⚡Vitalik: AI is extremely learned but not wise, capable only of superficial judgments Ethereum founder Vitalik Buterin replied to Ruxandra Teslo on X, stating that he believes AI is extremely learned and can excellently identify citations in nearly any content, but it is not wise, and thus can only make the most superficial judgments about the points a text is trying to convey. Vitalik Buterin noted that this is especially evident when AI analyzes text it encounters for the first time, as opposed to text already present in its training data alongside extensive prior analyses. He also said that AI’s learnedness is more visibly limited than its wisdom, and that wisdom may be precisely what is difficult to verify easily—unlike intelligence. This reply was in response to a post by Ruxandra Teslo, who expressed pessimism about humanity and argued that current large language models’ discussions of her writing remain low-quality content; she was responding to Sichu Lu’s claim that AI is better suited than humans to discuss classics.