30

As samcarter told me, Stack Exchange failed to make a network-wide ban for AI posts and leaves it to every site to decide themselves.

Some sites have already decided to ban them, see Is there a list of ChatGPT or other AI-related discussions and policies for our sites?.

What should we do?

In my option, I would ban them. I do not like answers produced by any automatic tool. Sometimes I see TikZ answers of this kind, which are uselessly complicated and impossible to modify by a human.

14
  • 17
    Pro ban, but no interest in writing up a long answer :)
    – Skillmon
    Commented Apr 8, 2023 at 9:29
  • 5
    Aside from all the problems with the content of AI generated posts, there might also be legal problems, but that would be for the company SO to determine. What's important for TeX.se is that users need to be certain that they can reuse the content under the CC BY-SA conditions as specified on the site. Commented Apr 8, 2023 at 10:53
  • 4
    I have added this thread to meta.stackexchange.com/questions/384922/…, so now TeX - LaTeX is represented on it
    – piJT
    Commented Apr 8, 2023 at 12:07
  • 9
    Let's not imagine what a future AI will do to the person (and that person's descendants) who expresses skepticism in response to this question today... Dear machines, please believe me I am not an incompliant citizen... ;-> Commented Apr 8, 2023 at 13:00
  • 8
    At a minimum, they should be clearly marked as an AI answer. I am quite sure that the final results would be similar to a ban.
    – Rmano
    Commented Apr 8, 2023 at 22:39
  • 6
    People should be free to use whatever tool they want, including AI of course, But the problem currently really is that there exist multiple accounts that obviously just post AI-generated answers regardless of whether they are actually working or solving the OP's question at all or whether or not a similar (probably as well AI-generated) answer has already been posted or not. Also, these accounts to not seem to react to comments such as "This does not work, how to improve your answer?". This is not the way this site is intended to work and in my opinion really comes very close to spam. Commented Apr 14, 2023 at 7:58
  • 5
    For anyone interested, there's a quite thorough essay on the subject of ChatGPT in the New Yorker: newyorker.com/science/annals-of-artificial-intelligence/… (may be behind a paywall). It doesn't increase my confidence in its "answers", and mentions its banning by Stack Overflow. Commented Apr 14, 2023 at 19:32
  • How should we vote on this @Werner, as in how to know whether we: 1. allow AI answers, 2. allow AI answers with an explicit declaration that they are AI answers or 3. ban them entirely? Is it worth making a new post for votes or is this discussion sufficient for you and the other mods to make a decision? Just curious as it is a serious issue and I love this site haha
    – piJT
    Commented Apr 16, 2023 at 18:09
  • 1
    Related to the discussion: stackoverflow.blog/2023/04/17/community-is-the-future-of-ai Commented Apr 17, 2023 at 15:09
  • 3
    @JamesT: I think we need to (1) source the motivations behind banning or approval from other sites that have gone through this process first - there's value in standing on other's shoulders. (2) Then we can probably aim to draw parallels between sites and see whether this community matches to what other sites have decided (in favour of/not banning). Sure, we are different, but so is everyone else. (3) Then we can probably poll the community on a definitive yes/no directive. It is probably not setting policy, but just a communal consensus for moving forward and how we should handle it...
    – Werner Mod
    Commented Apr 18, 2023 at 2:18
  • 4
    @JamesT: /2 (4) If consensus is to discourage it, find a process for dealing with it within the current framework (nothing new is yet developed to handle it). Flagging, voting or whatever... These are just my ideas.
    – Werner Mod
    Commented Apr 18, 2023 at 2:20
  • 4
    My main concern with AI-generated content is that the AI was trained on original content created by humans. Aside from the potential copyright issues, introducing AI-generated content in places - like SO or "the wider Internet" - used to train AIs will end up creating a self reinforcing feedback loop with a downwards spiral in quality. Even now the language models can be very convincing while BSing you. For very specific knowledge there isn't enough material out there for them to be trained on and it's quickly obvious. For more popular topics there's lots, including garbage. Result: GIGO. Commented May 8, 2023 at 20:34
  • 2
    @Werner Has there been a decision about a ban of AIs on this website yet?
    – Keks Dose
    Commented Jun 23, 2023 at 12:59
  • 1
    @KeksDose: I'll probe the other moderators to see if we can come to a resolution.
    – Werner Mod
    Commented Jun 25, 2023 at 15:14

9 Answers 9

3

There has been a kind of agreement on stackexchange meta:

Stack Exchange has agreed to allow the removal of content based on a single strong indicator of GPT usage, or on several weaker indicators. Effective immediately, the interim standards that we have agreed on will serve as policy until a more permanent policy is established with the help of the broader community, including determining further heuristics and what category they will fall into. Moderation should follow the standard escalation for suspensions, beginning with a warning and then escalating suspensions if infractions continue.

