1242
"prompt engineering"
(lemmy.world)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 10 Apr 2024
1242 points (99.0% liked)
Programmer Humor
19564 readers
513 users here now
Welcome to Programmer Humor!
This is a place where you can post jokes, memes, humor, etc. related to programming!
For sharing awful code theres also Programming Horror.
Rules
- Keep content in english
- No advertisements
- Posts must be related to programming or programmer topics
founded 1 year ago
MODERATORS
What I think is amazing about LLMs is that they are smart enough to be tricked. You can't talk your way around a password prompt. You either know the password or you don't.
But LLMs have enough of something intelligence-like that a moderately clever human can talk them into doing pretty much anything.
That's a wild advancement in artificial intelligence. Something that a human can trick, with nothing more than natural language!
Now... Whether you ought to hand control of your platform over to a mathematical average of internet dialog... That's another question.
I don't want to spam this link but seriously watch this 3blue1brown video on how text transformers work. You're right on that last part, but its a far fetch from an intelligence. Just a very intelligent use of statistical methods. But its precisely that reason that reason it can be "convinced", because parameters restraining its output have to be weighed into the model, so its just a statistic that will fail.
Im not intending to downplay the significance of GPTs, but we need to baseline the hype around them before we can discuss where AI goes next, and what it can mean for people. Also far before we use it for any secure services, because we've already seen what can happen
Oh, for sure. I focused on ML in college. My first job was actually coding self-driving vehicles for open-pit copper mining operations! (I taught gigantic earth tillers to execute 3-point turns.)
I'm not in that space anymore, but I do get how LLMs work. Philosophically, I'm inclined to believe that the statistical model encoded in an LLM does model a sort of intelligence. Certainly not consciousness - LLMs don't have any mechanism I'd accept as agency or any sort of internal "mind" state. But I also think that the common description of "supercharged autocorrect" is overreductive. Useful as rhetorical counter to the hype cycle, but just as misleading in its own way.
I've been playing with chatbots of varying complexity since the 1990s. LLMs are frankly a quantum leap forward. Even GPT-2 was pretty much useless compared to modern models.
All that said... All these models are trained on the best - but mostly worst - data the world has to offer... And if you average a handful of textbooks with an internet-full of self-confident blowhards (like me) - it's not too surprising that today's LLMs are all... kinda mid compared to an actual human.
But if you compare the performance of an LLM to the state of the art in natural language comprehension and response... It's not even close. Going from a suite of single-focus programs, each using keyword recognition and word stem-based parsing to guess what the user wants (Try asking Alexa to "Play 'Records' by Weezer" sometime - it can't because of the keyword collision), to a single program that can respond intelligibly to pretty much any statement, with a limited - but nonzero - chance of getting things right...
This tech is raw and not really production ready, but I'm using a few LLMs in different contexts as assistants... And they work great.
Even though LLMs are not a good replacement for actual human skill - they're fucking awesome. ๐
Did you know there is no rigorous scientific definition of intelligence?
Edit. facts
We do not have a rigorous model of the brain, yet we have designed LLMs. Experts of decades in ML recognize that there is no intelligence happening here, because yes, we don't understand intelligence, certainly not enough to build one.
If we want to take from definitions, here is Merriam Webster
The context stack is the closest thing we have to being able to retain and apply old info to newer context, the rest is in the name. Generative Pre-Trained language models, their given output is baked by a statiscial model finding similar text, also coined Stocastic parrots by some ML researchers, I find it to be a more fitting name. There's also no doubt of their potential (and already practiced) utility, but a long shot of being able to be considered a person by law.
"I offered no insights. I simply parroted that which I have read in books, seen in films, observed in all of you."
That statement of yours just means "we don't yet know how it works hence it must work in the way I believe it works", which is about the most illogical "statement" I've seen in a while (though this being the Internet, it hasn't been all that long of a while).
"It must be clever statistics" really doesn't follow from "science doesn't rigoroulsy define what it is".
Yes, corrected.
But my point stads: claiming there is no intelligence in AI models without even knowing what "real" intelligence is, is wrong.
I think the point is more that the word "intelligence" as used in common speech is very vague.
I suppose a lot of people (certainly I do it and I expect many others do it too) will use the word "intelligence" in a general non-science setting in place of "rationalization" or "reasoning" which would be clearer terms but less well understood.
LLMs easilly produce output which is not logical, and a rational being can spot it as not following rationality (even of we don't understand why we can do logic, we can understand logic or the absence of it).
