<aside> 🧭 Revised research queue

Photonics has been added after AI infrastructure because it is the most natural bridge from chips to data movement. These are research candidates only, not open positions or recommendations.

Workflow: score sector → research candidates → pick max 5 → unique catalyst/invalidation → only then open paper positions.

</aside>

Episode Order

1. AI Infrastructure / Chips
2. Photonics / Optical Interconnect
3. Water Treatment for Data Centers
4. Space
5. Prediction Markets
6. Crypto
7. Defense / Drones
8. Nuclear
9. Solar

Alpha Engine Integration

• Hermes Tape supplies public thesis/candidate intake; alpha-engine should track prices, market caps, volatility, beta, drawdowns, P&L, and eventual paper-position decisions.
• Do not turn these into live trades. Use alpha-engine as the data spine, journal, paper ledger, and opportunity-cost tracker.
• Candidate baskets should become watchlist/universe inputs first; opening positions requires a completed scorecard and episode thesis.

AI Infrastructure / Chips

• Core question: What suppliers get paid if AI capex keeps compounding?
• Why now? Compute demand, inference, HBM, networking, and custom silicon are still bottlenecks.