Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

ByteDance's Recommendation System

github.com

Post reply on HN
Print view View original

41–50 of 58 posts

Re: ByteDance's Recommendation System

#41

Post by xnx »

xnx wrote: post #7

Like Facebook before, everyone talks about the TikTok algorithm being some super secret and valuable mystery. In reality, both Facebook and TikTok succeeded because they were at the right place at the right time and didn't screw things up. The TikTok recommendation system is smart and very well implemented, but nothing novel that couldn't be implemented by a dozen other teams.

Additional evidence that TikTok doesn't possess special algorithm and infrastructure prowess is the complete failure(/gaming?) of search. Many (most?!) searches on TikTok have been returning irrelevant shock and/or porn content for weeks.

Re: ByteDance's Recommendation System

#42

Post by wbl »

xnx wrote: post #7

Like Facebook before, everyone talks about the TikTok algorithm being some super secret and valuable mystery. In reality, both Facebook and TikTok succeeded because they were at the right place at the right time and didn't screw things up. The TikTok recommendation system is smart and very well implemented, but nothing novel that couldn't be implemented by a dozen other teams.

What did Livejournal, Xanalga, Myspace, Tumblr, all screw up and Facebook do right? Facebook had stiff competition from the start.

Re: ByteDance's Recommendation System

#43

Post by mrfox321 »

xnx wrote: post #34
mrfox321 wrote: post #32

Earlier quoted context omitted.

It's simpler intuition but more complex from a data / ml perspective. Their algorithm is really built around their features. Specifically, temporal representations of user interest: https://ieeexplore.ieee.org/document/9458799/ The features used by their algorithm tells you what a user is interested, historically. Contrast this to Meta, which uses the social graph as their features. Imagine features like the number o…

That's a crazy design choice by meta if true. The interests of those in my social graph have very little connection to my interests.

It's because they originally built their recommendation system to recommend friends and their content. Here, the social graph makes complete sense as the foundation for their simple search algorithm. But as they expanded their recommendation capabilities, the features stuck around. It's the same reason why tech debt accumulates. Data sticks around in the same way code does. But data is even higher friction, since it's a superset of the code.

Re: ByteDance's Recommendation System

#44

Post by smokel »

xnx wrote: post #7

Like Facebook before, everyone talks about the TikTok algorithm being some super secret and valuable mystery. In reality, both Facebook and TikTok succeeded because they were at the right place at the right time and didn't screw things up. The TikTok recommendation system is smart and very well implemented, but nothing novel that couldn't be implemented by a dozen other teams.

Everyone talks about TikTok having some valuable algorithm, but I keep running into the most dimwitted nonsense every time I try it.

Then again, perhaps this is the algorithm's way of scaring me away to save precious bandwidth, knowing full well that I will never buy products from online ads anyway?

Re: ByteDance's Recommendation System

#45

Post by trogdor »

Majromax wrote: post #14
davidjfelix wrote: post #11

Earlier quoted context omitted.

My take: it's a mix of brand bundling and lack of data. They're roughly equivalent but shorts is bundled with youtube which has its own brand perception and reels are bundled with IG/FB and have their own brand perception. Additionally fewer users means less algorithmic data to keep viewers. Tiktok was allowed to establish its own brand and develop a community while shorts and reels are intrinsically tied to their pa…

> My take: it's a mix of brand bundling and lack of data. They're roughly equivalent but shorts is bundled with youtube which has its own brand perception and reels are bundled with IG/FB and have their own brand perception. Additionally fewer users means less algorithmic data to keep viewers. My intuition would work the other way around. I'd expect offerings from more established companies to have a big leg up in te…

> If boring-old-Youtube could promise shorts creators great exposure to an enthusiastic audience, it would win the platform regardless of its brand.

Just a guess, as someone who makes their living from YouTube: YouTube creators are driven to create content that earns them money. As compared to long-form content, YouTube shorts earn next-to-nothing, and it’s not clear that they drive significant new traffic to more-valuable content.

Most large creators on YouTube are focused on the bottom line, not exposure.

Re: ByteDance's Recommendation System

#46

Post by fragmede »

trogdor wrote: post #45
Majromax wrote: post #14

Earlier quoted context omitted.

