Resources for the AI Economy
Vitalik categorizes the intersection into four distinct areas: AI as a player in a game, as an interface, as the rules, and as the objective. He dives heavily into the risks of adversarial machine learning if models are fully open-sourced, and explains why decentralized market-based RLHF (Reinforcement Learning from Human Feedback) is one of the most promising avenues for the space.
A pragmatic look at how AI removes friction from data processing, while crypto provides the settlement guarantees AI agents need to transact autonomously. It breaks down the necessity of universal cryptographic identity for AI agents and the use of smart contracts to enforce strict, objective transactional boundaries on automated bots.
A heavy, data-driven report looking at actual funding flows, the explosive growth of decentralized physical infrastructure networks (DePIN), and early consumer-facing integrations. It is excellent for tracking the raw capital migration toward AI-crypto projects and understanding which sub-sectors are capturing liquidity.
Andreessen Horowitz's crypto arm argues that as the open web collapses into an "AI prompt bar," decentralized networks are required to prevent total corporate centralization. The research focuses heavily on "Proof of Personhood" infrastructure to fight deepfakes, distributed compute markets, and using blockchain to preserve persistent context for AI agents as they move across platforms.
A follow-up to his written blog, Vitalik Buterin unpacks how AI is transforming on-chain governance through "distilled human judgment." It digs into the technical reality of using Zero-Knowledge (ZK) proofs and Trusted Execution Environments (TEEs) to secure advanced AI-crypto interactions without leaking proprietary training data.
Published directly to arXiv in June 2026, this landmark 30-page survey catalogs the structural synergies and vulnerabilities of the convergence. It highlights how AI can make smart contracts more autonomous, while starkly examining the risks of AI-powered agents enabling market collusion and the necessity of blockchain to secure verifiable data pipelines for model training.
A June 2026 breakdown of Variant's latest $222M fund thesis, which pivots entirely from the Web3 "ownership economy" to "autonomy." It details the rise of the Agentic Economy, arguing that AI agents will soon become the primary drivers of internet traffic and will require decentralized crypto rails for money, persistent identity, permissions, and cross-platform ownership.