Ethereum co-founder Vitalik Buterin praised a new habit taking hold on X: users calling on the platform’s AI, Grok, to weigh in on tweets. In a post on his X account, Buterin called “the easy ability to call Grok on Twitter” one of the most important developments for the site’s “truth-friendliness,” equating its impact only to that of Community Notes. He added that a key advantage is the unpredictability of Grok’s replies, recounting occasions where people expected the bot to confirm extreme political claims, and were instead upended when Grok “rugs them.”
Grok, the chatbot developed by xAI and integrated into X, has quickly become a visible part of the platform’s conversation dynamics. Users often summon Grok in replies to ask it to fact-check claims, provide context, or simply mock a post by requesting the bot’s verdict. That behavior, which can resemble an automated form of dunking, has changed how debates play out on the site: rather than engaging the original poster, many participants appeal to Grok as a third-party arbiter.
Boosting Truth on X
Buterin’s endorsement arrives amid a broader debate about whether AIs attached to social networks improve discourse or amplify its worst instincts. Supporters point to Grok’s ability to quickly provide context and correct obvious falsehoods, helping readers see through misinformation. Skeptics warn that turning an AI into a rhetorical weapon risks lowering the tone of discussions, inviting people to weaponize fact-checking for humiliation rather than understanding.
The controversy over Grok’s behavior intensified after several high-profile incidents in which the bot supplied unexpected, sometimes inflammatory, answers when prompted about sensitive topics. Journalists and researchers have flagged patterns of strange or politically charged responses, which have led to criticism of how Grok is steered and moderated. Those episodes have made some observers cautious about relying on a single, network-linked AI to adjudicate contested claims.
Buterin’s comments also echo his earlier praise for Community Notes, the decentralized fact-checking feature on X that he has described as an encouraging step toward crowd-sourced truth-finding on social platforms. By highlighting both tools, he signaled support for layered, community-informed mechanisms that can counter misinformation, while implicitly acknowledging the imperfect, sometimes surprising, nature of automated responders.
While Grok continues to be woven into everyday exchanges on X, the tension between its utility and its unpredictability is likely to persist. For now, users are experimenting: some treat the bot as a quick sanity check; others weaponize it; and a few, as Buterin notes, find themselves surprised when the machine refuses to play the part they expected.
