Hacker News Reader

Comment by onphonenow | Hacker News Reader

TikTok is pretty diverse. I get zero view / like videos. Facebook got tiring all anyone did was call out / be offended / be outraged. Nothing fun anymore . TikTok has a surprising amount of fun stuff.

View on HN

TikTok has its share of enraging material if you go looking for it. The genius of their presentation is it effectively doesn't exist unless the algorithm decides to show you that content. When presented with that content, if you scroll away fast enough, it won't be presented again.

You can also long-press on a video and say "not interested" if you never want that stuff again.

> if you scroll away fast enough

I love that quote, its like being in the jungle and if you can't run away fast enough you'll be attacked by the flesh eating bugs.

That's exactly what it was like for me. I played with it for a couple of months. I had a tendency to frown at, SMDH at, rubberneck some dumb videos for a half second too long it seems. That magical FYP of great stuff never happened for me. I got like 70% stuff that makes me... huh, I don't know what the emotion is, but it's negative. Irked maybe.

At any rate I ultimately decided I absolutely do not need any part of TikTok in my life.

It's pretty pathetic how the algorithms were supposed to be this passive thing that just worked and if used properly would be showing people what they want to see. But instead it was so obvious how it works that people became very aware of it and change their behaviour to not give any bad input into the algorithm.

I don't know how many years it's been now that I pop into an incognito window to watch a YouTube video of which I don't want similar videos suggested to be for the next 3 months. Tiktok is the same, where if you see something that doesn't interest you, you have to not only not like it, but also don't give it too much attention or that will be taken as a sign that more of that will keep you more engaged.

Their algo is super reactive at times with the right threshold of tuning for expanding/exploring new content. It’s also great at leaning back into familiar stuff.

Overall it’s a really well done ML model.

I actually worked briefly with an engineer who claimed to have seen the inner workings of the TikTok algorithm when she was consulting for them (she was a software engineer and mathematician), and I was very surprised to hear her say that there was no ML involved in the algorithm! Of course, I'm not sure I believe her, but it definitely raises the question of whether an algorithm of TikTok's caliber is even possible without ML.

Hand crafted recommendation algorithms are on par with current ML offerings if not better - but they take much much longer to develop.

Each video is node. Figure out where to add edges to existing content, graph theory some shit and tune for similar content maximizing viewing time of short video clips. No ML necessary. Of course, given the magic ML sometimes seems to be, they could certainly be using it but the inputs and outputs are remarkably simple if you get the data structures right. To be clear, I'm not saying it's easy, just that ML's not necessary.

It's all if-then statements, like "If this video has a bird in it, then show it to people who like birds." No ML needed /s

You jest, but at this point YouTube’s recommendation system sees even less complicated than this. Just random videos from years ago will one day become blessed by the algorithm and receive millions of views.

YTs algorithm now just seems to amplify very, very small fluctuations in view counts. 3 people just happened to watch the same video at the same time? It must be trending! Shove it in front of everyone. It’s like Google has used all of machine learning to recreate taking a derivative.

Seeing it from afar, basically just Libs of Tik Tok, it doesn't seem that fun. (Although I guess the point of that is to highlight extreme stuff.)

Reply on news.ycombinator.com