The Surgeon Management Model
Spend over half your time making your top 10% feel like a supported surgeon who can't be blocked.
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 4
- Confidence
- 88%
Borrowing from The Mythical Man-Month's image of the software team organized like a surgical suite — one person doing the work while everyone supports them — Sherwin Wu builds his management around making each top performer feel like that surgeon: fully supported, with an unseen army looking around corners and pre-empting blockers. He deliberately spends more than 50% of his time on his top ~10% of performers, which matters even more now that AI removes the coding bottleneck and leaves organizational blockers as the main drag on shipping.
Origin
Adapted by Sherwin Wu from Fred Brooks' 'The Mythical Man-Month' (the surgical-team metaphor); reinforced by Marc Andreessen's line that AI makes good people better and great people exceptional.
Core principles
- 01AI makes good people better and great people exceptional, widening the spread of team productivity
- 02High-agency top performers compound their leverage most with AI tools
- 03With code no longer the bottleneck, organizational and process blockers become the main constraint on shipping
- 04The manager's job is to be the invisible support staff, not the surgeon
How to run it
- 1
Identify your top ~10% performers
Single out the highest-agency people who lean into tools and ship the most.
Pro tip In an AI world these are often the engineers going deep on tool best practices — a very high-leverage thing to let them do.
- 2
Allocate the majority of your time to them
Spend more than 50% of your management time with that top tier — unblocking them, keeping them happy, productive, and heard.
Watch out Don't default to spending your time on the lowest performers; the leverage is with the top.
- 3
Look around corners and pre-empt blockers
Anticipate organizational and process obstacles before they hit, so the tool or decision the person needs is already waiting — the scalpel ready before the surgeon asks.
Pro tip Try asking an AI hooked to company knowledge (Slack, Notion, GitHub) 'what are the active blockers on my team, and what will block us next month?' to surface and even anticipate blockers.
- 4
Let them set standards and share them out
Give top performers room to explore best practices, then propagate what they discover across the org via docs and knowledge-sharing sessions.
In the wild
Rather than manage the team maintaining a fully Codex-generated codebase closely, Wu let them 'kind of rip' and see what emerged, treating them as surgeons to support rather than direct.
→ The experiment paid dividends and generated best practices worth sharing across the org.
Common mistakes
Spreading time evenly or toward low performers
Distributing attention equally under-invests in the people whose AI-amplified leverage produces the most output, flattening the exact spread that top performers create.
Acting as the surgeon instead of the support staff
A manager who inserts themselves as the one doing the work removes the sense of ownership and support that makes top performers feel empowered, and becomes a bottleneck instead of a blocker-clearer.
Is it for you?
Best for
Engineering managers whose ICs now ship fast with AI and whose main constraint is organizational friction
Not ideal for
Early-career or struggling teams where the majority need heavy coaching rather than top-tier empowerment
From the transcript
“it's like more than 50% of your time with your top performers with maybe your top like 10%”
“looking around corners and unblocking people especially from an organizational perspective is extremely extremely useful”
“AI makes good people better and it makes great people exceptional”
“if the surgeon needs scalpel but you know the manager kind of already has a scalpel ready for them that that's the best case scenario”
From the episode
“Engineers are becoming sorcerers”
The future of software development with OpenAI’s Sherwin Wu