Tokenization of Latent Computing Power
summary
A lightweight compute marketplace posts matches and deals on-chain and records inputs, outputs and mediation results as an IPFS CIDs sequence so anyone can fetch and reproduce work; a solver market-maker matches client jobs to node offers, nodes run Dockerized modules (CPU or GPU) and return results to IPFS, demonstrated with a PDF→Markdown pipeline and Stable Diffusion runs that complete via the CLI. Integration with Filecoin-backed storage economics is proposed to fund long-term public datasets and tokenized contributions.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:23What is Co-op Hiive
- 1:47Project Bravo and Lilypad
- 2:53Co-op Hiive marketplace
- 5:19Run jobs with CLI
- 6:03History of volunteer computing
- 9:13Limitations of client-server
- 10:56Reproducibility crisis
- 11:59IPFS in decentralized storage
- 13:00Filecoin economics
- 14:28Reproducibility data structure
- 15:50Hackathon prizes
- 16:34Best public datasets use
- 17:16Best AI applications
- 17:57Tokenizing computing power
- 19:22Monetizing scientific projects
- 21:12GitHub setup guide
- 23:42PDF to Markdown converter
- 24:45Q&A
related talks
25m🛠 Making Data Useful with Decentralised Compute
31m🛠 Building Applications with Filecoin & FEVM
09mScaling Compute Supply 10x Through Crypto-Economic incentives
54mIPFS & Filecoin 101
01h 04mBuilding Web3 Startups
14mFilecoin - Building with the Filecoin Onchain Cloud
25mBuckets of Data: Strategies for Leveraging Off-chain Data Storage
15mThe State of Filecoin