Maxim Raginsky
@mraginsky.bsky.social
6.6K followers 950 following 2.4K posts
web: http://maxim.ece.illinois.edu substack: https://realizable.substack.com
Posts Media Videos Starter Packs
mraginsky.bsky.social
I only read him in Russian. His prose is powerful, but it is not “literary” in the conventional sense.
mraginsky.bsky.social
That could be an artifact of the translation.
mraginsky.bsky.social
Ok, I’ve read that Guardian article to the end. Serious Dugin vibes (modulo the appropriate inversions), except Dugin can produce texts like this basically at the drop of a hat.
mraginsky.bsky.social
Exactly, and here “illegible” has the precise meaning that the model cannot be represented as a composition of truth-preserving transformations acting on these symbols.
mraginsky.bsky.social
I hope he quotes Arendt‘s “The Human Condition” next.
mraginsky.bsky.social
I don’t know about others, but I always emphasize in my undergrads controls class that the Laplace transform matters because of initial conditions and the arrow of time. Fourier transforms only describe the steady state.
mraginsky.bsky.social
But what exactly is the katechon according to Thiel? Who or what is doing the restraining?
mraginsky.bsky.social
closing of the spritz season
mraginsky.bsky.social
Conan O’Brien had a skit about a superhero named Captain Punishment who did exactly that.
mraginsky.bsky.social
Right. Plus the kiwaguro ink is very high quality, I have not noticed any bleeding.
mraginsky.bsky.social
mraginsky.bsky.social
Sample-efficiency of human intelligence was paid for in aeons of evolution (“what’s a priori for an individual is a posteriori for the species,” like Konrad Lorenz said).
mraginsky.bsky.social
Yes, that’s the main attraction.
mraginsky.bsky.social
Yes. Matus Telgarsky converted me to dots many years ago. Before that, I used to buy regular grid.
mraginsky.bsky.social
Leuchtturm notebook with 120g paper (dot grid) and Sailor Kiwaguro ink. Nice!
mraginsky.bsky.social
It's excellent, and was a huge influence on Putnam's internal realism.
mraginsky.bsky.social
Dijkstra had a spicy take, way before LLMs, and much of what he says there just makes sense, especially when it comes to nondeterminism vs. robustness: www.cs.utexas.edu/~EWD/transcr...
E.W.Dijkstra Archive: On the foolishness of "natural language programming". (EWD 667)
www.cs.utexas.edu