2024-06-20-Thu

CuspAI

role: couple of things to think about. excited about someone like me joining. phil shaped hole in the org. no-brainer.
1. timing (when is the right time for Cusp) → early next year
2. affordability - not MSFT but
MSFT:
- thinking about compute: easy piece
- more interesting: relationship with Meta and others
  - MSFT largest carbon credits because of consumption
  - motivated to reduce cost of capturing carbon as low as possible
    - bring together Azure / MSFT / Meta / we fund Cusp’s research
    - partnered with best known carbon capture companies. develop
    - pay cash and compute for keeping cusp afloat. some revenue
      - carbon capture uses material
    - cost of carbon capture comes down
    - everyone wins
- would we engage? in capital or credits
  - Chris Bishop: can we work together?
    - they will speak in person

2025-03-11-Tue

last spoke in july
using 30m seed
now 27 people.
more joining in the coming months
treading a fine line on building vs. hiring
working together.
team is in a good place.
building the platform now
building a search enging for materials. 80% of way through building
take specifications, and take them througha. pipeline
agents building
first early customer engagements
accelerated prpogram: take mateial to ai design to testing to carbon capture facility in 6 months.
economics model / isomorphic labs?
technical risk
expanded application space: semiconductors and carbon capture and others.
cc is inherently risky
spent a long time with Microsoft - very frustrated - working with Jon Tinter’s team
- strategic partner but very slow movement
- Paul James in the UK
- build on Azure, sign up to compute deal, do an investment
doing more building than researchsolutions are understood
can do simulations at x cost, can train models and search with them
have to get as much search out of the experiment and into the computer
all the work to build the pieces and put them in the well-oiled platform.
might see opportunities on the research side.
a whole new way of doing x might emerge
dot on the horizon is finding new materials - point the cannon at those materials
rfp for dac for datacenters