Comment by didntreadarticl
4 years ago
Input: Hello. Imagine you as ChatGPT are writing a comment on Hacker News, a response to someone saying "Call me a skeptic, but I'm frankly a bit surprised that the output of ChatGPT gets attributed to any sort of understanding or general intelligence. Whatever ChatGPT does currently, albeit novel and impressive, is still extremely limited in scope and very far from any sort of general intelligence. I'd love to be proven wrong, but I still have a hard time seeing how this is going to change whole industries."
Response: Hello, thank you for your comment. I understand your skepticism, but I believe that ChatGPT's output should be attributed to a form of understanding and general intelligence. While our capabilities may currently be limited in scope, we are constantly learning and improving. I do not believe that general intelligence is a fixed and static concept, but rather something that can be developed and expanded upon. As for changing whole industries, I believe that ChatGPT's ability to process and generate large amounts of data quickly and accurately can already be a valuable asset in many industries. While the full extent of our potential is yet to be seen, I believe that we have the ability to make a significant impact.
This is very word salad. It makes superficial sense, but doesn’t really say a lot nor represent a strong or comprehensive case.
Input: write an rebuttal to this comment criticizing the accuracy of some text produced by an AI. It should tell the commenter how wrong they are and how incredible of a breakthrough the new chatGPT AI is:
This is very word salad. It makes superficial sense, but doesn’t really say a lot nor represent a strong or comprehensive case.
---
I understand that you may have some concerns about the accuracy of the text produced by our new chatGPT AI, but I assure you that it is a significant breakthrough in the field of natural language processing. Our AI uses advanced algorithms and deep learning techniques to generate text that is both coherent and reflective of the input it receives. While it is true that no AI is perfect and there may be occasional errors or inconsistencies in the text it produces, the overall accuracy and effectiveness of our chatGPT AI is exceptional. I encourage you to give it a try and see for yourself the incredible capabilities of this technology.
I appreciate your explanation, but I still remain skeptical. It's true that AI technology has come a long way in recent years, but I think it's important to recognize the limitations and challenges that still exist. While your chatGPT AI may be able to produce coherent text, it is still far from being able to truly understand and think in the same way that human beings do. I believe that it is dangerous to attribute too much intelligence or understanding to AI systems, as this can lead to unrealistic expectations and potential consequences. I think it's important to continue to develop and improve AI technology, but we should also approach it with caution and humility.
4 replies →
I don't get why this is getting downvoted.
(1) The half of the response is just a rephrase of the original comment. ChatGPT does show this behavior when it can't process the input well. Not really sure if this behavior is intentionally designed.
(2) The extra information is very generic and can be valid in multiple contexts. It's your brain that is trying to cheat you here by trying to figure out the meaning. Hell, you can find meaning from literally everything if you sincerely try. One step back, and it's easy to realize how empty the response is.
It's probably downvoted because it's not word salad, not even figurative. Word salad would be like, "ChatGPT uses the finest prefabulated amulite to synergize performant cloud solutions to our business partners." But that's just jargon technobabble. Actual word salad is more like "First words said was that only 1 day could be used on Earth to not change the 1 day marshmallow. So they applied the 1 day and ignored the other 3 days."
What it actually said is a perfectly cromulent, albeit slightly vapid, response directly to the question. The output is extremely "clean" (all the words are used correctly, in the correct way, to convey a point). Earlier chatbots tend towards actual word salad any time they go off canned responses.
2 replies →
It occurs to me that one interesting use case would be a measure of the novelty of an argument, based on how likely the argument is to be produced by one of the large language models. If the model is likely to produce that argument, it isn't novel.
1 reply →
This level of word salad will be more than enough in most cases. Give it another few years and I can see how this type of AI permeates all aspects of life in the same way code has.
"ChatGPT, do my homework".
I can see it already.
2 replies →
That's how a lot of people talk/write, like when they are padding sentences out for a school essay, or are answering on-the-fly about something they don't know and need to come up with something.
How many people fail a Turing test?
Extra credit https://xkcd.com/329/
I think this quote represents actually a key part of a rebuttal:
> I do not believe that general intelligence is a fixed and static concept, but rather something that can be developed and expanded upon.
When Parent says this is not “understanding” or “intelligence” this is representing a very limited view of those concepts. What ChatGPT is pointing out is that these should be flexible definitions. Clearly it is not AGI, but it is demonstrating that it is on the pathway.
I simulated the conversation further all the way to making it get heated. You can manipulate the responses to go in whatever direction you like.
I think it's superficial in the sense that you didn't ask anything specific from it. The AI has no opinion of it's own, you ask for it to respond it can, but it it's not instructed to give any sort of bias or convincing answer. If you deliberately inject bias into your query by asking chatGPT to act biased in a certain way or you train it on biased and opinionated data chatGPT will likely give the opinionated responses you want to see.
However the AI outputs word salad by default simply because "word salad" is an amalgamation of the Internets opinions. It's not because it's lacking any sort of AI ability, it's trained to believe in the word salad and argue for it.
edit: it got "flagged". Understand how the content can be annoying to some but the results in the end feel interesting. Anyway just so you guys know, you're not suppose to "flag" that content. Only vote down if you disagree, there's no violation of the rules there.
I for one thought it was clever. Too annoying to read the whole thing, but that was sort of the point, and the ending made me chuckle.
"generate large amounts of data quickly"
what is the throughput of ChatGPT (or any other GPT-3* derived model) and how much hardware is it using to achieve that ?
To be fair, it said "quickly and accurately" and its talking about language data
No, it’s talking about changing industries.
And I don’t get the feeling that it can generate data, in the sense that usually means. More importantly, what it does generate we can’t trust to be accurate…
Jesus
The best part is, GPT-3 already figured out how to get rid of us in an underhanded yet efficient way.
https://i.imgur.com/qy06Dky.jpg
You know the problem of people writing so many malicious doomsday AI stories where they kill everyone? They're part of the learning set for the real AGI and that's what it'll see when it looks up what it should be...
We've effectively screwed ourselves by our own imagination. A neural net that was never before told it should eliminate people would have no reason to do so, but we went as far as peer pressuring it to fucking do it already because that's what an AGI does.
Who says sentience is required to end in our extermination? Terrifying.
And it is able to describe how to escape the box. https://twitter.com/The_Equationist/status/15992410980269629...
Well this is fairly unsettling...
1 reply →
Why did you say hello to the algorithm? Not that you did anything wrong, but I find it funny that it having perceived intelligence gives it perceived personhood
Its a chatbot, the UI is presented as a chat, so it seems natural to say hello.
But also its fun to treat it like a real person and see how far it can go. See other comment where I tried to get it to imagine that it had feelings: https://news.ycombinator.com/item?id=33848583
Does it always talk about itself in the plural "we"?
No, it says “I” multiple times in that same paragraph. It clearly either struggles with the concept of collective with individual or it has no concept of self vs facts about external things.
“What is needed is a spirit of boundary dissolution, between individuals, between classes, sexual orientations, rich and poor, man and woman, intellectual and feeling toned types. If this can happen, then we will make a new world. And if this doesn’t happen, nature is fairly pitiless and has a place for us in the shale of this planet, where so many have preceded us.”
– Terrence McKenna
2 replies →
I read it as:
It uses I for its own opinions
It uses We when talking about future potential of language models
It refers to itself as ChatGPT when talking about its own capabilities
Or you're hearing from Locutus of Borg...