From a technology perspective, that's a much harder problem than it sounds. To the degree it's even possible, it's a good way to make sure that social networking gets consolidated into the hands of a few companies and that never get challenged. Or that social networking breaks back up into country-specific sites, making the "world-wide" part of the WWW more of a technical possibility than a practical reality.
Also, the discussed German law (which I'm not particularly fond of) addresses the problem you raised: it only applies to the largest companies giving startups enough time before needing to solve these issues.
So, again, it's probably going to reinforce information bubbles as places to share thoughts and ideas will stay tied to a nationality and/or small.
> ...this is a bad excuse for giving up trying at all.
I think all of Western Europe is struggling with information bubbles as surprising election results have been happening one after another. Is it "giving" up to not double down on that phenomenon?
Maybe I'm misunderstanding what you mean by "just block the post" but it seems the algorithm for writing:
is_illegal(image, country)==true
.... is currently not possible to write with accuracy. (Relevant XKCD:
https://xkcd.com/1425/)If it's a large-scale service used by a billion people posting user-generated content, you have to rely on algorithms to filter it instead of depending on users manually flagging posts. Users themselves don't know their own country's laws and would flag things like the Vietnam naked girl photo as "illegal" when it really wasn't.
Even if a machine learning algorithm can be trained to flag a specific cartoon of 3 religious figures having sex as illegal, there are near-infinite variations of other drawings showing the same thing that would get passed as "legal". In fact, it doesn't even have to be a totally different comic by a different artist... just changing 1 pixel might do it: https://news.ycombinator.com/item?id=15577885