Mercilessly tightly curated email, mastodon, diaspora, and reddit.
But really: books. Mostly books.
Over time, I’ve come to realize that this is fundamentally impossible without “mercilessly tight” moderation, and tools/services that give you that power (RSS, email), because the state of the world is 1 part signal to 99 parts noise.
Books, and by extension libraries, are a viscerally physical representation of this principle. Much of what’s in a library is noise, but the signals are sustained, dense, and often unlike anything you can find on the web. However, like the web, signal tends to cluster around certain authors and publishers. There’s still a moderation problem (picking the right book), but over time your verification algorithm (is this book signal or noise?) gets faster and “mercilessly tight”.
I don’t have much of a conclusion, but I’ll end with questions this raises for me. I wonder how this idea interacts with “echo chambers”? Aren’t we constructing personal ones with these tools and resources? Is that necessarily a bad thing?
I'm currently curating from a listing of 25k+ law articles on a topic of interest. Availability of those (free, online) has a huge influence. Thanks to the Library of Alexandra (Sci-Hub) and LibGen.
Adler's How to Read a Book is quite useful -- the synoptic approach especially.
Making rapid assessments of suitability is crucial. Being prepared to revisit those assessments, either way, later, also. Developing indicators of crap or possible gold help immensely. Again, see Adler.
I spend considerabe time with older sources. They're often known to be wrong or inaccurate, but:
* The path they reveal in development of understanding is often hugely useful. To understand the bug, it helps to know how it came to be.
* The rhetorical and ideological battles to which they may have been part is usually now known, often spent. Literature -- science, technology, fiction, philosophy -- is all tremendously ideological, and being removed from the frame under which it was constructed, or being aware of it, helps immensely.
* One often finds one's own crazy notions expresed there, sometimes partially, sometimes far more clearly. I've generally found that the questions I've been most concerned with are in fact long-standing ones.
* Much current discussion retreads older thinking, though aparently unconsciousy and/or in complete ignorance.
The Copyright Abyss -- many materials published since 1924 -- is aboth a hinderance and a blessing. It's not complete (LibGen, Sci-Hub, ZLibrary, Open Library via Archive.org), but it is sufficiently effective that it tends to force consideration of earlier works.
1. The US Library of Congress houses 24 million catalogued books and over 168 million items in total. Its annual reports show about 300,000 new book additions annually, a remarkably stable rate, since the 1950s. Bowker issues about 300k ISBNs annually to traditional publishers, and another million to 'untraditional" (self-published) authors. Google have estimated 129 million books were published, ever. And there's the question of how many books can matter, culturally. About 3/4 of Americans read at least a bok a year, which suggests a floor somewhere around 60 books over a lifetime. Given that there are other informational channels, how much information must be transmitted intergenerationally to preserve culture -- technically, socially, values, etc? "Of the making of many books there is no end, and much study is a weariness of the flesh."
4. https://loc.gov/about/reports-and-budgets/annual-reports/ http://catalog.hathitrust.org/Record/000072049 
5. One of the exceedingly rare cases in which Hathi Trust is actually useful.
I am however very much interested in how you find the email lists to curate.
Has to be someone with curiosity, something to contibute, willingness to correspond (I'm a pseudonymous space alien cat, acquired taste), consideration of my time (and likewise). It's a small set.