An auto-connective catalogue of all your favorite thinkers.
Here's a list of some of my favorite current-day thinkers:
I follow all of them via either their podcast or blog (rss feed), which in turn also occasionally makes me aware of their less trackable/searchable media content such as books or videos.
Recently I read an older piece by Cory titled Schizmogenesis. In that article he recommends the book The Dawn of Everything:
It’s a genuinely remarkable book, as is evidenced by the different angles stressed by reviewers. If you haven’t read Dawn and you tried to figure out what it was about from the reviews in the Guardian, the New York Times, and Crooked Timber, you could be forgiven for thinking that they were reviews of three different books. But if you’ve read it, it makes perfect sense that the reviews would have to pick an area of focus for any summary, and that there are many such foci to choose from.
This isn't the first time I've heard of this book. In fact, I've certainly heard it mentioned/recommended before by one or more of the other thinkers listed above. And that's key here. There are too gosh darn many good books to read, so I have to be very judicious about which ones I commit to.
My simple heuristic for choosing books is some version of the Rule of Three: The first time I hear an interesting book mentioned in a blog post or podcast, I make a mental note of it and move on. But if that book gets serendipitously mentioned (i.e. by happenstance rather than during a book tour media frenzy) by another thinker I follow, and then another, I feel strongly compelled to pick up the book. Even though I cannot exactly remember the other contexts in which I heard 'Dawn' mentioned, I have a general feel for it being brought up on numerous occasions on a positive note.
This is the sort of thing I want Thinks-app to do: Automatically draw connections between thinkers I follow, and make me aware of their common ideas/interests/influences. I don't want to put an end to my personal 'serendipity heuristic', but an automated discovery mechanism might cause me to pursue certain resources sooner rather than later.
How can it do that? Here are some preliminary ideas.
Partner with something like https://spiderwebai.xyz/
- Spider crawls the RSS feeds of of my listed thinkers, as well as the full websites those feeds live on.
- Spider looks for recurring topics across the thinker-feeds, and tells me about the highest scoring matches.
That's pretty much it. Ideally we could apply the same technique to podcast transcripts. Therefore..
Podcasts can't be searched without transcripts. There was a pretty good example of this called podtext.ai, but they've since shut down: https://web.archive.org/web/20231226230527/https://podtext.ai/planet-money/the-new-biden-plan-that-could-still-erase-your-student-loans
Podcasts could be transcribed on demand though, either via a paid API or a self-hosted version of something like Whisper:
- https://cloud.google.com/speech-to-text
- https://www.descript.com/tools/speech-to-text
- https://www.rev.ai/
- https://platform.openai.com/docs/guides/speech-to-text
- https://huggingface.co/spaces/lmz/candle-whisper
Thinks-app is essentially an RSS reader that (1) supports both article feeds and podcast feeds, and (2) draws connections between knowledge artifacts (articles & podcast episodes).
The latter is accomplished with the likes of Spider. The former would require a mashup of something like these two:
An extension of an rss reader like Lettura would largely suffice, since the app doesn't need to facilitate the listening of podcasts, but rather just the textual parsing of them.
Thinks-app would consist of two content primitives:
- Thinkers: any person with a website & consequent feeds.
- Topics: dawn-of-everything, meta-crisis, enshittification, localism, loving-kindness...
Each thinker would have a linkspage-like profile based on their feeds. Underneath the feeds would be a list of their most prevalent topics, coupled with outgoing links to other thinkers who've also touched on these topics. Both the thinker pages and topics would be amenable to some form of crowdsourced editing, but there's so much complexity there that it's best to first envision a version of the app that could be Good Enough with the most basic form of thinker-curation and topics-automation.
There are also major follow-on effects here by adding a social layer, such as curated collections of like-minded thinkers or topics. The possibilities are vast, but the most basic value-add from a sprinkling of social connectivity would be the accumulation of thinker profiles.
For instance, I'm not a fan of Jordan Peterson, but if he has talked about the meta-crisis, perhaps even together with one of the thinkers I am more aligned with, that's interesting to me, and that content might challenge my thinking with a point-of-view I don't usually pay heed to. Furthermore, I'd be doubly encouraged to listen to someone outside my sphere of intellecual comfort if that person/resource is recommended by one of my friend-peers in the Thinks-app network.
Possibly useful resources:
There's some similarity to Noosphere here ("A protocol for thought", or as I like to say, "A social network for your ideas"), but Noosphere is a tool for thought creation whereas Thinks is a tool for thought subscription. What they have in common is being tools for thought connection, so there's an avenue for convergence there over time.