Posted by peter_d_sherman 3 days ago
Sounds incredible, as in: can't believe it.
By the way, isn't Nvidia researching this?
Very real stuff (I have a PhD on this)
> Nonlinear phenomena become relevant only when the input light is very intense. Typically, in order to observe nonlinear phenomena, an intensity of the electromagnetic field of light larger than 108 V/m (and thus comparable to the atomic electric field of ~1011 V/m) is required.
I don’t know anything about this field. Can you address this point? How close are we to overcoming this in academia?
So also 1000x smaller integration
I'm not excluding that now we have more budget/time/AI to make it work (if it can), I'd just say exercise caution reading this (very AI slop) article (and anything that looks too good to be true)