Conversation

Replying to
Biden doesn't mention Tesla anytime he talks EVs/climate due to non-unionization (which is pretty petty), so musk has been pissed about that, so there's no way that doesn't factor into the calculus. "You won't acknowledge Tesla, I'll unmuzzle your nemesis."
3
34
Worst case he tries to run it himself in a spacex way… try to blow up a lot of rockets to learn to launch them well. As in let lots of twitter conflicts blow up. The problem is you can keep people a safe distance from trial and error rocket experiments but not social ones
4
31
Jack wanted to let anybody write their own feed algorithm against the api iirc. That’s really the big potential win here. If I can make (or more likely, select from a marketplace) a feed algorithm to suit me, I kinda don’t give a shit what the free speech crowd does.
3
115
It will also take the wind out of the sails of free speech posturing by creating a meaningful exit option. Right now the best in-platform exit option is protected tweets and that is no good. As it stands only crude mute/block mechanisms exist and they don’t scale.
Quote Tweet
Replying to @vgr
If he says that he’s going to open source the algorithm it’ll be unprecedented In the history of social media.
2
15
For “free speech” I’ve lately started mentally substituting “remove all ddos protection and sibyl resistance.” If you’re an internet-of-beefs knight with enough mooks at your command, you can swamp you’re opponents ability to hold a mute/block perimeter.
2
34
There are asymmetrically more (and more powerful) mechanisms available today to amplify reach than there are ways to tune people out *at scale*. I’d like, for instance a learning algorithm on my side, like a spam filter, that learns from *my* blocks, mutes, and muted words
1
31
And I’d like to share and merge my filters in useful ways with those set up by trusted people. Unfortunately every bit of user control over filters (“free tuning” perhaps, like “free speech”) erodes advertising potential, or we’d have all this already.
2
23