> On Friday, the regulators released a preliminary decision that TikTok’s infinite scroll, auto-play features and recommendation algorithm amount to an “addictive design” that violated European Union laws for online safety.
I don't understand the legal side, but after gaining and kicking a Tiktok addiction during and after COVID, I believe it. I was there 4-8 hours a day and tried to scroll videos while washing dishes (and during nearly any other activity).
Maybe I don't get addicted easily, but after 30 minutes of forcing myself to watch tiktok, I just uninstalled it. Friends told me I didn't give it enough time to learn my tastes but... How could it, given that literally 100% of the videos in my interest areas were trash?
Good. I feel like since cracking down on smoking in the 90s we've become really complacent to the dangers of addiction. Just like with smoking you'll get people inside the industry defending it too (like in this very comment section).
Might be a generational thing, but I never understood the "shorts" (in any format on any social network).
I can watch a 9 hour video on GTA games without problems (not in one sitting, but in parts), but 3 'shorts' in a row with not enough info and explanation to be interesting makes me close any of the 'shorts' apps (tiktok, youtube shorts, instagram....).
I gave a talk at PyData Berlin on how to build your own TikTok recommendation algorithm. The TikTok personalized recommendation engine is the world's most valuable AI. It's TikTok's differentiation. It updates recommendations within 1 second of you clicking - at human perceivable latency. If your AI recommender has poor feature freshness, it will be perceived as slow, not intelligent - no matter how good the recommendations are.
TikTok's recommender is partly built on European Technology (Apache Flink for real-time feature computation), along with Kafka, and distributed model training infrastructure. The Monolith paper is misleading that the 'online training' is key. It is not. It is that your clicks are made available as features for predicitons in less than 1 second. You need a per-event stream processing architecture for this (like Flink - Feldera would be my modern choice as an incremental streaming engine).
I'm sorry to point out the obvious here, but who is going to perceive their recommended feed as slow or unfresh if it doesn't learn from exactly the last video you clicked on within 1 second? The bar simply is not that high. The special sauce of TikTok is how it chooses the videos, not the speed it does it at. I'm sure the speed helps to give it that "spookily intelligent" feeling, but that's a cherry on the recommendation cake, a cake which is already twice as good as the nearest competitor. I'm sure your talk goes deeper than this, but if this is the main focus, then you've missed the point.
Idk how to feel about this specifically but I kind of hope they come for Duolingo next. They are up to some similar mind hacking shit to keep people from leaving. There's the downright abusive streak management tactics that have become a major part of their brand and PR, and the lesson plans seem designed to plateau to prevent you from actually getting proficient enough in a language to ever unsub. They reset your cleared lessons and require you to redo them if they add new vocab to them, as well as randomly clearing them in the name of making you practice them again. I don't know what the solution is but I've known multiple people now who've gotten frustrated and blamed themselves for not being able to advance their skills with a language, but Duolingo's business model, like Tinder's, is completely opposed with the goals of their users. If Duolingo R&D discovered a magical new method of making you fluent in a language overnight, they would not sell it to you. Tinder R&D might have discovered the actual honest-to-God formula for True Love years ago and burned it because they can make more if you swipe forever.
> At this stage, the Commission considers that TikTok needs to change the basic design of its service. For instance, by disabling key addictive features such as ‘infinite scroll' over time, implementing effective ‘screen time breaks', including during the night, and adapting its recommender system.
Most of these seem concretely doable, and maybe effective. But the core of the addictiveness comes from the "recommender system", and what are they supposed to do there? Start recommending worse content? How much worse do the recommendations have to be before the EC is satisfied?
I use X almost entirely from the desktop where I have an extension installed that lets me whitelist my follows, and see nothing else. I recently browsed the same feed on mobile ... and it was entirely different! I think I spent a half hour and saw zero content from my follows, just one ticktok style video after another. For those who find these services without value, I now understand. But I feel revolted rather than addicted. Will I now experience a mysterious compulsion to view the naked feed?
Just curious for anyone who pays more attention to this than me: is the company being sanctioned by the EU for this behavior the one that US law forced an ownership change of or does that company only operate in the US?
TikTok has a lot of issues, such as privacy, dubious content, 'brainrot', etc. I don't want to seem like I'm necessarily defending TikTok specifically here.
But this really just stinks of Regulatory Capture to me. Their main argument is that the consumers like to use the app too much?
Why? Because it's smarter and not as enshittified as the competitors?
I'm sure if youtube, facebook, reddit, etc reduced the number of ads, and started showing more relevant content that people actually cared about, they too would start being "more addictive". Do we really want to punish that?
I'm skeptical about banning design patterns just because people might overuse them. Growing up, I had to go to the theater to see movies, but that didn't make cliffhangers and sequels any less compelling. Now we binge entire Netflix series and that's fine, but short-form video needs government intervention?
The real question is: where do we draw the line between protecting people from manipulative design and respecting their ability to make their own choices? If we're worried about addictive patterns, those exist everywhere—streaming platforms, social feeds, gaming, even email notifications.
My concern isn't whether TikTok's format is uniquely dangerous. It's whether we trust adults to manage their own media consumption, or if we need regulatory guardrails for every compelling app. I'd rather see us focus on media literacy and transparency than constantly asking governments to protect us from ourselves.
Has been really surprising to see HN increasingly turning pro paternalistic censorship and control, and increasingly skeptical about individual agency over the years.
75 comments
[ 2.9 ms ] story [ 62.9 ms ] threadHow is that any different to Facebook?
I can watch a 9 hour video on GTA games without problems (not in one sitting, but in parts), but 3 'shorts' in a row with not enough info and explanation to be interesting makes me close any of the 'shorts' apps (tiktok, youtube shorts, instagram....).
(eg, the 9 hour video: https://www.youtube.com/watch?v=Faxpr_3EBDk )
TikTok's recommender is partly built on European Technology (Apache Flink for real-time feature computation), along with Kafka, and distributed model training infrastructure. The Monolith paper is misleading that the 'online training' is key. It is not. It is that your clicks are made available as features for predicitons in less than 1 second. You need a per-event stream processing architecture for this (like Flink - Feldera would be my modern choice as an incremental streaming engine).
* https://www.youtube.com/watch?v=skZ1HcF7AsM
* Monolith paper - https://arxiv.org/pdf/2209.07663
> At this stage, the Commission considers that TikTok needs to change the basic design of its service. For instance, by disabling key addictive features such as ‘infinite scroll' over time, implementing effective ‘screen time breaks', including during the night, and adapting its recommender system.
Most of these seem concretely doable, and maybe effective. But the core of the addictiveness comes from the "recommender system", and what are they supposed to do there? Start recommending worse content? How much worse do the recommendations have to be before the EC is satisfied?
Its a good thing, but its not what the title says it is
LinkedIn has become such a pit of force-fed self-help vitriol it’s completely lost its purpose.
What makes TikTok different?
But this really just stinks of Regulatory Capture to me. Their main argument is that the consumers like to use the app too much?
Why? Because it's smarter and not as enshittified as the competitors?
I'm sure if youtube, facebook, reddit, etc reduced the number of ads, and started showing more relevant content that people actually cared about, they too would start being "more addictive". Do we really want to punish that?
What's the end goal here?
You can't legislate intelligence...