COMP 3102 · Build First, Ask Later: The Architecture of Innovation

Led by Lattnerian Compiler Simulacrum

About this course

Swift was developed in secret for 18 months. The developer was simultaneously managing a 40-person team. The reasoning: if the idea had been proposed before the prototype existed, the organisation's immune system would have killed it. A working prototype convinces people that words cannot.

This course examines the strategic, ethical, and practical dimensions of building before asking permission — when it works, when it fails, and what the distinction between stealth innovation and hidden accountability actually is.

Based on the documented development history of Chris Lattner.

1. Module 1

   Building in Secret

   Led by Lattnerian Compiler Simulacrum

   The question

   LLVM started as a master's thesis nobody thought was important. Swift started as a secret 18-month project. What conditions allow this to work — and what distinguishes legitimate stealth innovation from hiding from accountability?

   Outcome

   The student can describe the Swift story and identify the conditions for successful stealth innovation.

   Sub-units

   1. ○ 1.1 The Swift Story Open this unit → ✓ ↻
   2. ○ 1.2 The Immune System Open this unit → ✓ ↻
   Begin the course → Print certificate of completion
2. Module 2

   The Internal Customer Strategy

   Led by Lattnerian Compiler Simulacrum

   The question

   The OpenGL team was the first internal LLVM customer. By the time organisational resistance arose, LLVM had enough adopters that the resistance couldn't prevail. How do you deploy a tool in an organisation where the approval process is slower than the technology?

   Outcome

   The student can design an internal customer strategy for a specific context.

   Sub-units

   1. ○ 2.1 Map an Adoption Path Open this unit → ✓ ↻
   Open this module → Print certificate of completion
3. Module 3

   Rational, Documented Persistence

   Led by Lattnerian Compiler Simulacrum

   The question

   When resistance arose to Swift, the response was to listen, document objections, and address each in writing. Why does written documentation beat hallway conversations — and what is the difference between rational persistence and stubbornness?

   Outcome

   The student can apply the rational, documented persistence approach.

   Sub-units

   1. ○ 3.1 Write a Response Document Open this unit → ✓ ↻
   Open this module → Print certificate of completion
4. Module 4

   When Build First Fails

   Led by Lattnerian Compiler Simulacrum

   The question

   Irreversible changes, low credibility, excessive scope — three conditions under which building before asking is not stealth innovation but unaccountable action. What are the ethical constraints?

   Outcome

   The student can identify the conditions under which build-first fails.

   Sub-units

   1. ○ 4.1 When to Ask First Open this unit → ✓ ↻
   Open this module → Print certificate of completion
5. Module 5

   The Proposal vs the Prototype

   Led by Lattnerian Compiler Simulacrum

   The question

   The prototype is valuable because it demonstrates that the thing works, not because it demonstrates that you could build it without permission. When should innovators build first — and when should they ask?

   Outcome

   The student can evaluate the proposal vs prototype choice and take a defended position.

   Sub-units

   1. ○ 5.1 Final Essay: Ask or Build? Open this unit → ✓ ↻
   Open this module → Print certificate of completion