> My take: it's a mix of brand bundling and lack of data. They're roughly equivalent but shorts is bundled with youtube which has its own brand perception and reels are bundled with IG/FB and have their own brand perception. Additionally fewer users means less algorithmic data to keep viewers. My intuition would work the other way around. I'd expect offerings from more established companies to have a big leg up in te…

> If boring-old-Youtube could promise shorts creators great exposure to an enthusiastic audience, it would win the platform regardless of its brand. Just a guess, as someone who makes their living from YouTube: YouTube creators are driven to create content that earns them money. As compared to long-form content, YouTube shorts earn next-to-nothing, and it’s not clear that they drive significant new traffic to more-va…

The reason shorts don't earn any money, as compared to Instagram and TikTok, is that they don't advertise crap for me to buy (I have YT premium), so I don't end up buying shit there like I do the other two.

Re: ByteDance's Recommendation System

#47

Post by xnx »

wbl wrote: post #42
xnx wrote: post #7

Like Facebook before, everyone talks about the TikTok algorithm being some super secret and valuable mystery. In reality, both Facebook and TikTok succeeded because they were at the right place at the right time and didn't screw things up. The TikTok recommendation system is smart and very well implemented, but nothing novel that couldn't be implemented by a dozen other teams.

What did Livejournal, Xanalga, Myspace, Tumblr, all screw up and Facebook do right? Facebook had stiff competition from the start.

> What did Livejournal, Xanalga, Myspace, Tumblr, all screw up

Don't forget Friendster!

Mainly bad-luck in being too early or too late, though MySpace had some self-inflicted problems with performance and overcomplicating the interface (by allowing customization).

Re: ByteDance's Recommendation System

#48

Post by xnx »

smokel wrote: post #44
xnx wrote: post #7

Like Facebook before, everyone talks about the TikTok algorithm being some super secret and valuable mystery. In reality, both Facebook and TikTok succeeded because they were at the right place at the right time and didn't screw things up. The TikTok recommendation system is smart and very well implemented, but nothing novel that couldn't be implemented by a dozen other teams.

Everyone talks about TikTok having some valuable algorithm, but I keep running into the most dimwitted nonsense every time I try it. Then again, perhaps this is the algorithm's way of scaring me away to save precious bandwidth, knowing full well that I will never buy products from online ads anyway?

Quickly skip past videos you're not interested in, and like or comment on those you like. Your "For You" page should noticeably change.

Re: ByteDance's Recommendation System

#49

Post by wbl »

xnx wrote: post #47
wbl wrote: post #42

Earlier quoted context omitted.

What did Livejournal, Xanalga, Myspace, Tumblr, all screw up and Facebook do right? Facebook had stiff competition from the start.

> What did Livejournal, Xanalga, Myspace, Tumblr, all screw up Don't forget Friendster! Mainly bad-luck in being too early or too late, though MySpace had some self-inflicted problems with performance and overcomplicating the interface (by allowing customization).

This just feels like special pleasing. Why did Facebook get the timing exactly right and what was different about that time vs. a better understanding of user wants, well planned rollout and promotion strategy that created desire to have it, and not being too greedy.

Re: ByteDance's Recommendation System

#50

Post by PaulHoule »

mrfox321 wrote: post #31
PaulHoule wrote: post #23

Earlier quoted context omitted.

It is the users and not the company. I haven't worked on sites as big as YouTube but on sites with 100,000 members who are very much engaged with one "game" you usually find they are mainly indifferent when you offer them another "game" to play. I like YouTube for what it is. I have interacted very little with shorts but Google has scarily seen into my imagination. I don't want to go into that rabbit hole.

IG users and Tiktok users were / are quite similar. Especially when Tiktok wasn't yet eating IGs lunch.

The definition of "similar" is the problem w/ vector search isn't it?

Two populations can be similar in terms of conventional demographics such as age, gender, race, what kind of clothes they wear, etc. but be different in their behavior. IG users are "players of the Instagram game" and TikTok are "players of the TikTok game" and a whole system of values and behaviors are involved.

To take an example playing the "engagement farming" game on Bluesky I can follow people and know some fraction of people will follow me back, but who do I want to follow?

I postulated that the people I want are people who will repost my photos so I can try following people who repost photos but I find that reposters are not "followers" whereas I get a much better response rate if I follow people who follow another social media photographer since those people are "followers". People have an online behavior signature like that which for me matters more than the color of your skin.

Post reply on HN