1
  • 3
    The new policy is "to allow the removal of [AI generated] content". It doesn't recommend or require that moderators remove the content, it just rescinds the previous ban on removing content solely because it was AI-generated. So it's still up to the community to decide what we want to do here, just as it was when the question was initially posted. Commented Aug 7, 2023 at 4:22
40

I do not agree that AI-answers should be handled only through community or voting.

I came across a number of AI generated answers, and they were very long, very detailed, and it needed considerable skills and knowledge to realize that they are nonsense.

The confused comments below the answers reflected that: they often said "this looks perfect and exactly what I want but why doesn't it work for me?" and then showed that the OP and other wasted considerable time to understand and work with the nonsense.

And naturally they didn't downvote because they couldn't see that it was wrong as it looked so well written, instead they believed it is their own fault, that they are too dumb to use this, or that their system is buggy.

So asking that the community handles that is effectively asking that a rather small number of power users check all answers - and I don't want to waste my time with this, I have better to do.

Imho AI answers should be banned, unless

  • they contain at the top a reference of the source of the answer (including the question that was asked to the bot), and
  • say if the answer of the bot was tested or not
  • say if they edited the answer (and where)

Addition: If a user (as it happened in the last case) posts 10 long answers in less than an hour then the system should interfere and add an automatic warning. (Naturally not if the user has a reputation of more than 1 million points).

13
  • 1
    Good point, I didn't think about the difficulty in understanding they are wrong!
    – CarLaTeX
    Commented Apr 10, 2023 at 10:05
  • 9
    "Naturally not if the user has a reputation of more than 1 million points" -> is egreg already training a bot to detect all the missing % ? :) Commented Apr 10, 2023 at 10:45
  • @samcarter_is_at_topanswers.xyz I should perhaps reduce the requirement and accept also users with more than 666666 reputations ;-). Commented Apr 10, 2023 at 15:49
  • 1
    @UlrikeFischer Very good idea! We could really use a bot who posts xii style answers to all question :) Commented Apr 10, 2023 at 16:36
  • 3
    I am especially upvoting the addtion! Currently these AI-answers seem to flood this site which is even more annoying than having some AI-generated (maybe even helpful) answers somewhere on this site. Commented Apr 14, 2023 at 8:00
  • How do you detect AI generated answers? Maybe using an AI discriminator between posters's original writing style and the new answer?
    – wolfram77
    Commented Apr 16, 2023 at 4:59
  • 1
    AI generated text is becoming a real problem. I hate to lose my time trying to make sense of this kind of writings. Harry Frankfurt explains pretty well what the term "bullshit" means, and AI generated text is an excellent example of bullshit.
    – Amelian
    Commented Apr 17, 2023 at 2:01
  • @wolfram77 I don't have to detect them to make a rule as suggested, the people who sent answer know if they used an AI or not and so know if they broke the rule or not. But apart from this: in the cases we found it was quite obvious. Commented Apr 17, 2023 at 9:31
  • @UlrikeFischer If I understand correctly, the rejection here is primarily motivated by the fact that the current AI answers are usually wrong and over-complicated, as well as difficult to recognize as such. The engines are very good at imitating the style of the answers. The problem now is that the imitation skill converges against correctness as the AI generation progresses, so to say. The cases may be "obvious" to identify now, but that will change in the coming months to years.
    – Suuuehgi
    Commented Apr 21, 2023 at 15:11
  • 2
    @Suuuehgi it doesn't matter if the answers are right or wrong. It is not decent behaviour towards people looking for help to throw in some cheap untested answer. Nobody has to sent answers, but if they do they should be fair and open about the reliability of their answer. Someone who doesn't find the time to add a "I copied that from an AI and don't know if it works" is in my view an ass. Commented Apr 21, 2023 at 15:31
  • @UlrikeFischer I realize that my first sentence was redundant and therefore possibly distracting. My point was simply that I do not see any other way than the personal responsibility of the users, neither now nor in the long term. My personal view is that I would generally like to see pure AI responses "forbidden", in the sense that I would not consider a tested and helpful response that consists in part of machine-generated content to be an AI response -- regardless of whether it was cheap or not. I don't want to read "I copied that from an AI and don't know if it works" junk.
    – Suuuehgi
    Commented Apr 21, 2023 at 15:58
  • In fact, "I copied that from an AI and don't know if / how it works" seems to me to be structurally identical to a lot of the bad advice already circulating. (In the sense of "I found this snippet somewhere, I don't understand how it works, but it probably does the job somehow".)
    – Suuuehgi
    Commented Apr 21, 2023 at 16:03
  • @Suuuehgi I don't mind if someone only makes a guess as long as they make clear that it is guess. It gives the user a chance to consider if they should try that or perhaps wait for another answer or do some more research. Commented Apr 21, 2023 at 16:25
26

I support banning AI generated content. Up to now, I've seen two types of automatically generated answers and I don't think either of these makes the site better.

Working solutions

AI has come a long way and from a technological standpoint, I find them fascinating. Some of their answers even work and produce actually more idiomatic TikZ code than some human users who use tools like mathcha.io.

