Do you have a dumb question that you're kind of embarrassed to ask in the main thread? Is there something you're just not sure about?
This is your opportunity to ask questions. No question too simple or too silly.
Culture war topics are accepted, and proposals for a better intro post are appreciated.
Jump in the discussion.
No email address required.
Notes -
How do you all interact with LLMs?
I’ve seen a few articles recently noting the rise of AI as a buddy / therapist or whatever. It’s usually beside the point of the article but an implicit notion is that lot of folks regularly ‘chat’ with AI as if it were a person.
Which I find baffling. Outside of the very early novelty, I find this way of interacting extremely boring and tedious, and generally find the fact that AI wants to get conversational with me a general frustrater.
If I’m not using AI as a utility ‘write X, troubleshoot Y, give me steps for Z’, and I’m using it recreationally / casually, it’s more akin to web surfing or browsing Wikipedia than chatting on a forum or whatever. I will use it as an open format encyclopedia and explicitly not as a conversationalist sounding board. And i genuinely find negative value in the fact that the former is constantly interrupted with the attempt to be the latter.
So my question is again, how far outside of the grain am I?
I generally use it as a search engine that I can ask more specific questions to than I can ask google.
I think its pretty helpful with travel planning, I feel like it lets me dictate more degrees of freedom than google does.
I'm taking a trip with my daughter next month, my daughter wants to go ziplining, "Can you help me find ziplining places, we're starting at A, ending at B, anywhere roughly along the route ..." I find LLMs handle that sort of thing better than google does. "We going to be in X for 2 days, what's some things we should do?" "Ehh, I don't think we would like that, what else" "Ehh, how expensive is that, is there something more affordable?". idk, couple iterations of that get you to something pretty workable.
This is the way.
You can't use LLMs for anything that you can't check yourself afterward; the hallucination rates are still just too high. But they're fantastic for cases where you'd like to use a search engine but there's just no way to turn your query into a list of words that (along with obvious synonyms) would define and sort the results.
"Tourist attractions in X" will get you to a TripAdvisor page that's fine. For "but not too expensive" you might be better off with the LLM than you would be manually searching a curated list. For "near a road trip route in between X and Y" and "oh, but we'd prefer to take a more northerly and high-altitude road in the summer heat" there was just no beating the LLM (actual example from my last vacation). It took surprisingly few queries like "here's what I vaguely remember about a fun trip with my parents in this state decades ago" to get to an answer like "here's the specific canyon and creek-side picnic site they probably took you to" (which, based on how familiar the drive felt when I took my own kids there, was probably correct).
You'd think that only works for fuzzy answers like vacation planning suggestions, but it's pretty good even for well-defined answers to fuzzily-stated questions. I'd never trust an LLM alone to tell me what Godunov's Theorem is or means, but when I couldn't remember the name "Godunov's Theorem" it was by giving a vague description to Claude that I found it.
I found that it is very good at telling me which book I am trying to remember from a few hazy recollection of what the book was about.
I was trying to remember which book about basketball stats I read about 15 years ago which had a chapter comparing the relative merits of Tyson Chandler and Eddie Curry. Google gives you a sea of links to those guys wikipedia and basketball reference pages.
Chat GPT immediately knew it was 'The Wages of Wins' by David Berri, had a command of the basic thesis of the book in a way that jived with my memory of the book, was able to contrast and compare the arguments from the book with other books on sports statistics, talk about the various assumptions the arguments from those books relied on.
I was honestly pretty blown away at how useful it was in contrast to google searching.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link