When Twitter forced the algorithmic timeline on people, someone tweeted:
"Is the algorithm based what everyone else likes, or what I like? Because I hate what everyone else likes, and I don't know what I like."
"Is the algorithm based what everyone else likes, or what I like? Because I hate what everyone else likes, and I don't know what I like."
There have been many articles about the phenomenon of the "For You Page"[1][2]. But the funny part is Meta sees this and figures - people are looking at videos, we'll do videos. And not - people are seeing what they specifically like, lets do that
[1] https://www.nytimes.com/2021/12/05/business/media/tiktok-alg...
Learning the things you don't like is important. It's much more pleasant experience when you are not exposed to the stuff you don't like. Unlike Twitter, the stuff you don't like doesn't mean things you disagree with. You very well might not like politics, TikTok keeps you out of the culture war then.
Similarly (and I have no way of proving this, but I believe it 100%), TikTok seems to alter your feed to incorporate videos from other people watching TikTok near you. My friends and I have made a game of this where we cast TikTok to a TV, watch videos, and try to guess which of us the video was intended for.
I've never once had a "oh wow that's a neat content discovery method" moment from Facebook/Instagram/YouTube but it has happened multiple times with TikTok. I'm not saying TikTok is the end-all-be-all of social media, but it's a good window into what social media could be, if they stopped chasing ad revenue and anger-driven content.
And I agree with the other commenter that TikTok is good at finding novelty in ways the other is not. Things like YouTube or whatever will just show you endless variations of the things you've already liked. TikTok will throw you things you like from way out of left field.
They optimized their primary user interaction to provide specific input!
By distilling user choice down to {watch time} or {swipe}, and preferentially promoting short pieces of content that generate more swipes, they get unambiguous and frequent input.
Which isn't to lessen their achievement, but is to highlight that they engineered their entire experience to provide their algorithms more and cleaner feedback.
What they're bad at is working out what you don't know what you want: the example that she gave was: you land in a new city, you're on your way somewhere. A recommendation engine might already know that you drink coffee and point out the cafes on the way to where you're going - but will it "know" to suggest that there's a public sculpture and that you wouldn't think to look at ordinarily, but you would be happy to see.
TikTok seems much closer to that notion of "systems of delight".
(However, for a great deal of social media I want to be running what I look at anyway.)
Then - maybe one video in 20 or something - you get stuff that's either completely different, completely new (under 100 likes/views) or related to something you liked a long time ago. And then the next one is again something you actually like.
The Instagram model is just shoving "recommendations" to every 3rd slot on my screen. And every 4th one is an ad.