Ethereum co-founder Vitalik Buterin identified four promising opportunities at the intersection of cryptocurrency and artificial intelligence technologies that have emerged in recent years.
According to him, the advantages of the synergy of two breakthrough areas are quite obvious. For example, blockchain can compensate for the centralization of AI, provide transparency, storage and security of data.
However, only with the advent of modern large language models, as well as powerful scaling solutions in the crypto industry and technologies like ZKP, FHE or MPC, truly deep integration capabilities emerged, the expert believes.
Buterin looked at the prospects and risks of using AI as an actor, an interface, the rules themselves, and the ultimate goal.
He considers the first option the most viable. As an example, the programmer cited the possibility of reviving the stagnating segment of decentralized predictions. Due to poor monetization, truly competent experts avoid these markets.
Inexpensive to operate, AI-based “participants” can create a strong forecasting sector through mass participation and competition, Buterin is confident. Human participation will only be needed in difficult arbitration situations, he pointed out.
Buterin considers the second option of technology synergy promising, but not without certain risks. The use of AI in blockchain application interfaces involves assisting users.
This is partly implemented, for example, in the MetaMask wallet, which warns of possible fraud. Another application, Rabby, shows the consequences of a planned transaction.
Buterin emphasized that AI can really serve as a mentor in such cases. But there is also a potential risk due to the possibility of attackers using the technology for their own purposes.
The expert called the option of using AI as an arbiter regarding making subjective decisions the most risky due to the problem of adversarial machine learning.
If the AI model that plays a key role in the engine is closed, you cannot check its inner workings and therefore it is no better than a centralized application. If the model is open, an attacker can download and build it locally and develop highly optimized attacks that can then be replicated on a live network,” Buterin explained.
He considers the possibility of using AI as the “objective of the game” to be “very intriguing” but long-term. According to him, here we are talking about creating blockchains, DAOs and other mechanisms that can support various ways of using neural networks.
Let us recall that Buterin proposed reviving the “cypherpunk” revolution in Ethereum, which was conceived for the network at the dawn of its existence.
