Could it be possible to counter it with another ML model that browses your feed?
For example, scraping your feed and presenting to you only the content that corresponds to some pre-defined labels (with a tiny bit of randomness to spice things up).
Although how could the automatic labeling work for videos from the user-end? Hashtags would be the simplest indicators, however also easily misleading.
If only somehow we managed to make social media uncool for the kids, that’s the most sure way they’d stay away from it.
I guess proper education on the real aspects of the social media phenomenon would be the real deal. For example, explaining how/why the companies use their algorithms to keep you in there; influencers only want to sell you a product; why posts/stories don’t reflect reality at all, etc.
But understanding all that would require quite some amount of emotional maturity from both the kids and parents themselves. Sadly, that doesn’t seem to be the reality at all, there are adults that still can’t see through the cracks..
For example, scraping your feed and presenting to you only the content that corresponds to some pre-defined labels (with a tiny bit of randomness to spice things up).
Although how could the automatic labeling work for videos from the user-end? Hashtags would be the simplest indicators, however also easily misleading.