For me at least, maybe not the people around me lol

Very cool. Now I'm daydreaming a CubeCon-like event.

Everyone brings their own shared deck/graveyard deck. Deck and starting player for game 1 is decided randomly and independently of each other; loser of previous game chooses deck and starting player for next game

(the whole thing is :chefskiss:) but what a sick-ass manabase

We wish we had the technology to observe the passage of time in a device we could carry on our wrists or in our pockets. Sadly we haven’t invented that yet

Love reading this on the way to the airport

Don’t do poor phinneas gage dirty like that! He just became a rude drunk, and not a self aggrandizing fool

If you're successful in that organizing, your incentives are to continue doing so, and to keep your position

If you're unsuccessful in that organizing, you lose out materially _and_ you'd feel discouraged from further organizing.

Not good!

(NB rank speculation) and then babby's first political organizing experience revolves around trying to maintain that advantage for themselves. Imagine where that leads!

I think the minimum description length principle is often described as a mix of "making kolmogorov complexity arguments tractable and landing at something like ockham's razor", and it has deep connections to bayesian considerations, yeah.

As always, @davisvaughan.bsky.social is the wizard who rescues me from the depths of tree sitter madness. We’re so back, folks

But also,

Use tree-sitter, they said. It has great Rust bindings, they said!

here I am, writing #define macros and hard-casting pointers like I'm in undergrad operating systems class again. Except debuggers don't work nearly as well as gdb in emacs did back in 2000!

Computers were a mistake

I think I understand now, thank you for putting up with my questions!

If I understand that right, then the robustness of the measure zero argument will become increasingly load-bearing the more "adversarial" your chosen prompt set is (which I think will defeat the empirical utility in many cases)

So the argument is entirely empirical, they're using "almost all" mathematical language to imply "the things we ran on it", and proving theorems?

Um...

There's no burnt toast, smell, or Carlos - it's all blocks and inlines and scanners that need to be as powerful as the grammar they're lexing for and the walls, they breathe, and they shimmer