Vitalik: AI is extremely knowledgeable but not wise, and can only make superficial judgments

Ethereum founder Vitalik Buterin responded to Ruxandra Teslo on X, stating that he believes AI is extremely knowledgeable and can excellently identify quotes in almost any content, but it is not wise, and therefore can only make the most superficial judgments about the viewpoints a piece of writing attempts to express.

Vitalik Buterin noted that this situation is particularly evident when AI analyzes text it reads for the first time, as opposed to analyzing text that already exists in the training data, which comes with a lot of analyses from others. He also mentioned that the knowledge of AI is more easily seen than its limited wisdom, as wisdom may be something that is difficult to verify easily, which is different from intelligence.

This reply was directed at Ruxandra Teslo's post. Ruxandra Teslo expressed that the related post made her feel pessimistic about humanity and believed that the discussions of large language models regarding her article are still low-quality content; she was responding to Sichu Lu's statement that AI is better suited than humans to discuss classics at that time.