niplav @niplav@schelling.pt

longe+loud=lounge

*hugs bison*

For me it's

s25::{x@<(#x){x,.rn()}:*[]}

It's a function that shuffles a list

the answer to all questions is either: “Can you feel your toes right now?”, “Start a long site” or “Talk to more girls”.

if there's one field ripe for huge progress by a large swath of theoretical physicists invading it, it's probably educational research.

looking from afar, the quality of science there looks quite abysmal, while the field is very important (finding out how to make humans learn things better! come on!)

but maybe i'm underestimating the research

humans have invented mathematics surprisingly early (while the status of the number zero was still being debated in ancient greece), and programming surprisingly late (although algorithms were around for a long time, they didn't catch on until later, even though they seem like an at least equally intuitive concept?)

who has taken the train to crazytown the furthest? (and hasn't suffered from nervous breakdown in the process)

to diversify its lingo, the rationality community should adopt the evidentials smṛti (“I read this in a blogpost”) and śruti (“someone told this to me in a one-on-one”).

pseudoquasisemimetric

if we solve distributional shift, do we also solve ontological crises?

imagine the teeth-grinding if they gave schmidhuber a turing award

listening to Gary Marcus right now on what would really impress him with AI advances, and, well,,,,,,,,,,

im gonna turn into schmidhuber if im not careful

maybe my website wants to become a collection of jupyter notebooks?

become unlinearizable