Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

meta.stackoverflow.com

Post reply on HN
Print view View original

31–40 of 143 posts

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#31

Post by autotune »

arthurofbabylon wrote: post #27

What is the mechanism by which a text is determined to originate from ChatGPT? This will be crucial for preventing unfavorable training loops among a myriad of other human-led control efforts.

To determine if a post on StackOverflow is generated by ChatGPT, you could look for certain characteristics that are typical of posts generated by the model. For example, ChatGPT posts may contain repetitive or nonsensical phrases, lack coherence or context, or have a distinctive "machine-like" writing style. Additionally, you can check the user profile of the poster to see if it indicates that the user is a language model, such as ChatGPT. However, keep in mind that not all posts generated by ChatGPT will necessarily exhibit these characteristics, and some human-generated posts may also exhibit similar traits. Ultimately, the best way to determine if a post is generated by ChatGPT is to look for signs of artificiality in the content and context of the post.

- obviously generated by ChatGPT

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#32

Post by bombcar »

arthurofbabylon wrote: post #27

What is the mechanism by which a text is determined to originate from ChatGPT? This will be crucial for preventing unfavorable training loops among a myriad of other human-led control efforts.

I read this policy as being, on the ground, "if you post an answer that is wrong and it smells like ChatGPT you're gonna eat a ban".

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#33

Post by bombcar »

ksaj wrote: post #28
cokeandpepsi wrote: post #14

Earlier quoted context omitted.

I'm not sure about that, it's a temporary ban and the reasoning presented in their post is sound they are swarmmed with a massive influx of incorrect answers to questions that 'look somewhat correct at first glance'

I found a great example to demonstrate. I asked how to transpile Lisp to C. It perfectly described what transpiling is, said (correctly) that not all Lisp structures will translate directly into C, so those parts of the code may be broken, then gave this method for transpiling: sbcl --script example.lsp --output example.c Now... that would be REALLY GREAT if it was true. But SBCL is not a transpiler (except for bytec…

Reminds me of a story I read where an advanced computer (we'd call it AI now) when asked a question it couldn't answer, would simulate an entire universe where the question was answered, and then determine the answer and report back.

And it'd get things like the sbcl command line above, because that was the right answer in the particular universe where the question could be answered.

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#34

Post by __d »

colesantiago wrote: post #3

StackOverflow and Google are finished. This is expected and it may not look like it now, but banning an advancement in technology (even temporary) is usually the first step to irrelevance of those pushing back. In a few years from now when ChatGPT and perhaps the much awaited GPT-4 are released, get better, (or even an implementation is open sourced) it will further plummet the usage for SO and Google. The genie is o…

I played with ChatGPT today. It doesn't replace Google or SO today, and based on the types of issues with what it generated, ChatGPT is not going to replace a competent programmer any time soon.

Where I fear you might be right though, is that the web in general (and so, Google) and sites like SO with their internal content, will inevitably lose value as people post ChatGPT-generated content which superficially looks reasonable, but is subtly wrong. Wading through all those subtly-wrong things looking for the actual answer will be a far worse experience than today.

Which is unfortunate. I don't see a good way to avoid it, sadly. SO's attempt to ban ChatGPT-generated content is useless: while people have the incentive to try to get kudos from SO, and ChatGPT requires skilled, detailed analysis to detect its errors, the ban is unenforceable.

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#35

Post by lovecg »

I think the ban on ChatGPT answers on Stack Overflow is a mistake. Sure, some of the answers might not be correct, but that's true for any tool. And the fact that ChatGPT answers are easy to produce means that more people can contribute, which can only be a good thing for the community. Plus, it's not like the moderators have to accept every answer - they can still weed out the bad ones. Let's not punish everyone because of a few bad apples.

— ChatGPT arguing its own case

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#36

Post by fragmede »

SkyPuncher wrote: post #19
jmugan wrote: post #6

Earlier quoted context omitted.

I don't think so, especially for StackOverflow. I want correct answers, not answers that just look correct.

Yep. Stable diffusion is powerful because there's no "right answer". Just better/worse variations of a prompt. GPT is going to be amazing for telling stories. Particularly, for "too literal" software engineers, like me. However, it doesn't seem to offer much in terms of hard, factual information that technical fields rely on.

There's plenty of text in the process of hard factual information. How about for commenting code? Or describing how a system works? Engineering isn't just a set of diagrams with length notations. Tell me how eg a Scharfenberg coupler operates with only symbols and numbers; no using text that ChatGPT might generate. Or how about all of this page: https://docs.aws.amazon.com/elasticloadbalancing/latest/appl...

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#37

Post by fragmede »

ksaj wrote: post #28
cokeandpepsi wrote: post #14

Earlier quoted context omitted.

I'm not sure about that, it's a temporary ban and the reasoning presented in their post is sound they are swarmmed with a massive influx of incorrect answers to questions that 'look somewhat correct at first glance'

I found a great example to demonstrate. I asked how to transpile Lisp to C. It perfectly described what transpiling is, said (correctly) that not all Lisp structures will translate directly into C, so those parts of the code may be broken, then gave this method for transpiling: sbcl --script example.lsp --output example.c Now... that would be REALLY GREAT if it was true. But SBCL is not a transpiler (except for bytec…

What happens when you give it the Lisp code and ask ChatGPT to do the transpiling?

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#38

Post by 9wzYQbTYsAIc »

insanitybit wrote: post #30
nextlevelwizard wrote: post #20

Earlier quoted context omitted.

Even if it is the same why should it be allowed? We want better answers not same shit or something even marginally worse

I'm asking what makes it worse.

From the post: “The volume of these answers (thousands) and the fact that the answers often require a detailed read by someone with at least some subject matter expertise in order to determine that the answer is actually bad has effectively swamped our volunteer-based quality curation infrastructure.”

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#39

Post by saagarjha »

samcheng wrote: post #26
rogerb wrote: post #22

Earlier quoted context omitted.

prompt: write a short essay as a scientist which is better, celsius or fahrentheit partial answer which some faulty reasoning about precision: First, the Celsius scale is based on a more intuitive and logical reference point. The Celsius scale sets the freezing point of water at 0 degrees and the boiling point of water at 100 degrees, whereas the Fahrenheit scale sets the freezing point of water at 32 degrees and the…

Perfect. This kind of junk is going to litter our collective knowledge bases for years. It looks just good enough to be authoritative, but is fundamentally incorrect.

So a Hacker News comment.

Re: Use of ChatGPT generated text for posts on Stack Overflow is temporarily banned

#40

Post by numbsafari »

esperent wrote: post #18

This seems reasonable to me and I expect this will become a permanent, although hard to enforce, rule on most forums. StackOverflow is supposed to be a place to ask programming questions and receive help from your human peers. It's not a place to ask questions and receive help from an AI. That would be a different product. It might in the future even be a better product, who knows? But it is a different product and t…

> That would be a different product.

That product is called "Google Cloud Support".

Post reply on HN