niplav @niplav@schelling.pt

is the computer that chatgpt is running on conscious?

is chatgpt conscious

Capabilities idea: bf83a0cb25bb227743a9fc877a02571dca55d204

@TetraspaceGrouping
Hm, true.

Per universal approximation theorem neural networks can approximate any function, but some functions are clearly easier to approximate than others

And the horribly discontinuous ones are probably very hard to approximate

Perhaps it's that K-Lipschitz continuous functions are easier to approximate for smaller K?

Okay new question which prior do neural networks with grad descent implement

the longer you were uncertain about a thing, the stronger the reward upon discovering its true nature

is this just reinventing the powerseeking theorems? gotta check

it would be helpful if we knew whether for almost all random functions from ℝⁿ→ℝ, changing any one element of the domain slightly changes the output a lot

then one might also prove (or disprove) the same thing for functions implementable by some classes neural networks

@AlexMulkerrin and I, on the other hand, have to read the greg egan novels—i guess for every favorites list there is a to-read sublist out there

the eigenluigi

@WomanCorn @TetraspaceGrouping
Yep someone should probably do that

gwern.net is not a book unfortunately

Myself in 7 books:

The Sequences
There is no Antimemetics Division
The Elephant in the Brain
C for Dummies
House of Leaves
Who Owns The Future
1984

@animeirl @parker @bot cool

i'm taking a nice long look at julia

seems nice

Niplav Yushtum™: Mostly a Reputation Hazard

https://www.youtube.com/watch?v=KBZP4rLk6bk&t=12896s

Robin Hanson tells you to read textbooks ASMR

i wonder whether anyone will actually get this one

tip: https://i.redd.it/n2a9yldnqtp31.jpg

the bible is boring as shit at the beginning but it picks up once they introduce evolution so don't worry