Hacker News Reader

Comment by tboyd47 | Hacker News Reader

It all comes down to the assertion made by the author:

> There is no way to run a targeted ad social media company with 40% margins if you have to make sure children aren’t harmed by your product.

View on HN

I find it hard to see a way to run a targeted ad social media company at all if you have to make sure children aren't harmed by your product.

don't let children use? In TN it that will be illegal Jan 1 - unless social media creates a method for parents to provide ID and opt out of them being blocked I think?

Wouldn't that put the responsibility back on the parents?

The state told you XYZ was bad for your kids and it's illegal for them to use, but then you bypassed that restriction and put the sugar back into their hands with an access-blocker-blocker..

Random wondering

Age limitations for things are pretty widespread. Of course, they can be bypassed to various degrees but, depending upon how draconian you want to be, you can presumably be seen as doing the best you reasonably can in a virtual world.

I'm not sure about video, but we are no longer in an era when manual moderation is necessary. Certainly for text, moderation for child safety could be as easy as taking the written instructions currently given to human moderators and having an LLM interpreter (only needs to output a few bits of information) do the same job.

That's great, but can your LLM remove everything harmful? If not, you're still liable for that one piece of content that it missed under this interpretation.

There are two questions - one is "should social media companies be globally immune from liability for any algorithmic decisions" which this case says "no". Then there is "in any given case, is the social media company guilty of the harm of which it is accused". Outcomes for that would evolve over time (and I would hope for clarifying legislation as well).

What about 0% margins? Is there actually enough money in social media to pay for moderation even with no profit?

At the scale social media companies operate at, absolutely perfect moderation with zero false negatives is unavailable at any price. Even if they had a highly trained human expert manually review every single post (which is obviously way too expensive to be viable) some bad stuff would still get through due to mistakes or laziness. Without at least some form of Section 230, the internet as we know it cannot exist.

Reply on news.ycombinator.com