ncase
patreon
Quick Update Sep 2025
🕑 Added 2025-09-24 18:41:24 +0000 UTCHi all! I've been under the weather this month, so instead of a vlog, here's a quick text-post update:
Stuff done this month:
๐๏ธ Talked with Audrey Tang(!), Digital Minister of Taiwan, about possibly doing comics/videos for her upcoming book on a more humane take on AI Alignment! More info if you're curious: https://6pack.care/
๐ฏโโ๏ธ Made a working prototype for an "LLM Clone" of myself! This LLM agent can 1) browse writings about my beliefs, values & memories, 2) do research-brainstorm & write-revise loops, then 3) imitate my thinking & writing style. For more info on this creepy experiment, see this pitch for a grant I did not get: https://docs.google.com/document/d/1s0AAHmWtP90vcz0_Cks1JSDEPJJ5SwgqFxZb1xY2UgE/edit?usp=sharing
๐ฟ Only a little bit on CatTuber, and near-nothing on AI Safety for Fleshy Humans Part 3. :/
Signal boosting:
๐ค The MATS Program, an AI Safety bootcamp, is taking applications for Winter 2026! It's 12 weeks long, there's a stipend, travel & accommodations are covered. There's one location in Berkeley California, and another in London England. Deadline to apply is Oct 2nd: https://www.matsprogram.org/
๐ง My friend Tobias Rose-Stockwell just started a new interview series about how algorithms & AI are reshaping humanity, for worse or... well, mostly worse. His first interview was with Jonathan Haidt, one of the few social psychologists whose work, like, actually replicates. Check out the first interview, it's full of useful ideas: https://www.youtube.com/watch?v=wFnrW9Pp9KI
That's all I got for this month! Sorry :/
~ Nicky
