moot✳
A little more connected.
Join Moot

openai's ai might have just disproved one of the most important assumptions in theoretical computer science...

something hundreds of research papers have relied on for the past 30 years.

it's called the exponential time hypothesis (ETH).

basically, computer scientists have long believed that certain problems, like finding a solution to a boolean satisfiability problem, fundamentally require exponentially increasing computation as they get bigger.

the model apparently found a way to solve these problems in subexponential time, and the result has been formally verified in lean.

this doesn't mean p = np, or that these problems suddenly become easy in practice. but it means a foundational assumption about how hard computation has to be may have been wrong all along.

the downstream implications could be enormous... graph algorithms, computational biology, databases, optimization, and hundreds of conditional hardness results may need to be revisited.

and openai has brought in two of the original researchers who proposed ETH to help understand the proof of its own refutation...

Greg Brockman@gdb

Announcing three major scientific discoveries, one each in mathematics (harmonic analysis), computer science (complexity theory), and physics (quantum field theory).

AI starting to help us improve our understanding of the universe! https://openai.com/index/building-scientific-understanding-with-academic-communities/

View on X
#theoretical computer science#ai#exponential time hypothesis
2
0 comments
0 comments