Agenda: 2/5 Sync

When: 2/5, 1:00-2:30PM PST

Where: Meet (see cal hold)

Agenda:

• Warm-up [5 min]
  • Confirm agenda
  • Goal-setting for remaining time. What are the non-negotiables before exiting the call?
• Analysis [45 min]
  • Each person presents three generalized and actionable insights that they learned in their research [10 min, 5 each]
  • Pick one of the insights as a starting point for broader discussion [5 min]
  • Deep-dive (free form time for follow-up questions, pushback, “golden nuggets,” shots from the hip) [15 min]
  • Do any of these insights lead to concrete product hypotheses? If so, let’s make an initial stab. [15 min]
• Process debrief [25 minutes]
  • Outline steps taken for analysis, focusing on use of AI tools [5 min]
  • Discuss what worked well/did not work well (prompting techniques, different models, etc.)? [15 min]
  • Brainstorm Improvements to implement for next time [5 min]
• Wrap-up [15 min]
  • Retro of today’s meeting (proper allocation of time? topics we missed? more engaging/less engaging bits?) [5 min]
  • Align on goals for next meeting (goal setting for next touchpoint, incremental deliverables, logistics of next touchpoint) [10 min]
    • Dustin: what happened in early 2022 supply shock? Produce report. Goal: understand it as a case study of supply/demand for the value chain. Specifics and generalizations, apply generalizations towards other value chains. Understand WTF the volatility in the industry, fundamental industry drivers + how they compare/contrast to oil.
      • Add report to notion, indicate sections B should read
    • Bliss: If we had every piece of data we wanted to understand and forecast supply/demand shocks for value chains, what are the high priority data pieces would we collect, how could we incentivize users to give it up, and what alpha-generating decisions would we enable through that data?
      • Create notion page for prompt sharing
    • EOWeekend - touch base on signal
• Buffer time [10 min]

Running notes:

Up to three hypotheses to test against other 4 supply chains

Very clear logistics for the next week

Dustin to study, quantify economic/business dynamics; bliss to think about data asset

supply and demand

If we had every piece of data we wanted for this, what would we assemble?

What will it take acquire that data?

Studying additional supply chains or not?

Use to verify and not just run at semis

Semiconductors as the New Oil_ A Strategic Supply Chain Analysis.pdf

Semiconductors & Risk_ Viability of Financial, Insurance, and Data Solutions.pdf

Notes:

Meeting Setup & Agenda

• Started late due to scheduling confusion (thought 1:30 vs 1:00)
• Extended meeting by 15 minutes to 2:50 PM hard stop
• Success criteria: Develop up to 3 hypotheses for testing with remaining 4 supply chains
• Agenda: Share insights → Deep dive analysis → Concrete product hypotheses → Next steps

Key Insights from Semiconductor Analysis

Bliss’s Three Insights:

1. Complexity vs. Resilience Distinction
   • Complex supply chains aren’t necessarily unresilient
   • Semis have structural bottlenecks: geographic concentration, high CapEx, human capital constraints, high lock-in
   • Oil comparison: complex but flexible (can reroute, blend sources, move refineries)
2. Resilience as Public Good
   • Foundries absorb shocks by passing prices to customers due to market power
   • No incentive for proactive resilience investment without compensation
   • Government subsidies currently only mechanism driving resilience projects
3. Market-Driven Resilience Opportunities
   • Missing mechanisms that exist in oil: OPEC-style coordination, futures markets, demand indexes, strategic reserves
   • Concept: “Resilience as a Service” (RAS)

Dustin’s Three Insights:

1. Value Chain vs Supply Chain Focus
   • Many value-additive opportunities don’t touch physical supply chain
   • Framework shift needed from supply to value chain thinking
2. Financial/Logistical Infrastructure Gap
   • Semis principally lag in financial and logistical mechanisms
   • Need to quantify volatility including 2020s chip shortage
3. Automotive as MVP Candidate
   • Hit hardest by chip shortage, understands pain
   • Likely willing to pay for solutions

Financial Products as Core Opportunity

• ChatGPT analysis identified missing layers: commodity trading, risk hedging, insurance, strategic buffering, capital formation, market intelligence
• All opportunities except strategic buffering are financial products
• One data product (market intelligence) similar to initial proposal
• Potential to build fintech/financial-first solution rather than pure supply chain tool

Specific Product Hypotheses Discussed

1. Semiconductor Commodity Exchange & Hedging Platform
   • High potential but very difficult execution
2. Independent Trading House (“Glencore of Chips”)
   • Challenges: chips aren’t fungible, depreciate over time, Moore’s law effects
   • Question: Does Moore’s law death make chips more commoditized?
3. Insurance & Risk Solutions
   • Multiple types: business interruption, parametric, yield/output, political risk
   • Example: Chip continuity insurance for automakers (20% delivery shortfall triggers payout)
   • Requires actuarial modeling and projections
   • SageSure model: issue policies, sell to insurers, no liability held
4. Real-Time Supply Chain Intelligence Platform
   • “Waze for chip supply chain” - crowdsourced information sharing
   • Marketplace for excess inventory swapping
   • Risk monitoring and alternative sourcing suggestions

Data Strategy & Business Model

• All financial products built on same underlying data asset
• Eric Schmidt model: collect data through one service, monetize across multiple products
• Entry point strategy: use one service to accumulate data capital, expand to other services
• Potential path: Intelligence platform → Trading → Commodity exchange → Full financial ecosystem

Critical Research Questions

Market Dynamics Analysis Needed:

• Is there actual volatility or just perpetual undersupply?
• Daily volatility vs. black swan events?
• Supply/demand yo-yo like oil, or structural shortage for next 40 years?

Data Requirements:

• What’s the most valuable data for forecasting?
• Likelihood and cost modeling for insurance products
• Incentive structures for data sharing

Next Steps & Assignments

Dustin’s Homework:

• Study 2022 semiconductor shortage as case study
  • One-page summary of what happened
  • Understand fundamental industry volatility patterns
  • Compare/contrast with oil industry dynamics
• Deliverable: Report on supply/demand dynamics and industry drivers

Bliss’s Homework:

• Define optimal data asset requirements
• Economic lens: what’s the most valuable data?
• What high-value decisions need data-informed support?
• How to incentivize data sharing?
• Test framework against 3 other supply chains for verification

Shared Considerations:

• Independent study opportunity with Jonathan Burke on financial tools in semiconductor space
• Decision pending: continue studying other supply chains vs. deep dive on semis
• Insurance expert contact available (SageSure connection)