Jonathan Stray
jonathanstray.bsky.social
Jonathan Stray
@jonathanstray.bsky.social
Knowing things is a solved problem. Getting along is not. Working on AI, media, and inter-group conflict @CHAI_Berkeley. Got here from computational journalism.
I don't know it, do tell
November 5, 2025 at 7:25 AM
Tell me about that? How would you set up your algo?
November 5, 2025 at 7:04 AM
I experience toxicity as predominantly coming from other users.
November 1, 2025 at 7:54 PM
I love that! What's your idea?
October 31, 2025 at 7:24 PM
If "media" isn't talking about it, how do you know about any of it?
October 25, 2025 at 4:02 PM
You may be interested in subscribing to my newsletter
greenearthsocial.substack.com/p/introducin...
Introducing GreenEarth
We're building advanced open source algorithms for social media
greenearthsocial.substack.com
October 25, 2025 at 4:00 PM
Idk it worked for me!
October 25, 2025 at 3:54 PM
Are you two working together again?
October 24, 2025 at 4:18 PM
Haha no. You don’t want to hire me as a professional programmer these days.
October 24, 2025 at 2:11 AM
That's fine! You will never be forced to use it. Or you can code up your own algorithms within our indexing infra. That's kinda the point.
October 23, 2025 at 5:01 PM
Yeah we imagine the "feed builders" will create many pre-configured feeds with good defaults.
October 23, 2025 at 5:00 PM
Lol that you'll have to take up with @support.bsky.team
I imagine you've tried docs.bsky.app/docs/tutoria... ?
Muting users | Bluesky
Muting a user hides their posts from your feeds. Mutes are private, and stored
docs.bsky.app
October 23, 2025 at 4:59 PM
And yet there are plenty of Marxists in the academy for economists to argue with, esp in the social sciences... www.econlib.org/archives/201...
The Prevalence of Marxism in Academia - Econlib
As the Iron Curtain crumbled, people often joked, “Marxism is dead everywhere – except American universities.”  The stereotype of the Marxist professor runs deep.  But is this stereotype grounded in s...
www.econlib.org
October 22, 2025 at 9:42 PM
Yeah that's pretty much exactly the point. Showing "bias" would require that more people are recommended content X over content Y even though *everything else is the same* between the two and everything else is never exactly the same.

To their credit, they end with a call for algo transparency.
October 22, 2025 at 4:40 PM