Ethereum co-founder Vitalik Buterin recently shared his perspective on the limitations of artificial intelligence, noting that while models possess vast knowledge, they lack true wisdom. He observed that chatbots excel at identifying references but struggle to grasp the deeper arguments within a text, often providing only surface-level analysis.
The discussion was prompted by a report on the Humanities 10 course at Harvard, which saw a 40% increase in applications this fall, with 180 students competing for 90 available seats. The course prohibits the use of digital devices, as students seek alternatives to AI-driven shortcuts in their academic work.
Buterin highlighted that AI models face particular difficulties when analyzing new texts that lack extensive human commentary in their training data. He previously warned that AI could potentially compromise crypto security within two years and has advocated for maintaining human oversight in technical processes.
Despite these concerns, many crypto teams continue to utilize AI for research, governance summaries, and smart contract audits. Buterin suggests that relying on AI for such critical tasks carries risks, as models may overlook subtle flaws that require human judgment to identify.
