Fair point. But don't you think that companies today have learned from the mistakes of the past? Also they have way more cash for aggressively buying out disruptors.
358 karma · joined April 29, 2020
On the other hand, they could provide a different function (GPUSort for instance) but they shouldn't replace the default one with different threading requirements.
It seems to me that some winners emerge, which makes other competitors want to replicate the winner's formula. Over time all products end up the same. Trying to deviate from that sameness will either result in customers not being interested or everyone copying you. This may be bad for innovation.
I am trying to build my own Language Model using Pytorch and it is much easier than a lot of the stuff I've had to do as a traditional SWE over the years. My understanding is that the difficulty comes from training massive models that cannot fit on a single GPU efficiently, so more of a software problem than a maths problem.
Am I suffering from the Dunning-Kruger effect ? Or is the genius of the field in the combination of this limited set of techniques ?
Funny because Google is famous for having a monorepo at scale.
...although I have to say that it is a little bit embarrassing that my "reputation" up to now is mainly from this thread. –
mikuszefski
Nov 25, 2014 at 18:19