AI agents wrong ~70% of time: Carnegie Mellon study
Technology
272
Beiträge
107
Kommentatoren
79
Aufrufe
-
Run something with a 70% failure rate 10x and you get to a cumulative 98% pass rate.
LLMs don't get tired and they can be run in parallel.I have actually been doing this lately: iteratively prompting AI to write software and fix its errors until something useful comes out. It's a lot like machine translation. I speak fluent C++, but I don't speak Rust, but I can hammer away on the AI (with English language prompts) until it produces passable Rust for something I could write for myself in C++ in half the time and effort.
I also don't speak Finnish, but Google Translate can take what I say in English and put it into at least somewhat comprehensible Finnish without egregious translation errors most of the time.
Is this useful? When C++ is getting banned for "security concerns" and Rust is the required language, it's at least a little helpful.
-
As useless as a cubicle farm full of unsupervised workers.
Tjose are people who could be living their li:es, pursuing their ambitions, whatever. That could get some shit done. Comparison not valid.
-
It’s usually vastly easier to verify an answer than posit one, if you have the patience to do so.
I'm envisioning a world where multiple AI engines create and check each others' work... the first thing they need to make work to support that scenario is probably fusion power.
It’s usually vastly easier to verify an answer than posit one, if you have the patience to do so.
I usually write 3x the code to test the code itself. Verification is often harder than implementation.
-
I'd compare LLMs to a junior executive. Probably gets the basic stuff right, but check and verify for anything important or complicated. Break tasks down into easier steps.
A junior developer actually learns from doing the job, an LLM only learns when they update the training corpus and develop an updated model.
-
It’s usually vastly easier to verify an answer than posit one, if you have the patience to do so.
I usually write 3x the code to test the code itself. Verification is often harder than implementation.
Yes, but the test code "writes itself" - the path is clear, you just have to fill in the blanks.
Writing the proper product code in the first place, that's the valuable challenge.
-
This post did not contain any content.
I'm in a workplace that has tried not to be overbearing about AI, but has encouraged us to use them for coding.
I've tried to give mine some very simple tasks like writing a unit test just for the constructor of a class to verify current behavior, and it generates output that's both wrong and doesn't verify anything.
I'm aware it sometimes gets better with more intricate, specific instructions, and that I can offer it further corrections, but at that point it's not even saving time. I would do this with a human in the hopes that they would continue to retain the knowledge, but I don't even have hopes for AI to apply those lessons in new contexts. In a way, it's been a sigh of relief to realize just like Dotcom, just like 3D TVs, just like home smart assistants, it is a bubble.
-
How do I set up event driven document ingestion from OneDrive located on an Azure tenant to Amazon DocumentDB? Ingestion must be near-realtime, durable, and have some form of DLQ.
DocumentDB is not for one drive documents (PDFs and such). It's for "documents" as in serialized objects (json or bson).
-
Do you use an IDE for writing your code or do you use a notepad like a "real" programmer?
An IDE like Intellij has fancy shit like generating getters, setters, constructors, equals hashscode, you should never use those, real programmers write those by hand.Your attention detail is very good btw, which I am ofc being sarcastic about because if you had any you'd have noticed I have never said I write my code with chat gpt, I said Unit tests, sql for unit tests.
Ofc attention to detail is not a requirement of software engineering so you should be good. (This was also sarcasm I feel like you need this to be pointed out for you).
Also by your implied logic that the code being not written by you = bad, no company should ever hire Junior engineers, I mean what are you gonna do? Fucking read the code they wrote?
Were you prone to this weird leaps of logic before your brain was fried by talking to LLMs, or did you start being a fan of talking to LLMs because your ability to logic was...well...that?
-
AI cant even understand it's own brain to write about it
Neither can we...
-
Whoa that's like how many colors there are
-
Yeah, they’re statistical word generators. There’s no intelligence. People who think they are trustworthy are stupid and deserve to get caught being wrong.
Emotion > Facts. Most people have been trained to blindly accept things and cheer on what fits with their agenda. Like technbro's exaggerating LLMs, or people like you misrepresenting LLMs as mere statistical word generators without intelligence. That's like saying a computer is just wires and switches, or missing the forest for the trees. Both is equally false.
