Yeah, it sure was convenient that there was a RAM pricing crisis right when Apple was making local inference viable. All because of a promise that AI companies will buy more of it... with money they don't yet have, whereas Apple does have lots of money.