Totally agree. The whole text-to-<stuff> trend really needs to be put into perspective. Fortunately, trends move through quickly these days, so I don't think we'll be struggling under this tedium for too much longer. It has its place, for sure, but ultimately it's only one piece in the puzzle. I think a lot of the hype is simply due to the fact that it's been a long-standing challenge that's finally been met with reasonable success.

I happen to be in music, where text just isn't particularly appropriate. Unfortunately, all the new music AI systems are grounded in text, at this point, so while there should be a ton of exciting stuff happening, there really isn't much to celebrate from an end-user perspective. The problem, of course, is that there are a bunch of challenges involved in effectively leveraging AI for different kinds of interactions, and that development isn't really happening right now, because everybody appears to be focused on text-to-music.