Thomas Wolf, co-founder of Hugging Face, challenges the notion that AI can revolutionize science as currently developed technologies merely serve as advanced students. He argues that without breakthroughs, AI will remain limited to being "yes-men on servers" and unable to create innovative solutions akin to what scientific geniuses like Newton or Einstein did. Wolf believes true advances in AI must involve not just knowledge retention but also the ability to ask unprecedented questions and connect unrelated facts—a capacity that current AI paradigms fail to provide, limiting its potential for significant scientific breakthroughs.
"The main mistake people usually make is thinking [people like] Newton or Einstein were just scaled-up good students, that a genius comes to life when you linearly extrapolate a top-10% student."
"To create an Einstein in a data center, we don't just need a system that knows all the answers, but rather one that can ask questions nobody else has thought of or dared to ask."
"Wolf thinks that AI labs are building what are essentially 'very obedient students' - not scientific revolutionaries in any sense of the phrase."
"Wolf's problem with AI today - and where he thinks the technology is heading - is that it doesn't generate any new knowledge by connecting previously unrelated facts."
Collection
[
|
...
]