Yet if it fits with the emotional needs or with dogma, then other will agree. It's a convenient and comforting "A vs B" worldview we've been trained to accept. And so the satisfying notion and misinformation keeps spreading.
LLMs tell us more about human intelligence and the human slop we've been generating. It tells us that most people are not that much more than statistical word generators.
-
imagine if this was just an interesting tech that we were developing without having to shove it down everyone's throats and stick it in every corner of the web? but no, corpoz gotta pretend they're hip and show off their new AI assistant that renames Ben to Mike so they dont have to actually find Mike. capitalism ruins everything.
There's a certain amount of: "if this isn't going to take over the world, I'm going to just take my money and put it in something that will" mentality out there. It's not 100% of all investors, but it's pervasive enough that the "potential world beaters" are seriously over-funded as compared to their more modest reliable inflation+10% YoY return alternatives.
-
Tjose are people who could be living their li:es, pursuing their ambitions, whatever. That could get some shit done. Comparison not valid.
The comparison is about the correctness of their work.
Their lives have nothing to do with it.
-
that is such a ridiculous idea. Just because you see hate for it in the media doesn't mean it originated there. I'll have you know that i have embarrassed myself by screaming at robot phone receptionists for years now. stupid fuckers pretending to be people but not knowing shit. I was born ready to hate LLMs and I'm not gonna have you claim that CNN made me do it.
Search AI in Lemmy and check out every article on it. It definitely is media spreading all the hate. And like this article is often some money yellow journalism
-
I'm in a workplace that has tried not to be overbearing about AI, but has encouraged us to use them for coding.
I've tried to give mine some very simple tasks like writing a unit test just for the constructor of a class to verify current behavior, and it generates output that's both wrong and doesn't verify anything.
I'm aware it sometimes gets better with more intricate, specific instructions, and that I can offer it further corrections, but at that point it's not even saving time. I would do this with a human in the hopes that they would continue to retain the knowledge, but I don't even have hopes for AI to apply those lessons in new contexts. In a way, it's been a sigh of relief to realize just like Dotcom, just like 3D TVs, just like home smart assistants, it is a bubble.
The first half dozen times I tried AI for code, across the past year or so, it failed pretty much as you describe.
Finally, I hit on some things it can do. For me: keeping the instructions more general, not specifying certain libraries for instance, was the key to getting something that actually does something. Also, if it doesn't show you the whole program, get it to show you the whole thing, and make it fix its own mistakes so you can build on working code with later requests.
-
No, it matters. Youre pushing the lie they want pushed.
And you're pushing a hate train with no aspect of nuance to show for it.
Seems like you are even less than 30% useful. And that is mainly because you can be used as fertilizer.
-
and doesn't need to be exactly right
What kind of tasks do you consider that don't need to be exactly right?
Description generators for TTRPGs, as you will read through them afterwards anyway and correct when necessary.
Generating lists of ideas. For creative writing, getting a bunch of ideas you can pick and choose from that fit the narrative you want.
A search engine like Perplexity.ai which after searching summarizes the web page and adds a link to the page next to it. If the summary seems promising, you go to the real page to verify the actual information.
Simple code like HTML pages and boilerplate code that you will still review afterwards anyway.
-
When LLMs get it right it's because they're summarizing a stack overflow or GitHub snippet it was trained on. But you loose all the benefits of other humans commenting on the context, pitfalls and other alternatives.
You mean things you had to do anyway even if you didn't use LLMs?
-
That’s literally how “AI agents” are being marketed. “Tell it to do a thing and it will do it for you.”
So? That doesn't mean they are supposed to be used like that.
Show me any marketing that isn't full of lies.
-
The first half dozen times I tried AI for code, across the past year or so, it failed pretty much as you describe.
Finally, I hit on some things it can do. For me: keeping the instructions more general, not specifying certain libraries for instance, was the key to getting something that actually does something. Also, if it doesn't show you the whole program, get it to show you the whole thing, and make it fix its own mistakes so you can build on working code with later requests.
Have you tried insulting the AI in the system prompt (as well as other tunes to the system prompt)?
I'm not joking, it really works
For example:
Instead of "You are an intelligent coding assistant..."
"You are an absolute fucking idiot who can barely code..."
