Rich Felker
Yesterday I encountered a "wrong-on-the-internet" rando professing his excitement for "using machine learning" in #3dprinting to throttle speeds in the right places to avoid quality loss.
While completely not worth engaging with, I feel like this is a useful example to understand why this idiocy is so infuriating...
Raven Onthill
@dalias they want badly to believe that minds can be reduced to simple stochastic models. This wasn't an entirely unreasonable hypothesis 20 years ago, but at this point it doesn't look like it's correct.
Rich Felker
@ravenonthill The concept is still vaguely plausible, but their idea for how to achieve it is utter bullshit.
If you compare how human minds are "trained", there are multiple feedback layers in the form of consequences, and most importantly, we select very carefully what training inputs are used rather than throwing giant mostly wrong and mostly evil corpuses at children, and most of the training is experiential not ingesting word soup.
Rich Felker
The proponents of this kind of shit want to throw away the whole concept of having and using scientific knowledge obtained by experiment, with documentation of how it was obtained, evidence supporting the resulting models, falsifiability, etc., and replace it with a worse version of the way humans tens of thousands of years ago came to believe things about the world: simplistic pattern recognition.
Rich Felker
This is a problem domain where the constraints and effects are pretty much entirely comprehensible in terms of known physical models. Any suboptimal behavior is entirely a matter of nobody having spent the time to apply known models. But sure, let's instead spend the time hooking up ML, CV to evaluate results, and waste tons (literally) of plastic training a model to learn a poor approximation of what we already know.
But this is a general pattern that's terrifying...
Chris Gioran 💔
@dalias I think you are attributing way too much agency to them.
It's not high aspirations for AI's capabilities. It's fundamental disinterest in the subject matter that's the problem.
My bet is that they thought - gee, this looks like a hard problem. I bet that no one has thought about it before, so instead of me investing the effort to understand it, I'll just have AI solve it for me.
Rich Felker
@chrisg That's kinda the whole thing, the anti-expertise sentiment behind it. 🤬
Rich Felker
@futurebird It's not that it couldn't work, but that we understand the physical reasons when/why going too fast can give poor results, and can just apply that rather than trying to coax a model into figuring out the same things without underlying models of thermal transfer, etc. It's really low hanging fruit.
myrmepropagandist
@dalias Can you tell us more about why machine learning isn’t a good fit for this? Though I didn’t even know throttling speeds could possibly improve quality. Knowing little it sounds possible? But what tipped you off that it would not work?
AK
@futurebird @dalias most of the time applying ML to physical models will produce an over-linearized black-box solution compared to using known equations to solve, simulate, and optimize whatever you’re trying to do. In situations where simulation is resource intensive to do at full fidelity we have tools like domain-specific reduced -order models.
hypolite
@futurebird @dalias I feel like it isn’t a good fit for the same reason it isn’t a good fit to make a machine learning calculator. Sure, it can work, but at what cost and reliability compared to what we can already make using existing knowledge?
Unfortunately it doesn’t stop people to try to make worse versions of Wolfram Alpha using ML or LLM systems. 😞
clacke@libranet.de is my main
@hypolite Oh, I like this explanation. It's a bit like how its possible to solve many applied mathematical problems with brute force, an expensive and inexact numerical solution, but if you know calculus you can create an exact symbolic solution, a nice formula instead of an iterative algorithm.
David Brin suggested in his Uplift books that in an advanced and ancient galactic civilization, computers would be far too powerful, everything would have already been invented and people would be relying on ancient libraries instead of researching anything. This would make them fall prey to using numerical solutions to everything, which could give an advantage to a scrappy upstart civilization with a culture that insisted on understanding what you're doing, and using calculus.
It seemed unrealistic to me, but it looks like we're already seeing that decline before we even reached the stars.
@futurebird @dalias
G. Wozniak
@steve @dalias Based on my exposure to all this AI stuff, I find those most enthusiastic are those who want to make fast gains in a field they know nothing about. It's not about it getting better, it's about that (perceived) immediate gain.
Steve Canon
@dalias at least in part, this is testament to our collective failure to make actual engineering accessible to normal people. I can believe that for a lot of people a usually-good-enough ML solution is easier, even though we know it’s profoundly stupid and wasteful.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.