23
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 22 Mar 2024
23 points (100.0% liked)
SneerClub
982 readers
9 users here now
Hurling ordure at the TREACLES, especially those closely related to LessWrong.
AI-Industrial-Complex grift is fine as long as it sufficiently relates to the AI doom from the TREACLES. (Though TechTakes may be more suitable.)
This is sneer club, not debate club. Unless it's amusing debate.
[Especially don't debate the race scientists, if any sneak in - we ban and delete them as unsuitable for the server.]
founded 1 year ago
MODERATORS
Such a simple way to give away that you know absolutely fuck-all about actual Machine Learning as a science. Neural networks (ML term of art) don't fucking expand with data, they have a fixed size. A neural network is just a spicy matrix with real numbers.
This also seems to suggest the author thinks "bigger = better", which is just false. The whole field would probably be much easier if it were true, but you learn this im ML 101, just making the matrix bigger can reduce accuracy.
Neural networks are such an unfortunate name. They kinda look like a layered network and propagation-backpropagation kinda might give you "neurons firing" vibes, so scientists named it like that; only for dipshit pundits to immediately go "neuron means brain means machine thinks!!!" years later...
Yeah, I work in a field where machine learning has incredible applications, but also we're facing a deluge of trash. It's been clear for a while now that smaller, domain specific models are where it's at