Ye (formerly known as Kanye West) is buying Parler, a social network that some describe as an alt-tech alternative to Twitter (focusing on the conservative side of the political map), and others as a massive echo chamber of right-wing loyalists.
A while back, YouTube removed the ability to see the number of dislikes a video has received; I wonder if they’ve since seen a decrease in engagement since viewers can no longer see how controversial a video is?
Also, I am curious to hear your thoughts on the feasibility of calculating a “consensus score” with an asymmetrical reaction system. In other words, instead of giving users the ability to like/abstain/dislike, you only give them the ability to like/abstain (such as with Instagram), or maybe have a superlike/like/abstain (such as Twitch’s subscribe/follow/abstain system for livestreams). Maybe the tighter spread of values in these asymmetrical reaction systems is not wide enough to estimate consensus?
Love that xkcd comic! I think one can do the same for asymmetric systestem. We may need then to use more NLP to also understand the actual sentiment. But can be interesting.
Perfect summary paragraph which reinforced this question I had throughout: I wonder if arenas prone to be echo chambers algorithmically introduce even more extreme content to reinvigorate engagement amongst the population they do have versus providing a more neutral debate forum to drive that engagement amongst a more heterogenous population.
Great read! Reminds me of this classic xkcd comic: https://xkcd.com/386/ :)
A while back, YouTube removed the ability to see the number of dislikes a video has received; I wonder if they’ve since seen a decrease in engagement since viewers can no longer see how controversial a video is?
Also, I am curious to hear your thoughts on the feasibility of calculating a “consensus score” with an asymmetrical reaction system. In other words, instead of giving users the ability to like/abstain/dislike, you only give them the ability to like/abstain (such as with Instagram), or maybe have a superlike/like/abstain (such as Twitch’s subscribe/follow/abstain system for livestreams). Maybe the tighter spread of values in these asymmetrical reaction systems is not wide enough to estimate consensus?
Love that xkcd comic! I think one can do the same for asymmetric systestem. We may need then to use more NLP to also understand the actual sentiment. But can be interesting.
I would love to see this on reddit!
Feel free to post it!
My subreddits are more on coffee than academics :)
Perfect summary paragraph which reinforced this question I had throughout: I wonder if arenas prone to be echo chambers algorithmically introduce even more extreme content to reinvigorate engagement amongst the population they do have versus providing a more neutral debate forum to drive that engagement amongst a more heterogenous population.