← all repositories
steven2358/awesome-blockchain-ai

A field guide to the blockchain-AI Venn diagram

A curated list of projects trying to decentralize, monetize, or otherwise ledger-ify machine learning.

1.1k stars Learning
awesome-blockchain-ai
Velocity · 7d
+0.4
★ / day
Trend
steady
star history

What it does This is an awesome-list that catalogs roughly two dozen projects and a dozen academic papers at the intersection of blockchain and AI/ML. It sorts them into buckets: algorithms, data marketplaces, compute marketplaces, finance, medicine, supply chains (empty), autonomous agents, and research papers.

The interesting bit The list is a useful temperature check on which buzzwords are still breathing. You can watch the field evolve from 2017-era “decentralize all the things” (SingularityNET, OpenMined) to 2024-era “AI agents with on-chain identities” (Hashgraph Online, AgentFund). The academic section is arguably the most concrete part—actual papers on proof-of-learning consensus mechanisms and energy-recycling blockchains.

Key highlights

  • Covers live projects (Bittensor, Fetch.ai, Ocean Protocol) alongside dormant ones (Raven Protocol, Thought Network)
  • Includes Microsoft’s 0xDeCA10B framework for training models inside smart contracts
  • Academic research section spans 2019–2021, heavy on “proof-of-useful-work” variants
  • CC0 license—no attribution friction if you fork it for your own research
  • Supply chain section is currently blank, which is its own kind of signal

Caveats

  • Curation appears static; no clear update cadence or last-reviewed date
  • Project descriptions are mostly copied from marketing copy, not independently verified
  • Some listed projects use words like “revolutionize” unironically (the README quotes them faithfully)

Verdict Grab this if you’re writing a literature review or trying to distinguish the dozen legitimate projects from the hundreds of whitepaper graveyards. Skip it if you want working code—this is a signpost, not a toolkit.

heatdrop uses Google Analytics to see which pages get read — nothing else. Your call. How we handle data.