Flagging what children can see/can do is a necessary part of education, the question is where to draw the line and what is considered damaging enough that it needs to be marked as off limits

Yes, but that's a choice ought to be made by the parents. Not the government or some other party. It's the parents who choose how to raise their kids.

If something because a problem for the whole society, parents won't have their say in it. For instance violent content would arguably fall in this bucket.

For this specific discussion, the question would be if children exposed to some of the most predatory apps have provably destructive behaviour (for instance if it affects their health such that they are not able to participate in regular social activities, or get depressed and because bring a high cost for the whole society etc.)

I'd see censoring at the society level or not depend on the answer to that question.

I noticed you didn't want to use the word government there and instead used "society". More often than not it's unelected bureaucrats doing the regulating, so it's not "society".

I think that more often than not the people pushing these decisions follow an idea of what most people would agree with.

For instance even if a regulator deeply believes eating babies is OK, they wouldn’t push a law in that direction, understanding nobody would back them up.

They would still be biased and try to limit punishment on baby eating, but there would be enough other people countering them.

In a way, if nobody resists or succesfuly rejects a proposal, I’d argue that proposal goes along what society accepts as valid, even if if asked in a poll a lot of people would disagree with it.

Calling it censorship makes it sound bad. But if you think about it, we have these restrictions of behaviours outside the digital world already, sometimes with options for parents to override it.

Where I live, alcohol can be consumed by children if given by their parents, but child pornography is illegal regardless of consent.

Predatory digital apps have a certain level of danger, and we need to figure out how much society should protect us from it.

I'd guess that most parents agree that they don't want their kids drinking alcohol and don't want others selling it to the kids. That's why this behavior is made illegal.

However, I do not think that there would be such consensus on other questions, like when you should tell how babies are born, what politician is not doing a good job, or what books and TV shows should be banned for kids at home, etc. There is no agreement on these matters so there should be no laws governing them.

At least, I assume that the people you want to give the freedom to decide matters by themselves, are also people who would accept changing their mind (at a greater scale changing the consensus) when faced with credible facts going against their beliefs.

To take your “when you should tell how babies are born” example, I think we agree at some point kids need to know. There is no consensus on when it should happen exactly, but we know a portion of the kids will have life or death issues if it’s not done at puberty.

As I see it, idealy we’ll want to put laws to enforce sex education at puberty, and let parents do what they want before that.

You can censor the prevalence of something on a specific platform, but users will always be able to link to any content, if they choose to do so.

You would also fail to make the internet safe for children. I know, children do stink, but they are probably better at locating information they are looking for than their parents. Only direct parenting can help here because there is no feasible technical solution.

So we could draw lines all day and it wouldn't really matter anyway.

The claim that socially media don't select and filter is nonsense. Of course they do. They're a huge part of a social feedback loop that selects and reinforces the most average "popular" behaviours, while filtering out more challenging high-effort social edge cases.

IMO the edge cases are where all the interesting things happen, so keeping them out of the public eye is a form of censorship itself.

I can think of many improvements:

* Boosting instead of retweeting can avoid viral outrage, explanation - https://www.codesections.com/blog/mastodon-elevator-pitch/

* Reduce the incentives & rewards for signalling and curating an online identity, e.g. avoid Like buttons and Share'd content. Snapchat shows it's possible for Like-less social media without curated identity to still be popular, despite lacking Like-button dopamine hits, and the vacuum left by viral content and image grooming might¹ be replaced with socializing with your friends and ability to be yourself (¹ see https://news.ycombinator.com/item?id=11828280)

* Should "goes on your permanent record" always be the platform's default behavior for social interactions?

But I'm just a mere coder watching social media make my social circles ill, I wonder what systems real psychologists could come up with and test.

You know what else gives people dopamine shot? The next piece of chocolate cake, the cigarette that's always overdue, having the drink after long day of work, finally being able to hit the bong and so on.

"Social media" is as much of an addiction on a biological level as drugs. We know that prohibition doesn't work.

We should start with helping people find ways to feed their (dopamine) reward system with more meaningful stuff. Help accomplish them small and bigger things, promote real human interaction.

Or, put another way, if there's a wasp buzzing in the room, plugging one's ears to its sound doesn't make it go away. You just no longer know when you're about to get stung.

As for how to navigate these platforms responsibly- people have to care about consequences, and social media of the modern form is still too new for society to have come to grips with it. I think it will happen with time, as a mix of cultural and legal adaptations. As to what form that will ultimately take, well... Good question, that.

If anything, you should parcel out some examples of what that would look like instead of +1'ing some abstraction notion of it.