At a time when Anthropic is vacuuming up stock market money, the AI company has made a pretty grandiose claim about its Claude chatbot’s scientific capabilities.
The San Francisco-based artificial intelligence firm published a lengthy blog post this week claiming that its Claude Fable 5.2 model had made some progress on figuring out two long-standing mysteries surrounding black holes. The post goes into an incredible amount of scientific detail that might fly over your head (as much of it did mine), but I’ll try my best to summarize it in plain English.
According to Anthropic, the first conundrum Claude was able to solve involved the concept of heating up a hypothetical universe. When the universe reaches a certain temperature, it would form a black hole through a sudden “phase transition,” as initially described by Stephen Hawking and Don Page in 1983. Hawking and Page’s findings concerned only a vast universe described by general relativity, while this theoretical universe is much smaller and might work differently.
Mashable Light Speed
The best tech deal this week
Products available for purchase through affiliate links. If you buy something through links on our site, Mashable may earn an affiliate commission.
It turns out, as Claude found, that the phase transition is also abrupt in a smaller universe. As Anthropic explained, this implies that “black hole formation isn’t just a feature of relativity but is actually built into the underlying quantum theory.”
As for the second finding, it revolves around another theoretical universe made up of two dimensions instead of three. In this universe, physicists have spent years trying to figure out how heavy the lightest possible black hole could be using a technique called “bootstrapping,” in which the ceiling for the lightest black hole gets lowered each time a theory breaks it. Anthropic claimed that Claude Fable 5.2 was able to notice a pattern that physicists had missed in their calculations and use it to “calculate the precise ceiling for the simplest version of this technique and prove that it could not be lowered.”
As Anthropic was quick to point out in the blog post, these findings do not apply to our universe or have any practical application but could give us an idea of how relativity and quantum mechanics work together. It’s also worth noting that these findings were released alongside Anthropic’s much-anticipated IPO, AKA a time when Anthropic wants as much public money as possible.
Topics
Artificial Intelligence
Anthropic

