this post was submitted on 07 Sep 2025
162 points (87.2% liked)
Bluesky
1631 readers
18 users here now
People skeeting stuff.
Bluesky Social is a microblogging social platform being developed in conjunction with the decentralized AT Protocol. Previously invite-only, the flagship Beta app went public in February 2024. All are welcome!
founded 9 months ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Medicine discovery is one of the few fields with ai where the juice is worth the squeeze. Filtering through billions of different molecule geometries to compare them all to eachother is a task that ai does well and humans do extremely slowly
The issue is the overloading of the word “AI”.
“Machine Learning using Neural Networks” is a technique that can come up with decent but rough solutions to problems where it’s hard to come up with any solution.
“Large Language Models” is the application of “Machine Learning using Neural Networks” to natural language processing, and it is incredibly good at that.
The problem comes when people apply models trained for natural language processing onto other random problems just because you can formulate anything as a natural language problem.
That's a fair distinction. That being said, at their core llms are just big functions. You could cover a dartboard in subfields of physics, toss a dart randomly, and I'd bet money you hit a field that finds use for the bessel functions for instance. I am not informed enough on the specifics of llms to say either way, but there's definitely precedent for "we found this really powerful function and it turns out it accurately predicts 10 shitloads of unrelated systems."
To double down on my devils advocacy, the projects I have personally seen or been consulted for that fit the form "use llm to solve non-nlp problem" are 99% propelled by "funding for Ai buzzwords flows freely" and "understanding of the limitations of different kinds of ai is rare"
I think the lesson in that is that these things are tools that experts who are actually focused on checking the outputs can use to great benefit.
The average schmuck trying to write vibe code way beyond their understanding is the one fooling themselves about their utility.
When a dork at work gives me a proposal that is clearly hot LLM garbage they hardly read? I make sure they know they are still responsible for producing shitty work that needs to be re-done.
Yeah. That's the most frustrating part of it to me. Statistical learning is really limited in the applications it's good at, but it excels wildly where it excels. Its a specialized scalpel, but the people who are the public face call it "ai" and market it as a cure all for the pesky problem of having to pay workers