“Almost nothing” is not the same as “actually useless”. The former is saying the applications are limited, which is true.
LLMs are fine for fictional interactions, as in things that appear to be real but aren’t. They suck at anything that involves being reliably factual, which is most things including all the stupid places LLMs and other AI are being jammed in to despite being consistely wrong, which tech bros love to call hallucinations.
They have LIMITED applications, but are being implemented as useful for everything.
To be honest, as someone who’s very interested in computer generated text and poetry and the like, I find generic LLMs far less interesting than more traditional markov chains because they’re too good at reproducing clichés at the exclusion of anything surprising or whimsical. So I don’t think they’re very good for the unfactual either. Probably a homegrown neural network would have better results.
GPT-2 was peak LLM because it was bad enough to be interesting, it was all downhill from there
Agreed, our chat server ran a Markov chain bot for fun.
In comparison to ChatGPT on a 2nd server I frequent it had much funnier and random responses.
ChatGPT tends to just agree with whatever it chose to respond to.
As for real world use. ChatGPT 90% of the time produces the wrong answer. I’ve enjoyed Circuit AI however. While it also produces incorrect responses, it shares its sources so I can more easily get the right answer.
All I really want from a chatbot is a gremlin that finds the hard things to Google on my behalf.
I’m in the same boat. Markov chains are a lot of fun, but LLMs are way too formulaic. It’s one of those things where AI bros will go, “Look, it’s so good at poetry!!” but they have no taste and can’t even tell that it sucks; LLMs just generate ABAB poems and getting anything else is like pulling teeth. It’s a little more garbled and broken, but the output from a MCG is a lot more interesting in my experience. Interesting content that’s a little rough around the edges always wins over smooth, featureless AI slop in my book.
slight tangent: I was interested in seeing how they’d work for open-ended text adventures a few years ago (back around GPT2 and when AI Dungeon was launched), but the mystique did not last very long. Their output is awfully formulaic, and that has not changed at all in the years since. (of course, the tech optimist-goodthink way of thinking about this is “small LLMs are really good at creative writing for their size!”)
I don’t think most people can even tell the difference between a lot of these models. There was a snake oil LLM (more snake oil than usual) called Reflection 70b, and people could not tell it was a placebo. They thought it was higher quality and invented reasons why that had to be true.
Like other comments, I was also initially surprised. But I think the gains are both real and easy to understand where the improvements are coming from. [ . . . ]
I had a similar idea, interesting to see that it actually works. [ . . . ]
I think that’s cool, if you use a regular system prompt it behaves like regular llama-70b. (??!!!)
It’s the first time I’ve used a local model and did [not] just say wow this is neat, or that was impressive, but rather, wow, this is finally good enough for business settings (at least for my needs). I’m very excited to keep pushing on it. Llama 3.1 failed miserably, as did any other model I tried.
For story telling or creative writing, I would rather have the more interesting broken english output of a Markov chain generator, or maybe a tarot deck or D100 table. Markov chains are also genuinely great for random name generators. I’ve actually laughed at Markov chains before with friends when we throw a group chat into one and see what comes out. I can’t imagine ever getting something like that from an LLM.