[nonbirithm]

> YouTube's downvotes, invisible though they now are, can at least influence the recommender algorithm.

YouTube has this feature, but the information it signals is now hidden from the public. The hiding of dislikes have essentially nullified the disapproving glares of the surrounding audience. I'm wondering what the ultimate effects on viewership will be, and the effects on the algorithm might not be in the favor of viewers.

A theory I have is that the lack of dislikes could make potential viewers more anxious that they could be stepping into misinformation at any time and have no easy (even if unreliable) way of checking veracity, so they choose to engage with less content overall.

I have also heard that if someone dislikes a video after watching it all the way through, it still counts them as "engaged" and doesn't negatively impact the channel's exposure. That would mean that a viewer could be mislead into thinking that a dislike means "nobody should watch this", whereas the effect is actually the opposite for the platform's algorithm.

I would really like to see hard data on this from a YouTube channel with significant viewership, detailing how the view count and engagement have changed since dislikes were hidden. If YouTube claims that dislikes don't influence viewership, there should at least be evidence that it is so.