Personally, I haven’t found LLMs to be helpful for the internal dialogue. Even with a lot of exposition, code samples, and documentations, it always provides either obvious solutions (store items in a vector!), pointless modifications (use map and filter instead of reduce!), or it just makes up APIs that don’t exist.
I think it’s really good with the 101-level academics side. Learning the basics of anything through a conversational manner can be massively helpful.
As soon as your situation exceeds textbook level, I’ve found them to always be a waste of my time, and nothing I’ve seen as of late makes me think they’re trending in a direction to be helpful in this scenario
> As soon as your situation exceeds textbook level, I’ve found them to always be a waste of my time
Doubly so. A very knowledgeable and helpful tutor would say ”you’re asking for advanced or detailed guidance. I don’t know the specifics, but I can point you to some places that you’d be able to find these answers on your own”.
What the AI does is continue its babbling confidently while being incredibly wrong and non-sensical, like a person suffering a stroke or concussion where the mannerisms are normal but they don’t remember their name. It seems completely unable to judge its own knowledge (probably because it is).
Yes, it would have to be unable to reason about its own confidence levels, wouldn’t it? It produces content—and, as far as I understand, sort of simulates basic reasoning—by making predictions based on a huge corpus of text. The larger this corpus becomes, and the more sophisticated its method of analyzing it, the better it becomes at “reasoning about” the things described in its corpus.
But the question “How sure are you?” inherently refers to something—the LLM’s “mental state”, if such a thing can be said to exist—that isn’t referred to anywhere in the corpus. No improvement in the quality of the corpus or the power of the predictions made based on the corpus can have any impact on this problem.
I want LLMs to answer simple, but tedious questions that arise when I do the thinking. I want them to help me find relevant sections in multi-thousand page datasheets regardless of whether I happened to use the same synonym that documentation's author has used. I want them to remind me the meaning of a term that was defined 12 chapters ago without me having to context switch and look for it again. I want them to consolidate information spread over 6 PDFs that I need to look at to understand something.
I want them to be an interface between me and reliable resources. I want them to essentially facilitate ad-hoc fact retrieval without requiring me to master field-specific jargon first. I don't need them to do any thinking, that's my job. I don't want them to try to answer my questions, I want them to point me to resources that let me answer them and save me the time on searching them in the process. You don't need to know where to look beforehand if you're a machine that can ingest whole libraries in seconds - so let me actually benefit from this power rather than try to provide me with a sketchy equivalent of a clueless intern trying to make a good impression and not realizing how tremendously bad they're at it.
I believe LLMs (or, more specifically, complex systems utilizing LLMs) can end up being incredible productivity boosters, but right now they're being so hopelessly misapplied that it will take a good while for them to get there. LLMs can already be somewhat helpful if you approach them carefully, but they're still far from life-changing - unless your life can be changed by reducing the amount of boilerplate you need to type to code, then I guess you're already happier.
I think it’s really good with the 101-level academics side. Learning the basics of anything through a conversational manner can be massively helpful.
As soon as your situation exceeds textbook level, I’ve found them to always be a waste of my time, and nothing I’ve seen as of late makes me think they’re trending in a direction to be helpful in this scenario