Kinda silly to follow your “prove it” challenge with an absurd claim you most certainly cannot prove, much less support with evidence.

It was not a "prove it" challenge.

I'm pointing out that there's no known information theoretic constraint about the impossibility of frontier AI models being improved to fit/run on a small GPU.

Please do not make up plausible sounding science facts.

I will not claim a 5070, but there is already evidence in nature that you can get very good general intelligence with an order of magnitude less wattage.

There are constraints of course- training takes way longer.