Ethereum co-founder Vitalik Buterin‘s latest vision for Ethereum’s intersection with artificial intelligence sees the two working together to improve markets, financial safety and human agency.
In an X post on Monday, Buterin said his broader vision for the future of artificial intelligence (AI) sees humans being empowered by AI, rather than replaced, though he said the shorter term involves much more “ordinary” ideas.
Buterin pointed to four key areas where Ethereum and AI could intersect in the near future: enabling trustless and/or private interactions with AI, Ethereum becoming an economic layer for AI-to-AI interactions, using AI to fulfill the “mountain man” ideal by verifying everything onchain and improving market and governance efficiency.
Source: Vitalik Buterin
Buterin argued that new tooling and integrations are required for AI use to be truly private, without leaking data or revealing personal identities.
Private data leaks by large language models (LLMs) have become an increasing area of concern since the rise of AI chatbots. Cointelegraph Magazine highlighted in an article last month that while ChatGPT can give you legal advice, your chat logs can be used against you in court.
He pointed to the need for tooling to support the use of LLMs locally on personal devices, utilizing zero-knowledge proofs to make API calls anonymously and improving cryptographic tech to verify work from AI, among other things.
Buterin also envisions AI becoming a users middleman to the blockchain, suggesting that AI agents could verify and audit every transaction, interact with decentralized apps and suggest transactions to users.
AI verification could be a major boon for crypto and other sectors, with increasingly sophisticated scammers on the rise. Address poisoning scams, just one attack vector, have seen a major uptick since December.
“Basically, take the vision that cypherpunk radicals have always dreamed of (don't trust; verify everything), that has been nonviable in reality because humans are never actually going to verify all the code ourselves. Now, we can finally make that vision happen, with LLMs doing the hard part,” he said.
Adding to that, Buterin sees AI bots being able to “interact economically” to handle all onchain activity for users and make crypto much more accessible.
He said bots could be deployed to hire each other, handle API calls and make security deposits.
“Economies not for the sake of economies, but to enable more decentralized authority,” he said.
Finally, Buterin thinks AI can enhance onchain governance and markets if LLMs are used to overcome the limits of human attention and decision-making capacity.
He said that while things like prediction markets and decentralized governance are “all beautiful in theory,” they are ultimately hampered by “limits to human attention and decision-making power.”
“LLMs remove that limitation, and massively scale human judgement. Hence, we can revisit all of those ideas,” he said.