1. Introduction: Valuation Chaos

• OpenAI valued at $300B by SoftBank, $157B by Microsoft, $97B by Musk.
• No systematic framework exists for AI company valuation.

2. The Need for a New Framework

• Traditional ARR multiples don't capture AI's rapid obsolescence risk.
• AI companies face "existential threats" where value can go to zero overnight.
• Displacement Risk Factor (D): 0-1 scale measuring probability of core value being replaced.

3. Zero-Value Threshold

• Foundation models: when open-source catches up in performance.
• Applications: when foundation models replicate their core features.
• Value drops vertically (cliff), not gradually.

4. New Valuation Formula

• Value = ARR x Multiple x (1 - D)
• D depends on: technical lead over open-source, efficiency, defensibility of specialized features.

5. Open-Source Catch-Up and the "Death Zone"

• Open-source models close the gap in 5-22 months (trending shorter).
• Usage drops cliff-like when open-source reaches parity.

6. Technical Moat Limitations

• Technical moats last quarters, not years, due to talent mobility and paper publication.
• D estimates: Tech leader (6+ months ahead) = 0.1-0.2; Marginal lead = 0.3-0.5; Parity/behind = 0.9-1.0.

7. Escape Strategies

• Domain specialization, infrastructure pivot, retreat to application layer -- all involve ~90% value decline.

8-10. Valuation Application

• OpenAI bull case: $208.8B; bear case: $139.2B.
• Only 2-3 companies can maintain tech leadership; the rest risk falling into the death zone.

"The only unforgivable mistake is misunderstanding the fundamental dynamics that determine whether your company becomes zero or hero."