I thought this was an interesting take. On the one hand it's peak AI hype hysteria... and part of the thesis feels like it could argue for the opposite: democratization of software development means more software... But I guess the article is really about the viability of medium scale software companies in a hyperscaler world...