wild how researchers are like, “ya know, the problem is matrix multiplication, lets just not do that part”, and it works great
i imagine #nvidia is watching this with anticipation and anxiety and will probably manage to completely miss the development, because big companies aren’t that good at changing course
@kellogh Nvidia seems at least somewhat different from other large companies in that dimension afaict. The Acquired podcast episodes / interview with Jensen were informative (though probably a bit biased) in that regard.
@spoltier right, the one thing i know about Jensen is he’s got a VERY weird management philosophy and so it’s probably hard to predict where Nvidia is going to fail
@spoltier right, the one thing i know about Jensen is he’s got a VERY weird management philosophy and so it’s probably hard to predict where Nvidia is going to fail