2

At this stage, we don't have a policy relating to answers written by ChatGPT or similar AI.

  • On one hand, ChatGPT is often able to correctly answer questions. So it can be used as a time-saving method. Perhaps it can write better English than some users, or think up better example sentences. AI-generated text is likely the future: ChatGPT and its ilk will be used increasingly broadly, and be an essential tool across multiple domains, much like Google. We're going to have to adapt. (I've recently been using ChatGPT a lot for learning Chinese---it's truly amazing!)

  • On another hand, ChatGPT makes mistakes. Also users are able to blindly copy/paste questions into ChatGPT, and copy/paste ChatGPT-generated answers, which feels fairly close to plagiarism or "let me ChatGPT that for you", and such users may have no idea what it actually says. ChatGPT could answer large swaths of questions at this site, and I don't know how much value there is in having every question answered (possibly multiple times) by ChatGPT.

  • On a third hand, it can be quite difficult to accurately detect ChatGPT-written posts. You can literally ask ChatGPT to make the answer sound like it's not written by ChatGPT. Or you can edit it yourself. I don't want to be making accusations willy nilly. I sometimes read the ChatGPT Reddit, and people get falsely accused of using ChatGPT---it's not nice.

Question: Do we want to allow answers written wholly by e.g. ChatGPT?

At the moment, I feel like users should at least acknowledge e.g. "this answer was generated by ChatGPT". This seems required by Stack Exchange staff:

If it wasn't created by you, attribution is always required here. --- animuson

And OpenAI:

Indicate that the content is AI-generated in a way no user could reasonably miss or misunderstand. --- OpenAI: Sharing & publication policy

Beyond that, if a correct, well-formulated answer goes in the answer box, what's the problem? It's been very controversial network-wide: Ban ChatGPT network-wide was declined, and use of ChatGPT generated text for content on Stack Overflow is temporarily banned.

(Questions involving ChatGPT-generated text are another matter again; we should talk about in another post.)


I asked ChatGPT (GPT 3.5 turbo) about this:

At a Q&A site about the Chinese language, we are discussing whether or not to allow people to submit answers written by ChatGPT. Do we want to allow answers written wholly by e.g. ChatGPT?

And it replied:

As an AI language model, I believe it is important to consider the usefulness and accuracy of the answers written by me or other similar AI models. While I am capable of generating responses to questions about the Chinese language, my answers may not always be perfect or comprehensive. It is important to keep in mind that language is complex and nuanced, and there may be cases where a human's understanding and interpretation of a question or topic may be more appropriate or valuable than an AI-generated response.

That being said, allowing answers written by AI language models like ChatGPT could potentially provide some benefits, such as providing quick and readily available responses to questions that are straightforward or have clear-cut answers. However, it may be more appropriate to allow such answers only as a supplement to human-generated responses, rather than as a replacement for them.

Ultimately, the decision of whether or not to allow answers written by ChatGPT or other AI language models is up to the moderators of the Q&A site. It is important to carefully consider the potential benefits and drawbacks of such an approach before making a decision.

2
  • becky, thanks for the help 😸 somehow, i can’t @ you from new comment in my question. btw, the “pushing back pictures” is considered as **superstitious” in “that area” 🙀 Commented May 13, 2023 at 2:32
  • Quote:- "While I am capable of generating responses to questions about the Chinese language, my answers may not always be perfect or comprehensive" So then should all answers given by ChatGPT have this statement as a "Preface" to all its answers, thus putting all users on "Notice"? Commented Jun 19 at 14:02

2 Answers 2

0

According to Stack Overflow's announcement: https://meta.stackoverflow.com/q/421831/20228153.

Overall, because the average rate of getting correct answers from ChatGPT is too low, the posting of answers created by ChatGPT is substantially harmful to the site and to users who are asking and looking for correct answers.

I do not have data to assert if this is true for a language related community like this one, but my personal feeling is that it might not work well either. What bothers is that, ChatGPT always tries to create answers even it does not know the answer or does not know how to solve the problem, which leads to a lot of garbage, especially if the question is difficult.

Also, I feel that allowing such answers eventually discourages users from giving thoughtful answers.

1
  • (For those who don't know, the Stack Overflow policy is not applicable here. The Stack Exchange policy is: we've decided no network-wide, general policy regarding banning ChatGPT, or other AI generated content, is necessary or helpful at this time. Thus, we can decide either way.) I've copied/pasted a few questions from this site into ChatGPT, and sometimes (not always) it gives great answers; it also sometimes gives total nonsense. Maybe I should create a post showing how ChatGPT answers questions at this site.
    – Becky 李蓓 Mod
    Commented May 12, 2023 at 0:14
0

copied from a comment:

I've copied/pasted a few questions from this site into ChatGPT, and sometimes (not always) it gives great answers

interesting 😸

becky, would you ask the chatgpt this question:

how to comprehend the verse “黑免走入青龍穴”?

it’s the first verse of the 43th text of the prophecy book “pushing back pictures” (推背圖)

i wonder what will the a i say 😼

enter image description here


Edit (becky): I copy/pasted the question into ChatGPT, and it wrote:

The verse "黑免走入青龍穴" is written in Chinese and can be translated to "The black rabbit walks into the blue dragon's cave."

In Chinese culture, the black rabbit is often associated with bad luck or misfortune, while the blue dragon is a symbol of power and strength. The phrase "walking into the dragon's cave" implies walking into danger or taking a risk.

Therefore, the verse can be interpreted as a warning or cautionary advice to avoid taking unnecessary risks or making foolish decisions that could lead to unfortunate consequences. It's similar to the English expression "walking into the lion's den" which means to enter a dangerous situation or confront a powerful adversary.

(PS. I have no idea if this is accurate.)

1
  • @becky, such a i answer is absolutely rubbish 😾 though this 43th “phenomenon” is an oracle, anyone can interpret it freely. most interpretations that i’ve read are: it’s a hint of the time. that, black rabbit refers to 癸卯, blue dragon refers to 甲辰. further, most would say it indicates the year, that, it’s roughly 2023 & 2024. that, some gigantic changes would happen in “that area”. considering the current political tensions between taiwan & “that area”, the situation is worrisome. Commented May 13, 2023 at 2:25

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .