Ethereum co-founder Vitalik Buterin says artificial intelligence (AI) could be crucial to solving one of Ethereum’s “biggest technical” risks: bugs buried deep within its code.
In a Feb. 18 post on X, Buterin shared his excitement for AI-powered auditing to identify and fix buggy code in the Ethereum network, describing it as the “biggest technical risk” to the network.
Buterin’s comments come as Ethereum closes in on implementing its long-awaited Dencun upgrade, which is currently slated for launch on March 13. Dencun was implemented on the Goerli testnet on Jan. 17, but a bug in Prsym prevented the network from finalizing on the testnet for four hours. Upgrades to the Ethereum network are crucial to the long-term roadmap of the blockchain.
However, not everyone agrees that AI is a reliable tool to detect bugs in Ethereum-based code.
In July 2023, OpenZeppelin conducted a series of experiments that utilized OpenAI’s GPT-4 to identify security issues in Solidity smart contracts — the native language of Ethereum code.
During these experiments, GPT-4 successfully identified vulnerabilities in 20 out of 28 challenges.
When GPT-4 failed to identify flaws, it could often be prompted to correct its mistakes quickly. However, at other times, OpenZeppelin found that the AI had actually invented a vulnerability that had never existed in the first place.
Similarly, Kang Li, the chief security officer at CertiK, told Cointelegraph that using AI-powered tools — such as ChatGPT — in coding often creates more security issues than it solves.
Overall, Li recommends that AI assistants should be used only as assistants to experienced coders, as they can be helpful in quickly explaining to developers what a line of code means.
“I think ChatGPT is a great helpful tool for people doing code analysis and reverse engineering. It’s definitely a good assistant, and it’ll improve our efficiency tremendously.”
While Buterin is largely optimistic about the future of AI, he has previously warned developers to exercise caution when implementing AI with blockchain technology, particularly when deploying it alongside “high-risk” applications, such as oracles.
“It is important to be careful: if someone builds e.g. a prediction market or a stablecoin that uses an AI oracle, and it turns out that the oracle is attackable, that’s a huge amount of money that could disappear in an instant.”
Source : Cointelegraph by Tom Mitchelhill / Feb 19, 2024