[vonneumannstan]

I think that would be a good precedent given the current lack of rules around AI Safety. These models don't seem to be plateauing yet and could be much more dangerous than Mythos in 1-2 years.