Yeah, I certainly want people who want to see new people, topics, and points of view to be able to do so. Fortunately, I think networks of trust are pretty good at getting you to far away ideas pretty quickly though. Veritasium actually did a video on this quite recently but basically even a small number of connections outside the "bubbles" create bridges that connect networks really well.
I think a system that put this into practice would need precise controls over who gets added to your network (e.g. friend this person but don't add their friends to the network of trust, or block this person and prevent any of their friends from being added to the network of trust, or friend this person but privately so others can't use that connection in their own networks of trust, etc.). That would help with balancing seeing a lot of posts while still "exploring" the network at your own pace and comfort level.
And today discoverability is still pretty weird and closed off. You either get algorithmic feeds that just show you shit they think you'd like, and you're likely to see posts from influencers a bunch, causing para social relationships, or nobodies who you've never seen before and never will again, making it hard to form a connection. Or, you have closed off groups like discord servers that aren't very "permeable". You can't, say, use the server you like and use it to "explore" similar communities.
So alongside the network of trust I think being able to "traverse" the network would help discoverability. I picture this as like a digital neighborhood, where being friends with someone is like them being your neighbor, and you can "walk down the street" so to speak, and the further out you walk the further the people may be from your "bubble". I don't know exactly what that would look like in practice, but one idea I've played around with is myspace style pages. So instead of a global feed, you go to someones page, see their friends and some posts that spark your interest, so you go to the friends page, and so on.
But IDK, this is all theoretical and IDK exactly what such an app would look like. But I think the discoverability problem is solvable in a network of trust based platform
I'm really into youth liberation and the various ideas surrounding rethinking education. I'm a parent and am sending my kid to a Montessori, but I'd really love a whole system where child raising and education are seen less as responsibilities for the nuclear family unit to have full control over and a communal responsibility that everyone participated in. Letting kids guide their own education (under the general advice that anything important to learn will be intrinsically motivating, and anything that isn't can be encouraged through instilling values rather than forcing curriculums), making things mastery focused rather than using grades and limiting ones future prospects based on their performance on tests (which, more than anything else, simply correlate with things like food security and access to tutors and stuff), generally getting rid of the concept of kids being rankable in intelligence, and above all defining a box of what's "neurotypical" and what's not (in favor of the neurodiversity framework).
With that in mind, I liked that this article made some good points about how this process should be decentralized and involve communities determining what they think will work best for them, and being okay with some not achieving the goals they set out with.
But I have to say, it was pretty hard to read. In particular, the negative parallelisms every 2nd sentence just interrupted the flow with constant thoughts of "this is likely written with the help of ChatGPT".
Also, man it can be hard to find local community members to try any of this stuff with. A lot of the resources in unschooling are also tied to religious and/or conservative families wanting to homeschool so the parents can better enforce specific ideas on their kids, rather than seeking child liberation.