That's exactly it. The whole point of YTK was to protect our kids from the fire hose and provide a trusted safe space. About a year ago it was. Our son was watching Blippi and videos about banana slugs. We deleted the app once we saw the dark shit slipping in there and it was clear YouTube was not acting to address it.

The real question is: "Can an algorithm identify child-friendly content?" The answer is currently unknown. Families are basically the "beta testers" of the algorithms in question, but the downsides of that approach, like exposing kids to really disturbing content, are horrifying.

Child-friendly isn't even a well-defined objective category; different parents will disagree violently (sometimes, quite literally) over what is appropriate.

Marketing your product as kid-friendly and then demonstrably exposing children to abusive content, through “algorithmic negligence” or otherwise, should incur serious penalty.

It never ceases to amaze that the kind of psychological engineering these companies engage in is outside the bounds of any established ethics, but somehow they escape unscathed because money.

How about "Do we need algorithms to find child-friendly conent?"

I was surprised to read those posts re: YouTube Kids. My daughter has watched a few thousand hours of YT Kids over the last two years while she's been bed-ridden with leukemia. My wife or I have been within earshot for almost all of it. 99% of what she watched was human made content and the most risque thing she saw was a video where a "host" pretended to make some stuffed animal poop gumballs. A good chunk was actually educational—she went on a kick learning how to make healthy smoothies and learning about the solar system, state capitols, etc. Perhaps we've just been extraordinarily lucky, but the reporting re: TY Kids this week reminds me of the 1980s panic around Dungeons and Dragons.