this post was submitted on 04 Oct 2023
22 points (86.7% liked)
Asklemmy
43852 readers
699 users here now
A loosely moderated place to ask open-ended questions
Search asklemmy ๐
If your post meets the following criteria, it's welcome here!
- Open-ended question
- Not offensive: at this point, we do not have the bandwidth to moderate overtly political discussions. Assume best intent and be excellent to each other.
- Not regarding using or support for Lemmy: context, see the list of support communities and tools for finding communities below
- Not ad nauseam inducing: please make sure it is a question that would be new to most members
- An actual topic of discussion
Looking for support?
Looking for a community?
- Lemmyverse: community search
- sub.rehab: maps old subreddits to fediverse options, marks official as such
- [email protected]: a community for finding communities
~Icon~ ~by~ ~@Double_[email protected]~
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
These models aren't great at tasks that require precision and analytical thinking. They're trained on a fairly simple task, "if I give you some text, guess what the next bit of text is." Sounds simple, but it's incredibly powerful. Imagine if you could correctly guess the next bit of text for the sentence "The answer to the ultimate question of life, the universe, and everything is" or "The solution to the problems in the Middle East is".
Recently, we've been seeing shockingly good results from models that do this task. They can synthesize unrelated subjects, and hold coherent conversations that sound very human. However, despite doing some things that up until recently only humans could do, they still aren't at human-level intelligence. Humans read and write by taking in words, converting them into rich mental concepts, applying thoughts, feelings, and reasoning to them, and then converting the resulting concepts back into words to communicate with others. LLMs arguably might be doing some of this too, but they're evaluated solely on words and therefore much more of their "thought process" is based on "what words are likely to come next" and not "is this concept being applied correctly" or "is this factual information". Humans have much, much greater capacity than these models, and we live complex lives that act as an incredibly comprehensive training process. These models are small and trained very narrowly in comparison. Their excellent mimicry gives the illusion of a similarly rich inner life, but it's mostly imitation.
All that comes down to the fact that these models aren't great at complex reasoning and precise details. They're just not trained for it. They got through "life" by picking plausible words and that's mostly what they'll continue to do. For writing a novel or poem, that's good enough, but math and physics are more rigorous than that. They do seem to be able to handle code snippets now, mostly, which is progress, but in general this isn't something that you can be completely confident in them doing correctly. They make silly mistakes because they aren't really thinking it through. To them, there isn't really much difference between answers like "that date is 7 days after Christmas" and "that date is 12 days after Christmas." Which one it thinks is more correct is based on things it has seen, not necessarily an explicit counting process. You can also see this in things like that case where someone tried to use it to write a legal brief, where it came up with citations that seemed plausible but were in fact completely made up. It wasn't trained on accurate citations, it was trained on words.
They also have a bad habit of sounding confident no matter what they're saying, which makes it hard to use them for things you can't check yourself. Anything they say could be right/accurate/good/not plagiarized, but the model won't have a good sense of that, and if you don't know either, you're opening yourself up to risk of being misled.
They can definitely be made to work out arithmetic and similar though
If you were to say in the preprompt something like: When asked a mathematical question, please respond with the equations used to achieve the result
For example if you asked it what 3x4 is it could respond with "The answer is {3x4}" and then the {3x4} could be evaluated in software afterwards and dropped in for the user to see
I think that might be what chatGPT does now as they somewhat recently fixed it always getting maths wrong
Or alternatively you could ask it to simply write a script to work out whatever problem it's given that isn't linguistic and execute that in a sandboxed environment (though still might be too risky incase it generates some bad code)