YouTube Is Using AI to Alter Content (and not telling us)

[xocetaceans]

This would be a serious breach of trust on the part of YouTube. As Rhett Shull points out, this potentially creates a sense of distrust between content creators and their respective audiences (that comment starts at 09:26 in the video).

However, at this point, it is unclear why this is being done by YouTube.

In my opinion, YouTube should immediately stop this practice. It makes me question whether YouTube will be a reliable source of any information in the future if they are willing to do this sort of thing: that is, this practice might conceivably be a pathway to normalize a future insertion of large volumes of AI slop content, and so at some point, the time cost of having to try to distinguish what is real from what will be AI slop will render using YouTube as no longer worthwhile. (YouTube's ads without an ad blocker are truly bad enough already: many of them belong on the pages of a pseudo-journalistic ad sheet which would be a cross between the National Inquirer and a Chick tract.)