翻訳待ち：The Sequence Opinion #909: Return on Token: The New Economics of AI-Native Engineering

AI サービスが一時的に利用できないため、復旧後に翻訳を補完します。

For most of software history, engineering capacity was easy to sketch on a whiteboard. You had a certain number of engineers. Each had a certain amount of time and talent. The basic equation held: Engineering capacity ≈ people × time × talent. AI breaks that equation. An engineer can now assign one agent to investigate a production bug, another to write tests, a third to prototype an architecture, and a fourth to document the result. The agents can run for hours and work in parallel. They do not appear on the org chart, ask for equity, or attend the planning offsite. They do, however, consume tokens. The modern engineering organization now has a second, elastic workforce. The human workforce is measured in headcount. The machine workforce is measured, imperfectly, in tokens. And because companies love measurable things—especially things that produce dashboards—we have entered the era of token maxing. Token Maxing Is the New Lines of Code Read more