Can't mind read the guy obviously, but the usual motivation that I'm aware of is that you pretty much fuck over everyone else that comes later. Upscalers improve over time, but in terms of distribution, recency bias is strong and visual treats are inviting. So when those much better upscalers eventually come around, what's more likely to still be available is the secondary source you distributed, which is already upscaled once with a then-inferior upscaler. This leads to a form of generational rot.
Other likely explanations are:
- them not liking how these upscalers look: you can imagine if they can nitpick minor differences between different encodes that most people don't notice, they'll hate the glaring artifacts these filters usually produce
The reasons stated against upscaling were that (re-)encoding video files should generally be done in a way that preserves as much of the original information and intent as possible. AI upscalers add information where there is none, thus modifying the video in a way that goes against that goal.
Other likely explanations are:
- them not liking how these upscalers look: you can imagine if they can nitpick minor differences between different encodes that most people don't notice, they'll hate the glaring artifacts these filters usually produce
- boycotting AI