

Alternatives that may be fundamentally more effective than the familiar process
Plan → Act → Review Claude or Fable builds a plan, Codex implements it, and Kimi independently reviews the result.
One powerful agent One expensive model receives full context, suitable tools, and project skills.
Swarm Ten expensive agents propose solutions independently, and a final agent synthesizes the best parts.

The last option may be expensive in compute yet cheap in human attention: the task is completed in one shot, and the developer spends fifteen minutes accepting the result. But this hypothesis cannot be tested within one personal quota.

A personal quota distorts the selection criterion: instead of searching for the most effective approach, the developer tries not to exhaust the available limit. They end up attached to the first accessible process and improve it locally without testing fundamentally different alternatives.