Project TBD Memory Vault

Golden Nuggets 🍗

4 min read

Golden Nuggets 🍗

1/29, Oil Meeting:

  • Data asset can be exchanged for additional customer data that will strengthen the original asset, creating a self-reinforcing cycle
  • Analyzing supply chain re: AI in a top down and bottom way. Top down - how does AI change the exogenous factors determining outcomes of oil revenue streams; bottom up- how does AI change the way those revenue streams can operate and boost productivity in relation to each other

2/5, Semis-meeting

  • What if the solution to completing the semis value chain is a fintech problem (exchange, marketplace, derivative, hedging, etc.)
  • As Moore’s Law falters, do chips become more commoditized? → enables reserves, storage, risk transfer through time

2/7

  • Met with Eoin from Mundane robotics who was complaning about how he needs a software platform for supply chain intelligence/simulations + the ability to overlay it with macro OSINT. Happy to chat with us
  • Should have a 3 hour strategy session just thinking about creative ways to incentivize contribution of data
  • Avoids competition (a la Thiel)

2/18

  • Are chiplets the equivalent of blending oil?
    • multiprocessors/cores atomic particle of GPUs and CPUs → microprocessor index rather than GPU index
  • Half live of chips combination of physical properties + market assessment (we think)

2/21

  • The companies that are built AI-native will out-compete all of the other companies — we need to push the envelope in terms of AI integration and utilization
    • Legacy software is struggling because they are not built AI-native — they literally cannot adapt to an AI mentality and way of working
    • We have a real opportunity to out-compete if we can leverage AI better than other companies
    • This is per conversation with Jonathan Heiliger; he’s seeing his portfolio companies who are performing well simply not able to keep up because they’re not adopting AI as fast as the competition

2/27

  • Companies pay more for things that help them grow revenues vs shrinking expenses
    • We need to create something that enables growth
  • Compliance as standalone platform or third-party integration, or both

3/2

  • We should create [price or other] index based on the data asset we build, and then issue parametric insurance policies based on this proprietary index (or indices)

3/4

  • Macro tailwind: the US on both sides of the aisle is likely leaning into industrial policy
    • That means more compliance and more need for risk transfer mechanisms
  • Backup idea: Can you create some sort of financial/risk product that derisks manufacturing (pool risk/more efficiently sell hard assets should things go wrong)

3/8

  • The future of commerce and purchasing will be driven by agents
    • What does the world look like when it’s being driven by agents?
    • What are the implications of building a product that is intended to be sold to mathematical agents instead of people?
  • Ontology/digital twin — build a product around/incorporating this
  • Polymarket is going to be a part of this
    • Could we package polymarket bets into financial instruments that purchased by institutions?
    • combine polymarket with the digital twin to constitute middleware between the polymarket and wall street (ie laundering polymarket)

GOLDENEST NUGGET SO FAR

  • We are building a multi-tenant proprietary digital twin/ontology for the semiconductor industry
    • Our MVP is an agentic simulation of shocks (built on hand wavy data)
      • Product roadmap…
        • agentic simulation of shocks
        • alerting
        • prediction
        • remediation
        • recommendation marketplace based on visibility across multiple digital twins
          • Transaction recommendations (buy from this guy)
          • Insurance recommendations (issue custom parametric policies based on digital twin data; as MGA; sell off liability)
          • Hedging/derivatives/risk exposure (potential polymarket integration)
    • We waive the shiny digital twin agentic simulation model in front of peoples’ face as a way to get in the door, and then sell compliance as the first step in getting there
      • Vectors of compliance…
        • Supplier onboarding portal (can’t do business with restricted entities)
        • Shipment diligence (import/export control; can’t import/export controlled things)
        • Mandatory disclosures (have to file paperwork)

5/14

  • Why don’t we just build a new data-driven factory? IT fits the right to compete of capex physical assets for Dustin and data for Bliss

Graph View