Why Unicorn Logic Breaks in AI, and What Investors Should Do Instead

Machines Can Think brought together investors, builders, and policymakers in Abu Dhabi in January to examine AI’s most pressing questions. Managing Partner Dany Farha joined Eddy Farhat of e& Capital and Ahmad Ali Alwan of Hub71 to explore whether traditional unicorn investment logic still holds in the age of AI.

Unicorns historically scaled by building compounding moats through network effects, switching costs, and distribution advantages. AI disrupts this model at both extremes. Foundational models require billions in capital before generating revenue. At the application layer, teams of 5-20 people can hit $100M ARR in under two years, yet moats can erode in months as commoditization accelerates and big tech moves with startup speed.

The conversation surfaced critical tensions between hype and reality:

• Why inflated valuations hurt talent retention in the long run
• How to measure durability when moats erode faster than revenue compounds
• Where regional investors should focus when they can’t compete on foundational models

Watch the panel discussion to understand how conviction-driven investors are navigating AI’s supercycle.