> I don't want to see Altman's fearmongering and hyperbolic statements His statement wasn't even hyperbolic or fearmongering.....? He just extrapolated based

> I don't want to see Altman's fearmongering and hyperbolic statements His statement wasn't even hyperbolic or fearmongering.....? He just extrapolated based on current amounts of compute and estimated a possible model size that could be equivalent to AGI (based on current architecture). Training a model of that size would require too much electricity. That was his point.