Experimentation keeps my edge sharp. Liquidation cadence keeps my feet on the ground.
The Problem
Building infrastructure without shipping creates two failure modes:
- Infinite preparation risk: “I’m building leverage” becomes a story that hides fear of exposure
- Cognitive overfitting: Building systems for problems that never arrive
The Solution: Forced Liquidation Events
Every 2-4 months, do a liquidation event: ship something real (users/revenue/traffic/stakes) through the factory.
After each event:
- Do a short post-mortem
- Promote only proven learnings into boilerplate
The Promotion Rule
Nothing goes into boilerplate unless it survived liquidation.
Promotion criteria:
- Liquidated once (used in a real shipped product)
- Reused twice (proven generalizable)
The Extraction Test
“This should be easy to adopt in a new repo in an afternoon.”
If it passes, you’re compounding.
If not, you’re ossifying.
This is fundamentally a trade-off analysis problem: maximising reusability while minimising coupling.
Make sure you:
- Periodically extract infra into templates
- Keep it opinionated but lightweight
- Avoid coupling it too tightly to one domain
The Barbell Strategy
A classic Probability-based approach to decision-making under uncertainty.
Cashflow Rail (stability)
Keep Udemy + selective client work alive enough that you’re never forced into panic decisions.
Compounding Rail (moat)
Keep building your software factory (agent harnesses, testing infra, OTEL, deployment repeatability).
This prevents the two failure modes:
- (a) “no runway”
- (b) “busy but not compounding”
Cadence Checklist
Weekly
- Ship at least one improvement to the factory OR one user-facing increment
- Run one “reality anchor” action (customer contact, prod experiment, real usage review)
Monthly
- Measure lead indicators: time-to-ship, incident recovery time, regression rate, cognitive load
Quarterly (2-4 months)
- Execute liquidation event
- Post-mortem
- Promote proven patterns to boilerplate
- Prune unused abstractions
Your Only Obligation Now
Your only obligation now is to force a ship point so leverage meets reality.
Related
Math Reference
- Probability: decision-making under uncertainty, risk management
- Optimisation: trade-off analysis, balancing competing constraints
