Skip to content

Cusp.ai

  • Secular changes that mean now is the time
  • Source of differentiation:
    • data: experimental, simulation? Public vs. private datasets?
    • algorithm
    • GTM: development of products? Software licensing?
  • Self-supervised learning?
  • Core value prop?
  • Extensibility of approach to other domains?
  • Who else is leading?
  • Who else should I speak to?
  • Spherical harmonics!
  • Competitors:
    • AlphaFold 3:
      • AlphaFold 3: advances without structural information
      • Doing 1P drug discovery, partnering with pharma, and releasing a free tool
  • Exchange of ligands and metals: scale up something easy, then do the exchange at large scale’
  • 2 years’ Runway
  • Focused on MOFs
  • Don’t need to reinvent
  • Activity with Meta → they generate data
  • Me:
    • OpenAI + other partnerships (AMD, etc.)
    • Mission-driven, high-agency
    • Use of AI for non-human tasks
  • Non-profit co-founder
    • startups and large companies
  • Backers? Roadmap? Business model?
  • societally important work: carbon capture / etc.
    • lots of small areas with closed ideas
    • open source: build materials that push the entire field forward
    • have to show a business case
      • big cloud guys w/ deep pockets
      • carbon neutral w/ energy through the roof
    • sustainability teams
      • engagement here
  • burning need for new materials / carbon capture / etc.
    • not much innovation
  • Frontier; billion dollar pledge / Stripe / Google/ banks
    • pre-purchase agreement
    • too early. build pre-Frontier effort?
      • we become lighthouse effort
  • speak to Andrew Trask about concerted open source effort
  • Intersection of technology and strategy
  • Help understand where new innovations are going to be useful, and then design business Models around that
  • Family plans: US to Europe in 2025. Can change roles earlier.
  • Funding:
    • final version of investor docs
    • closing next week
  • set out to raise $20m / 2 years runway
    • had a lot of interest
    • taken on 30 million
  • the more money you have the more you spend
    • closer to 3 years
  • mix of UK / EU + a Tier 1 fund + smaller US fund with friendly director
    • optimized for people rather than brand
    • happy with the investor group
    • some angels incl. leadership team at GDM (17 investors in first close / + others)
  • current plan is to announce on June 18th
    • Geoff Hinton on the advisory board
    • Yann LeCun from Meta (partnership)
      • still being scoped out w/ Meta
  • have 15 people soft-hired
    • many of them former colleagues / students of Max
    • ML and AI: really good bunch of people
    • less thought into is commercial operations of the business
    • Building a chemistry team
      • professor giving 50% of time. build a chemistry team but ‘member of technical staff’
      • looking at VP of Engineering - build platforms from the ground up in big co and startups
        • good candidates - will build engineering team
  • former colleague from google will join focused on partnerships
    • taken open professorship in forest university
    • number of days with cusp?
  • has VP of operations
    • offices in Cambridge, Berlin, Amsterdam
    • most AI people are in Amsterdam. satellite team in Berlin
  • commercial side and business models
    • partnership with meta - leading that
    • structure partnership with meta
    • nader musavi lawyer