I wanted a prediction market on when Asahi Linux would have some stuff done, and now it exists (albeit with play money): https://manifold.markets/DanielFilan/when-will-asahi-linux-be-ready-enou?r=RGFuaWVsRmlsYW4
But algorithmic choice looks cool, as does domain names as handles as an identity verification method. https://blueskyweb.xyz/blog/3-6-2023-domain-names-as-handles-in-bluesky
Here you can read about my "everyday carry" https://danielfilan.com/2022/12/29/edc.html
Fine video but worse than I would have guessed - was hoping for a discussion of when mechanistic reasoning works, why it fails so much, etc, but it's mostly case studies in people being wrong. https://youtu.be/NmJsCaQTXiE
@alexaltair I had the vague impression that they aren't yet great for the domains of math that people actually work in.
@joshyou Hey
Oh also Words Bubble Up is nice and colourful (Mishima film also has nice use of colour)
Just a LWish dude.
I make podcasts that I think are worth recommending: if you like hearing about AI x-risk research, check out AXRP (the AI X-risk Research Podcast, axrp.net), and if you like random interviews with whoever I want to talk to, check out The Filan Cabinet (thefilancabinet.com).