[infurl]

https://scitechdaily.com/ai-magic-just-removed-one-of-the-biggest-roadblocks-in-astrophysics/

Using a bit of machine learning magic, astrophysicists can now simulate vast, complex universes in a thousandth of the time it takes with conventional methods. The new approach will help usher in a new era in high-resolution cosmological simulations...

I think of this as being like fuzzy function memoization. When you memoize a function you put a wrapper around it to intercept calls to the function and compare the parameters to a list of previously seen function calls. If the function has already been called with those same parameters, you return the result that it got with those parameters last time. If it hasn't already been called with those parameters, you run the function and add the new result to the list of past results. Obviously it is only useful with immutable functions, those not having side effects. It is the same technique that makes chart parsers and fast fourier transforms so efficient. You are storing all your intermediate results for reuse as the program runs.

[8pla.net]

Not surprising, since AI, can generate an extremely convincing video from a single frame of the Mona Lisa painting.