We are all in awe of how LLM's and their human assistants seem to have solved the Riemann hypothesis among many other Theorems.
The amazing achievements of these LLM's and people is undeniable but one thing is often left out in the credits, even though it seems to me it's the most important step to get here: Lean.
Lean is understandable yet so expressive that it allows proving the Riemann hypotheses through a plethora of mathematical abstractions. More importantly it has allowed LLM's to "brute force" proofs, without the need to understand those abstractions simply by understanding Lean.
I'd say the creation of Lean has been the most important step to make these proofs possible and I am a bit surprised that it's creator is rarely credited or well known. To be honest I am not sure how large his role was in Lean's development but it seems he is the primary author.
But maybe people more closely related to Lean development can chip in?
tomtomtom777•10m ago
The amazing achievements of these LLM's and people is undeniable but one thing is often left out in the credits, even though it seems to me it's the most important step to get here: Lean.
Lean is understandable yet so expressive that it allows proving the Riemann hypotheses through a plethora of mathematical abstractions. More importantly it has allowed LLM's to "brute force" proofs, without the need to understand those abstractions simply by understanding Lean.
I'd say the creation of Lean has been the most important step to make these proofs possible and I am a bit surprised that it's creator is rarely credited or well known. To be honest I am not sure how large his role was in Lean's development but it seems he is the primary author.
But maybe people more closely related to Lean development can chip in?