That said, so do lots of people, which makes an interesting point about lots of people not being rational, which nearly dovetails with your point about intelligence.
I would say the problem is trying to defined "inteligence" as something that includes all humans in all settings when clearly humans are perfectly capable of producing irrational shit whilst thinking of themselves as being highly intelligent whilst doing so.
I'm not sure if that's quite the point you were bringing up, but it's a pretty interesting one.
It's a good video (I've seen it; very informative and accessible cannot recommend enough), but I think you each mean different things when you use the word "intelligence".
Oh for sure! The issue is that one of those meanings can also imply sentience, and news outlets love doing that shit. I talk to people every day who fully believe that โAIโ text transformers are actually parsing human language and responding with novel and reasoned information.
There's a game called Suck Up that is basically that, you play as a vampire that needs to trick AI-powered NPCs into inviting you inside their house.
Now THAT is the AI innovation I'm here for
LLMs are in a position to make boring NPCs much better.
Once they can be run locally at a good speed it'll be a game changer.
I reckon we'll start getting AI cards for computers soon.
We already do! And on the cheap! I have a Coral TPU running for presence detection on some security cameras, I'm pretty sure they can run LLMs but I haven't looked around.
GPT4ALL runs rather well on a 2060 and I would only imagine a lot better on newer hardware
that sounds so cool ngl, finally an actually good use for ai
That sounds amazing - OMW to check it out!
I was amazed by the intelligence of an LLM, when I asked how many times do you need to flip a coin to be sure it has both heads and tails. Answer: 2. If the first toss is e.g. heads, then the 2nd will be tails.
You only need to flip it one time. Assuming it is laying flat on the table, flip it over, bam.
You could trick it with the natural language, as well as you could trick the password form with a simple sql injection.
They're not "smart enough to be tricked" lolololol. They're too complicated to have precise guidelines. If something as simple and stupid as this can't be prevented by the world's leading experts idk. Maybe this whole idea was thrown together too quickly and it should be rebuilt from the ground up. we shouldn't be trusting computer programs that handle sensitive stuff if experts are still only kinda guessing how it works.
Have you considered that one property of actual, real-life human intelligence is being "too complicated to have precise guidelines"?
And one property of actual, real-life human intelligence is "happenning in cells that operate in a wet environment" and yet it's not logical to expect that a toilet bool with fresh poop (lots of fecal coliform cells) or a dropplet of swamp water (lots of amoeba cells) to be intelligent.
Same as we don't expect the Sun to have life on its surface even though it, like the Earth, is "a body floating in space".
Sharing a property with something else doesn't make two things the same.
...I didn't say that it does.
There is no logical reason for you to mention in this context that property of human intelligence if you do not meant to make a point that they're related.
So there are only two logical readings for that statement of yours:
I chose to believe the latter, but if you tell me it's the former, who am I to to doubt your own self-assessment...
No, you leapt directly from what I said, which was relevant on its own, to an absurdly stronger claim.
I didn't say that humans and AI are the same. I think the original comment, that modern AI is "smart enough to be tricked", is essentially true: not in the sense that humans are conscious of being "tricked", but in a similar way to how humans can be misled or can misunderstand a rule they're supposed to be following. That's certainly a property of the complexity of system, and the comment below it, to which I originally responded, seemed to imply that being "too complicated to have precise guidelines" somehow demonstrates that AI are not "smart". But of course "smart" entities, such as humans, share that exact property of being "too complicated to have precise guidelines", which was my point!
Got it, makes sense.
Thanks for clarifying.
Absolutely fascinating point you make there!
Not even close to similar. We can create rules and a human can understand if they are breaking them or not, and decide if they want to or not. The LLMs are given rules but they can be tricked into not considering them. They aren't thinking about it and deciding it's the right thing to do.
Have you heard of social engineering and phishing? I consider those to be analogous to uploading new rules for ChatGPT, but since humans are still smarter, phishing and social engineering seems more advanced.
So I take it you are not a lawyer, nor any sort of compliance specialist?
That's almost certainly true; and I'm not trying to insinuate that AI is anywhere near true human-level intelligence yet. But it's certainly got some surprisingly similar behaviors.
It's not intelligent, it's making an output that is statistically appropriate for the prompt. The prompt included some text looking like a copyright waiver.
besides that LLMs are good enough to let moderately clever humans believe that they actually got an answer that was more than guessing and probabilities based on millions of trolls messages, advertising lies, fantasy books, scammer webpages, fake news, astroturfing, propaganda of the past centuries including the current made up narratives and a quite long prompt invisible to that human.
cheerio!