In a recent discussion, Vitalik Buterin, co-founder of Ethereum, posited that artificial intelligence (AI) has the potential to significantly accelerate the network”s development timeline while simultaneously enhancing security standards. This statement follows an experiment where a developer managed to “vibe-code” the entire Ethereum roadmap for 2030 in a matter of weeks.
Buterin remarked that the advancements in AI coding capabilities were unthinkable just six months prior, emphasizing the importance of understanding the direction of these trends. In a practical demonstration, he used AI to replicate his blog software within an hour on his laptop, underscoring the rapid advancements in coding efficiency.
He suggested that developers should allocate half of the increased speed gained from AI towards strengthening security measures. This would involve creating more extensive testing, formal verification, and multiple implementations. “People should be open to the possibility (not certainty! possibility) that the Ethereum roadmap will finish much faster than people expect, at a much higher standard of security than people expect,” Buterin stated.
Additionally, AI has shown promise in enabling formal verification of intricate cryptographic proofs. A collaborator involved in the Lean Ethereum project successfully used AI to create a machine-verifiable proof for one of the complex theorems that STARKs depend on for security. The Lean Ethereum initiative aims to formally verify all components, with AI enhancing the ability to achieve that goal.
Buterin highlighted that merely generating a larger set of test cases is crucial beyond just formal verification. However, he also cautioned that the two-week roadmap experiment carried “massive caveats,” including potential critical bugs and incomplete components. “Do not assume that you”ll be able to put in a single prompt and get a highly-secure version out anytime soon; there WILL be lots of wrestling with bugs and inconsistencies between implementations,” he warned.
He expressed optimism regarding the prospect of attaining bug-free code, which has long been viewed as an idealistic goal. Buterin noted that achieving bug-free code could become a basic expectation, essential for establishing trustlessness within cryptocurrency systems. While total security remains an elusive target, specific claims can be verified in particular cases, effectively mitigating over 99% of negative outcomes associated with faulty code.
Overall, Buterin”s insights illustrate that AI could serve as a dual-purpose tool, not only hastening development timelines for Ethereum but also elevating security protocols to unprecedented levels. This evolution suggests that the trade-off between speed and security may soon be rendered obsolete, allowing for thorough security verification at previously unattainable development speeds.
