arbitrage-audit

## The Framework Five types of gaps AI is closing: speed, reasoning, fragmentation, discipline, knowledge asymmetry. When a gap closes, a new one opens upstream — always toward judgment, taste, relationships, and system-level thinking. **The 7.6% vs 92.4% distinction:** Access to AI is universal. Advantage is not. "Bolted AI onto existing process" vs "rebuilt process around what AI makes possible" is the gap that separates winners from losers. Systems compound; pilots don't. ## The 3 Diagnostic Questions ### Question 1: What inefficiency is this built on? Name the gap. Every business, product, or career advantage sits on top of an inefficiency — an information asymmetry, a speed gap, a fragmentation problem, a knowledge bottleneck. Ask: "What would this business/product/role look like if perfect information existed and everything happened instantly?" The distance between that hypothetical and reality IS the gap. **Gap types:** | Gap Type | Example | AI Closing Speed | |----------|---------|-----------------| | Knowledge asymmetry | "Only specialists know this" | Fast — AI democratizes knowledge access | | Speed gap | "This takes weeks of manual work" | Fast — automation and agents | | Fragmentation gap | "Data lives in 5 systems" | Medium — integration tools accelerating | | Discipline gap | "People skip steps" | Medium — AI can enforce process | | Judgment gap | "Someone needs to decide what matters" | Slow — still requires human taste | | Relationship gap | "Trust takes time to build" | Slow — fundamentally human | | System-level thinking | "Seeing how pieces connect" | Slow — AI assists but doesn't replace | ### Question 2: How fast can AI close it? Not "can AI do this?" but "how quickly does the gap shrink?" **Informational gaps** (research, synthesis, aggregation) close fast. If your advantage depends on knowing something others don't, and that knowledge is in text or data, the clock is ticking. **Structural gaps** (judgment, taste, domain expertise, relationships) close slowly. If your advantage depends on making calls that require experience, context, and the ability to say "no" — that's durable. Ask: "If a smart person with Claude and six months of context tried to replicate this, could they? How long would it take?" ### Question 3: What new gap does the closure create? When one gap closes, a new one opens upstream. Always. - When AI closes the knowledge gap, the NEW gap is: who can distinguish good knowledge from confident-sounding wrong knowledge? (Judgment) - When AI closes the speed gap, the NEW gap is: who can decide what's worth doing fast? (Taste) - When AI closes the synthesis gap, the NEW gap is: who can decide what the synthesis means for THIS situation? (Context) Ask: "Once this gap is closed, what becomes the new bottleneck? Who owns that bottleneck?" The answer tells you where to position. ## Output ```markdown # Arbitrage Gap Audit: [Industry/Business/Role] Date: [date] ## The Gap [What inefficiency this is built on — named specifically] ## Gap Type [Knowledge / Speed / Fragmentation / Discipline / Judgment / Relationship / System-level] ## Closing Speed [Fast / Medium / Slow] — [rationale] ## Upstream Gap [What new gap opens when this one closes] ## Position Assessment - **If you're in the closing gap:** Migrate upstream. Build on judgment, not information. - **If you're in the opening gap:** You're positioned well. Encode your advantage into systems. ## Recommended Action [One specific thing to do based on the audit] ``` Based on Nate Jones' Reinvention of Arbitrage framework. ## Next step Gap mapped → run **kill-go** for a proceed/pivot/kill verdict, or **decision-brief** to write the positioning up. --- Part of the [ThinkHaven Method Kit](https://github.com/hollandkevint/thinkhaven-method-kit) by Kevin Holland. Full Board of Directors experience: [ThinkHaven](https://thinkhaven.co/try)