Vitalik Buterin AI Challenge Solved in 2 Hours: Can Developers Stay Anonymous?
Vitalik Buterin has confirmed a winner in his AI problem to unmask his nameless writing, after researcher Franklyn Wang traced a hidden EIP-7503 revision again to the Ethereum co-founder.
Buterin launched the experiment on June 22, providing up a part of his personal anonymity to check whether or not AI stylometry can establish hidden authors. For 13 days, nobody succeeded.
AI Challenge Winner Matched Reasoning, Not Prose
Wang ran the search by way of Co-Invest, an AI analysis engine, and flagged a December 2024 revision of EIP-7503, the Zero-Knowledge Wormholes privateness proposal. A throwaway account submitted the rewrite, which now types roughly 75% of the proposal’s textual content.
The disguise practically held. Keyvan Kambakhsh, an unique EIP-7503 writer, reviewed and authorised the nameless edit on the time. According to Wang’s evaluation, his mannequin gave the decide simply 20% confidence, but 10 occasions its subsequent candidate. The search reportedly took about two hours.
Follow us on X to get the newest information because it occurs
Buterin drafted the revision in Chinese, translated it regionally with Alibaba’s Qwen2.5 mannequin, and manually mounted the output. However, that camouflage lined solely his prose.
“Notice that the stylistic hints that his AI picked up on had been mental habits and magnificence of math and algorithm clarification, which bypassed my obfuscation technique (which solely lined prose) utterly.”
Buterin confirmed the end result on Monday. Wang, for his half, argued the identical engine may hunt buying and selling alerts throughout information and on-chain knowledge.
What the Result Means for Pseudonymous Developers
Stylometry has unmasked authors earlier than. In 2013, forensic linguists recognized J.Okay. Rowling as crime novelist Robert Galbraith by way of vocabulary and phrasing. Buterin’s check suggests detection now reaches deeper, into how an writer causes relatively than how they write.
That shift issues for an business constructed on pseudonyms, from Satoshi Nakamoto down. Ethereum alone just lately handed 1 million developers, whereas European regulators already gasoline crypto privacy fears.
Buterin has championed privateness for years, from co-authoring the Privacy Pools paper to his Lean Ethereum roadmap. His self-experiment additionally sharpens the talk over AI safety rules as fashions develop extra succesful.
Whether obfuscation can catch up, or whether or not this “thoughtprint” detection retains bettering, could grow to be clearer as researchers rerun Wang’s methodology on different nameless work.
The submit (*2*) appeared first on BeInCrypto.
