Oddbean

I swallowed the black pill on algorithmic social media, so doomer warning is in effect. Proceed at your own risk. In my mind, everyone is an alcoholic, drunk, at the bar, and a bunch of these drunks are saying things would be better if we brewed our own beer. Most people will use the algorithms to see what they want to see. Dopamine will flow. Social bubbles will form. They won't be challenged. We'll see distributed npub block lists. More people will be phone zombies. Fedi showed us this will happen. I think chronological feeds are a solution, along with great search functionality. It's not bad to have to put in work to find content. But that's just my opinion. At the end of the day, it's really cool that people get to do it however they want. Even if my doomsday scenario happens, at least there's the option of just turning it off. Can't do that anywhere else.

I’ll start by saying “thank you” to you because without nostr we wouldn’t have the freedom/ability to run such experiments. I agree that a bunch of people will revert to dopamine, social bubbles, and phone-zombie behavior like you suggest if given an easy glide path. But I wonder, who will build/optimize those algorithms? If developers are not incentivized by being coupled to an ad serving platform maybe we’ll see different results — maybe we’ll get some of the same, but we could also get a lot of wild experimentation that turns out to discover other things people want. I suspect we could see different algorithms at different points in a user lifecycle/journey. One algo to onboard new users to their first few follows/connections. Maybe a new interface that lets people preview content before selecting who to follow. Maybe there should be a lot more “for you” content driven by notes and following suggestions should be a side-effect of which notes you react/zap from people you aren’t yet following. There’s just so much greenfield to explore and nostr makes this possible. Corporate controlled algos will all trend toward serving the corporation’s goals, but open/transparent/switchable/pluggable algorithms can be built to serve end-user goals which will almost certainly be different since no one’s primary goal in life is to view an ad. I think “proof of work” on finding the good people/content is fine for now — slightly annoying, but not devastating. We put in the effort. Though we miss so many valuable ideas/connections this way. But I believe nostr has the right architecture to serve most publishing needs on the Internet and create a whole new layer of what is “the web”. I would estimate there are at least 1B+ public messages shared on the Internet each day. There’s not enough time in my day (no matter how committed I am to “doing the work”) to filter them all for the stuff I care about or might like to see. I need computational help (i.e. an algorithm) to do this properly. What great people/ideas/relationships could come into my life if only there were a way to match us with the right context? Open protocols rule. Thanks for making nostr, man!

My brother, I'm not sure at this point, but I may have accidentally deceived you. 🫂 nostr:nevent1qqsp0evgqkt92v4lacd25hyvmsazqd6tc60fgkmtfvpe48x6y9xdc6gppemhxue69uhkummn9ekx7mp0qgstwz8h8yh43pqxyykr3qh8kw7qmxcg6chet7shp5yezflvufmsuhsrqsqqqqqpuc7su7