Ethereum founder Vitalik Buterin said that artificial intelligence can accelerate the development of the network while improving security standards.
Summary
- Vitalik says AI can speed up Ethereum’s roadmap and delivery times.
- Part of the benefits of AI should go to rigorous testing and formal validation.
- AI can help make near-bug-free crypto code an anticipated reality.
Responding to an experiment in which someone “vibe-coded” Ethereum’s entire 2030 roadmap within weeks, Buterin wrote that “six months ago, even this was out of the realm of possibility, and what matters is where this trend is going.”
Buterin experimented with AI coding firsthand by building software for his hour-long blog using his laptop.
The Ethereum founder proposed taking half of the speed gains from AI and applying them to security with multiple test cases, formal verification, and multiple implementations.
“People should be open to the possibility (not a certainty! a possibility) that the Ethereum road will end much faster than people expect, with a much higher level of security than people expect.”
AI enables the formal verification of complex cryptographic evidence
A contributor to the Lean Ethereum project managed to code AI a machine-verifiable proof of one of the most complex theorems STARK relies on for security.
The Lean Ethereum initiative aims to validate all components, with AI improving the ability to achieve that goal.
Buterin noted that simply generating a large body of test cases is more important than formal validation.
The two-week roadmap test contains “big caveats: probably a lot of serious bugs, and maybe in some cases ‘stub’ versions of something where the AI didn’t even try to make a complete version.”
The right approach divides the benefits of AI between improvements in speed and security. “Don’t think you’ll be able to plug in one piece of information and get a more secure version anytime soon; there WILL be a lot of fighting with bugs and inconsistencies between implementations,” warns Buterin.
Bug-free code can go from logical manipulation to basic expectations
Buterin expressed excitement at the possibility that bug-free code, “which has long been considered a hoax, will finally become accessible first and then become the basic expectation.” He listed this as the need to distrust crypto systems.
Absolute security remains impossible as it would require direct communication between lines of code and mental content, Buterin estimated at many terabytes of information.
Certain security claims can be made and verified in certain cases, eliminating more than 99% of negative results from broken code.
The statement shows AI as a tool for both accumulating development opportunities and increasing security barriers at the same time.
Buterin’s framework suggests that AI can eliminate that trade-off by enabling comprehensive security assurance at a speed of development not previously possible.
