[gjvnq]

A big issue here would be defining social media.

Are forums social media? What about reddit? What about YouTube?

I think what we really need is a ban on algorithmic recommendations that seek to encourage engagement or total time spent on the app.

[trainerxr50]

I don't see how this would work either.

Banning algorithmic recommendations would need to ban search engines.

Social media is ultimately just a website. Anything I can think of quickly gets down the road of banning the web browser and/or banning email.

The only solution is people have to have the freedom to use these stupid platforms or not. People have to have the freedom to be stupid on stupid platforms.

Brazil is going in a much different direction.

[terribleperson]

My take would be something like this: Any public or public-ish website, or website with more than x user which presents algorithmically sorted or curated content must make readily available the source feed for their algorithms as well as any other information they use. On any page where algorithmically sorted or curated content is presented, they must fully describe the algorithm used. Ditto anywhere you select an algorithm or navigate to an algorithmically curated or sorted page - it must be described fully in the button or selector, or adjacent to it. If that is impractical for space reasons, it should be described as best as possible with footnotes expanding on the explanation. Furthermore, the explanation, source feed, and additional information should be complete and clear such that someone could reasonably recreate a page or sorting given the algorithm, source feed, and additional info. This would be the test used in court if someone alleged infringement.

My hope is that such a law would heavily bias sites towards simple, less manipulative algorithms.

[infotainment]

TBH, I think that wouldn't solve the core problem.

If you forced, for example, TikTok to do this right now, they would presumably add a page to their app with their recommender algorithm. Then what? Meta or other competitors might be interested in copying aspects of it, but normal users would likely ignore it and continue being addicted to TikTok.

[terribleperson]

A page on the app wouldn't cut it. Has to be on the same page. I'm tempted to say "has to fit in the button".

[porridgeraisin]

If you go to the root cause, the reason they want to maximise user-minutes is because it is in turn proportional to ad minutes.

Banning targeted ads will greatly reduce the benefit of ads (to the social media company) since they are rendered less effective. This will tip the scales of the cost-benefit tradeoff that the company makes. In this case, the cost of the ad is that it's annoying to the user. Every ad company chooses a tradeoff. If you made the benefit smaller, then they would have to reduce the cost too, which would lead to lower ad volume, which would reduce the incentive for engagement.

Any other way to reduce the effectiveness of ads also works. I'm sure the method I suggested(banning targeting) is not bulletproof, but they key thing that needs to be done is artificially reducing the effectiveness/relevance/quality of ads.

[rjmunro]

Surely if you ban targeting ads they'd have to catch up the revenue, and that would mean serving you more ads, worse ads and using stronger algorithms to ensure you stay longer. I don't think it makes any difference.

[Spivak]

Yes, yes, and yes. You picked basically the three most obvious examples of social media.

Is it a site that hosts user generated content and makes that content available to others in any 1-n fashion? Great, you have social media.

[t-writescode]

Is a group chat? Is Discord? What if the Discord invite link is public?

[infotainment]

I think this is a reasonable take; a good start would be banning all forms of algorithmic "discovery" recommendations, and things like "for you" feeds.