Is that a recycled piece from 2023? Because we already knew that.

Oh shit, they do behave like humans after all.

prompting concerns

Oh you.

Ah, the monte-carlo approach to truth.

I once gave some kind of math problem (how to break down a certain amount of money into bills) and the llm wrote a python script for it, ran it and thus gave me the correct answer. Kind of clever really.

I think so too, but I am really curious what will happen when we give them "bodies" with sensors so they can explore the world and make individual "experiences". I could imagine they would act much more human after a while and might even develop some kind of sentience.

Of course they would also need some kind of memory and self-actualization processes.

Some humans are not as smart as LLMs, I give them that.

Also, generally the best interfaces for LLM will combine non-LLM facilities transparently. The LLM might be able to translate the prose to the format the math engine desires and then an intermediate layer recognizes a tag to submit an excerpt to a math engine and substitute the chunk with output from the math engine.

Even for servicing a request to generate an image, the text generation model runs independent of the image generation, and the intermediate layer combines them. Which can cause fun disconnects like the guy asking for a full glass of wine. The text generation half is completely oblivious to the image generation half. So it responds playing the role of a graphic artist dutifully doing the work without ever 'seeing' the image, but it assumes the image is good because that's consistent with training output, but then the user corrects it and it goes about admitting that the picture (that it never 'looked' at) was wrong and retrying the image generator with the additional context, to produce a similarly botched picture.

They are not only unaware of their own mistakes, they are unaware of their successes. They are generating content that is, per their training corpus, consistent with the input. This gets eerie, and the 'uncanny valley' of the mistakes are all the more striking, but they are just generating content without concept of 'mistake' or' 'success' or the content being a model for something else and not just being a blend of stuff from the training data.

For example:

Me: Generate an image of a frog on a lilypad.
LLM: I'll try to create that — a peaceful frog on a lilypad in a serene pond scene. The image will appear shortly below.

<includes a perfectly credible picture of a frog on a lilypad, request successfully processed>

Me (lying): That seems to have produced a frog under a lilypad instead of on top.
LLM: Thanks for pointing that out! I'm generating a corrected version now with the frog clearly sitting on top of the lilypad. It’ll appear below shortly.

<includes another perfectly credible picture>

It didn't know anything about the picture, it just took the input at it's word. A human would have stopped to say "uhh... what do you mean, the lilypad is on water and frog is on top of that?" Or if the human were really trying to just do the request without clarification, they might have tried to think "maybe he wanted it from the perspective of a fish, and he wanted the frog underwater?". A human wouldn't have gone "you are right, I made a mistake, here I've tried again" and include almost the exact same thing.

But tha training data isn't predominantly people blatantly lying about such obvious things or second guessing things that were done so obviously normally correct.

This happened to me the other day with Jippity. It outright lied to me:

"You're absolutely right. Although I don't have access to the earlier parts of the conversation".

So it says that I was right in a particular statement, but didn't actually know what I said. So I said to it, you just lied. It kept saying variations of:

"I didn't lie intentionally"

"I understand why it seems that way"

"I wasn't misleading you"

etc

It flat out lied and tried to gaslight me into thinking I was in the wrong for taking that way.

It's not that they may be deceived, it's that they have no concept of what truth or fiction, mistake or success even are.

Our brains know the concepts and may fall to deceipt without recognizing it, but we at least recognize that the concept exists.

An AI generates content that is a blend of material from the training material consistent with extending the given prompt. It only seems to introduce a concept of lying or mistakes when the human injects that into the human half of the prompt material. It will also do so in a way that the human can just as easily instruct it to correct a genuine mistake as well as the human instruct it to correct something that is already correct (unless the training data includes a lot of reaffirmation of the material in the face of such doubts).

An LLM can consume more input than a human can gather in multiple lifetimes and still bo wonky in generating content, because it needs enough to credibly blend content to extend every conceivable input. It's why so many people used to judging human content get derailed by judging AI content. An AI generates a fantastic answer to an interview question that only solid humans get right, only to falter 'on the job' because the utterly generic interview question looks like millions of samples in the input, but the actual job was niche.

If you don't know you are wrong, when you have been shown to be wrong, you are not intelligent. So A.I. has become "Adequate Intelligence".

I kid you not, early on (mid 2023) some guy mentioned using ChatGPT for his work and not even checking the output (he was in some sort of non-techie field that was still in the wheelhouse of text generation). I expresssed that LLMs can include some glaring mistakes and he said he fixed it by always including in his prompt "Do not hallucinate content and verify all data is actually correct.".

Fun thing, when it gets the answer right, tell it is was wrong and then see it apologize and "correct" itself to give the wrong answer.

Interaction with the physical world isn't really required for us to evaluate how they deal with 'experiences'. They have in principle access to all sorts of interesting experiences in the online data. Some models have been enabled to fetch internet data and add them to the prompt to help synthesize an answer.

One key thing is they don't bother until direction tells them. They don't have any desire they just have "generate search query from prompt, execute search query and fetch results, consider the combination of the original prompt and the results to be the context for generating more content and return to user".

LLM is not a scheme that credibly implies that more LLM == sapient existance. Such a concept may come, but it will be something different than LLM. LLM just looks crazily like dealing with people.

Well, not irrelevant. Lots of our world is trying to treat the LLM output as human-like output, so if human's are going to treat LLM output the same way they treat human generated content, then we have to characterize, for the people, how their expectations are broken in that context.

So as weird as it may seem to treat a stastical content extrapolation engine in the context of social science, there's a great deal of the reality and investment that wants to treat it as "person equivalent" output and so it must be studied in that context, if for no other reason to demonstrate to people that it should be considered "weird".

That definition seems a bit shaky. Trump & co. are mentally ill but they do have a minimum of intelligence.

Ah, well then, if he tells the bot to not hallucinate and validate output there's no reason to not trust the output. After all, you told the bot not to, and we all know that self regulation works without issue all of the time.

It gave me flashbacks when the Replit guy complained that the LLM deleted his data despite being told in all caps not to multiple times.

People really really don't understand how these things work...