Potential Businesses

  • On-chain product reviews.
    • On purchase, link purchaser_pk with purchase_id
    • Can incentivise users in middle NPS to give reviews
    • Open access to analyse bad reviewers.
  • Open-source alternate software platforms tooling that are (at least) one of: {10x cheaper (e.g. use cloud storage), 10x faster (e.g. rewritten in Rust, better threading/SIMD)}.
    • Current examples
      • Dragonfly DB: better threading, caching and other decisions make it 25x more throughout than redis.
      • OpenObserve: ElasticSearch but 140x cheaper via object store, stateless nodes, Rust.
    • Possible opportunities:
      • kdb: Implement KDB’s q programming language, backed on clickhouse or other open source tools. Sell enterprise support, or common cloud platform with the self-hosted data. Apparantly the price of KDB starts at 25k for 2 cores, 100-200k/core (source).
  • App “once” Store for Open Source, full
  • Grandparents will be early adopters for Web3 products.
  • Farcaster for Real estate
    • Two parts:
      • Farcaster protocol: Build core protocol and network for real estate interactions (i.e. houses as NFTs, interactions with properties, user identity, signature and verification)
      • Warpcaster: Boostrap protocol by building the first commercial application that provides value. Encourage others to build on the protocol.
    • Current Real estate DeFI build projects/products that aren’t sufficiently decentralised –> don’t scale.
  • Robotic arm (or 3D printer like) painting
    • Simpliest process: Image, k-means, for each k: paint a dot per pixel in cluster, change to next colour. Optimise colour order and choice to minimise paint reuse.
  • Generative AI version of unsplash.com but with proper credit/profit assignment.
  • Sports “asset” marketplace for a new betting paradigm
    • e.g. $8.2 / stock of NBA team, 10c dividend per win
  • Use generative AI for “live” sports matches.
    • Revenue via betting
    • Start with simpler dynamics with good betting markets (e.g. horse/dog racing)
  • Platform for valuable aggregate data that individuals have already made, e.g:
    • Applications to ycombinator (un/successful)
    • College applications
    • Amazon purchasing history, used-goods marketplace
    • In sum: Any platform/monopoly that gets data from individuals that could be valuable to other businesses.
  • Parameteric climate insurance for individuals + SMBs
  • Proof-of-work email spam prevention
  • QR code coffee cup advertising marketplace
  • Introducing a ‘LLMLIKE’ SQL operator for text embedding search
  • Edge distributed, AI search engine service
    • Search query/results highly cacheable (~1% search terms -> ~60% search frequency)
    • LLM model size can be reduced, and will continue to.
    • Large tail of customers do not have large data; traditional/current vector search engines are going after large enterprise.
    • Read only on cache, async write
  • Low earth satelite cattle tracking ear tags
  • Hand curated tech news platform
  • AI podcasting tools
  • Realtime satelite camera data for:
    • Wild Animal Tracking
    • Wildfire Detection
  • Ecommerce Recs-as-a-Service
  • P2P Coupon exchange platform
  • Beer-to-good-people beer-company advertising campaign

Open-source

  • IaC framework for automatic SOC 2 compliance.
    • E.g. A terraform provider/plugin such that all infrastructure provisioned is SOC2 compliant by default. May have to be more than just IaC.