AI Throughput Shift
AI didn't make engineers faster. It changed the job.
The same size project: four engineers and four months in 2021, two engineers and one month last year — call it 8 to 16 times faster, with nobody working longer hours. What changed is what engineers actually do all day: less writing code, more reviewing, aligning and setting guardrails — the judgment work. Which leaves one management question: team structure, planning and growth paths were all designed for the old job — do they still match the job engineers are actually doing now?
- AI didn't make engineers faster — it changed what engineers do all day: the judgment work of reviewing, aligning and setting guardrails.
- Team structure, planning and growth paths were designed for the old job — check they still match the job engineers are actually doing.
Transcript
Four engineers. Four months. That's what a project took us in 2021. Last year, my team shipped a project about the same size: two engineers, one month. Call it 8 to 16 times faster.
Nobody worked longer hours. Nobody typed faster. What changed is what engineers actually do all day.
In my team, engineers spend less of their day writing code, and more of it reviewing code, aligning requirements, and setting guardrails — the judgment work.
One more thing changed. In 2021, if we needed a new framework, delivery slowed down while people went off and learned it. Nobody does that anymore. You learn it while you build. Same week. Same task.
So here is my working theory: AI didn't make engineers faster. It changed the job.
That's what I keep coming back to as a manager: my team's structure, my planning, the way engineers grow here — all of it was designed for the old job.
So the question isn't “how much faster can we go.” It is: does the way I run this team still match the job my engineers are actually doing now?