OpenAI and Anthropic models challenge exponential time hypothesis, scientists rethink
AI models from OpenAI and Anthropic have managed to solve problems once thought impossible to crack quickly, challenging a 25-year-old theory called the exponential time hypothesis (ETH).
This is making scientists rethink how much we should rely on human expertise versus AI when it comes to big discoveries.
MIT's Ryan Williams vindicated, peers concerned
MIT's Ryan Williams, who doubted ETH for years, feels validated, saying, "I did not fully believe the hypothesis, and I'm on the record as a prominent skeptic. So I felt vindicated."
But not everyone's thrilled. Impagliazzo wishes humans had made the breakthrough themselves, while Paturi encourages students to use AI as a creative partner.
There are also worries about private companies controlling powerful AI tools and the importance of understanding how these AIs make decisions.