Deliberate generalism

Spin-out from AI-Native Product Building. Informed by A reading list for generalists (Dylan Bowman).

Claim

The shortage is not "people who know a little about everything." It is people who deliberately practice the meta-skills that let them own cross-cutting work — the projects specialists will not start because no lane clearly owns them.

When building is cheap, this matters more: AI widens ideation; choosing and driving the right problem still requires strategy, resourcefulness, and legible progress.

Three skill buckets

Bucket	Question it answers	Examples from the list
Dispositional	How do you show up?	Relentless resourcefulness, agentic bias, impatience with idle waiting
Strategic	What do you do with what you know?	Humans are not automatically strategic; do things that do not scale for information
Project leadership	How do you ship on the ground?	Run major projects, make work legible, scope timelines for managed IC work

Organizational psychology is a fourth lens — accountability sinks, identity traps, how groups actually signal — but the handbook use is picking problems worth owning.

The decision

Act as a generalist when the work sits between functions, the falsifier is unclear, and progress requires pulling context from users, code, and economics in one thread.

Narrow to specialist mode when the failure mode is technical depth — security, data correctness, performance — and breadth would blur the gate.

Pressure-test

• Are you a generalist because you avoid committing to a problem bet, or because you can close loops others drop?
• Can you name what you learned this week in behavior counts, not activity?
• Is your strategy explicit — or are you "busy" without a falsifier?
• Does your work stay legible to collaborators and future you (docs, ADRs, metrics)?
• Are you optimizing a local metric while the cross-cutting project stalls?

Related

• Handbook: Finding problems
• Handbook: Mindset
• Handbook: Iteration
• A reading list for generalists — full bibliography