> I find it funny that casual observers on the outside are so morally opposed. ... It's really going to help mid market and below, for films that don't have Disney budgets.

Agreed. I also have a few decades of experience in film and television production, mostly in creating and deploying new digital tooling paradigms from 'desktop video' in the 90s to virtual sets to real-time 3D environments. New digital production tools have almost always had the biggest impact enabling low-end and mid-tier creatives, not big budget studio productions. In the early 90s the Amiga-based Video Toaster enabled upstart productions like Mystery Science Theater 3000 and Babylon 5. The Toaster also enabled about 95% more cable-access crap and bad porn but the other 5% was fantastically creative new stuff which couldn't have existed on indy budgets. Dramatic new production paradigms tend to unleash both democratization and disruption. Most people welcome the democratization yet reflexively fear the disruption. Today, few recall the early 90s predictions from the professional production industry of desktop video causing economic and creative doom, despite being widespread and echoed across mainstream media.

While machine learning-based production tools aren't flexible or granular enough yet for more than limited experiments, there's no reason they won't become increasingly useful for real work. IMHO they'll likely have the same kind of democratizing impact as desktop video and the Toaster - 95% more regrettable crap, some of which we're already starting to see but, eventually, also 5% more wonderfully creative stuff which wouldn't have existed without it. The crap will quickly fade away but the bold new stuff will remain pushing creative frontiers and shaping tomorrow's classics.