This seems like it could be substantially better than current systems. It could be effectively a crowdsource / peer recommended model where you opt in to aspec

This seems like it could be substantially better than current systems. It could be effectively a crowdsource / peer recommended model where you opt in to aspects of the bubbles. So my elderly parents could get some of the safety guidance mirroring mine (eg weeding out conspiracy thought or spam) but then also picking up aspects that mirror older peers (eg reflecting their more traditional tastes) Right now there's a remarkable amount of undesirable content that gets through that ought to be easy to filter. On Twitter you see it a load with searches for regular terms that show up inappropriate content for the term, yet you don't want to set content filtering or it cuts things that are genuinely interesting but edgy!