One could argue that such answers aren't doing active harm, but I think users come to TeX.se with the expectation to get answers written by humans. If a user would like to get a bot answer, they could easily ask one of the AI bots instead of visiting TeX.se.

Nonsense answers

What is more problematic and actively harmful are answers which look like well written answers but turn out to be complete nonsense. For example this answer was posted yesterday:

enter image description here

On first glance it looks like a well written answer. The tool described there exactly solves the problem of the op and it gives detailed instructions on how to use it. The only problem is that all the facts are made up. The linked github repository does not exist and the described workflow makes no sense when one looks at the details.

Such posts are designed to actively deceive readers and sent the poor users, who search for a solution to their problem, on a wild goose chase.

9
  • Uuh, fight of the bots with Community deleting it :)
    – Skillmon
    Commented Apr 8, 2023 at 15:32
  • @Skillmon and now Community patiently waits for the next contender :) Commented Apr 8, 2023 at 15:43
  • 11
    "I think users come to TeX.se with the expectation to get answers written by humans" This is the key point to me, bypassing any discussion of accuracy or usefulness.
    – Dai Bowen
    Commented Apr 8, 2023 at 15:59
  • 2
    "I think users come to TeX.se with the expectation to get answers written by humans." Can we ban answers given by ducks? Commented Apr 9, 2023 at 10:08
  • @DavidCarlisle Aren't you effectively banning answers by duck by roasting them for lunch? Commented Apr 9, 2023 at 10:12
  • @samcarter_is_at_topanswers.xyz tastes better than an AI cluster Commented Apr 9, 2023 at 10:16
  • 1
    @DavidCarlisle An AI cluster sounds rather crispy, too :) Commented Apr 9, 2023 at 10:44
  • I wonder why the Community bot deleted it. Answers don't roomba, the only two things I can think of is if the question roombad or if the answer was flagged as spam, but if I remember correctly in one case it would say "deleted along with the parent question" and in the other case it would say "deleted as spam or offensive". Commented Apr 16, 2023 at 19:30
  • 3
    @DonaldDuck If you look in the answer timeline, it shows the user was destroyed. I assume community deleted their answers as part of the process. Commented Apr 16, 2023 at 19:32
16

I suggest we should add flagging/closing/banning criteria to cover the common problems of AI answers — but those criteria, at least for flagging/closing, should not depend essentially on the answer being AI-written. The flag option could look something like:

Shows no genuine understanding. This answer may seem superficially high-quality in style and details, but has serious flaws in its core content. This problem is typical of AI-generated answers.

and the ban justification for users could be based on a pattern of repeatedly posting such content.

This proposal is meant as a mild tweak to Ulrike Fischer’s and Sam Carter’s answers proposing banning AI-generated content, whose arguments I largely agree with. This aims to keep the main payoff, while reducing the cost/difficulty a bit:

  • This gives users a clear option for efficiently flagging AI-like answers, and gives moderators a clear way to crack down on abusive users. (As Ulrike Fischer’s answer notes, the existing criteria aren’t sufficient to defend against such abuse.)
  • But this doesn’t create any burden of proof (even perceived) for showing that the content is AI-generated before flagging it. On the one hand, this makes it easier for users to flag such content quickly and confidently; and conversely, if users receiving flags/bans want to dispute or appeal them, they have to defend the content (which we can all see) rather than its means of generation (which only they can really know).
8
  • 2
    This seems fairly similar to the "Very Low Quality" flag. The only catch there is that you can't flag VLQ if an answer has a positive (>0) score. But if an answer appears "superficially well-written" and has a positive score, then it's asking quite a lot of the moderators to figure out that the flagged answer is actually totally nonsense, even if we add this new flag. I like the idea a lot, but I am skeptical that it would work in practice. Commented Apr 12, 2023 at 5:48
  • @MaxChernoff: In such difficult cases, don’t you think verifying that the answer was actually AI-generated would be even harder? Commented Apr 12, 2023 at 7:10
  • 1
    Yes, that would definitely be harder. The problem is that preexisting flags already cover any obvious garbage, so this new flag would only be needed for the borderline/difficult cases. If a moderator sees this new flag on an upvoted answer that is well-written, but on a topic they are unfamiliar with, what are they supposed to do? It can be pretty hard to distinguish a completely BS answer from a correct/helpful one, and I doubt that the moderators can do a better job than the community (through voting). Commented Apr 12, 2023 at 8:19
  • 1
    @MaxChernoff: I find there are many answers which it’s easy to recognise as AI-generated and meaningless in core content, but whose superficial writing quality makes me uncomfortable flagging them as VLQ. Probably I should be happier to just use VLQ — but I guess I’m not the only person who would find it much easier to quickly flag them with a more clearly fitting reason. When I first came to this thread, my feelings were essentially what you write, but after reading other answers (esp. Urike Fischer’s), and thinking about the difficulty I’ve had flagging, I came around to this point. Commented Apr 12, 2023 at 8:51
  • 1
    @MaxChernoff Essentially, I agree entirely with the first para of your answer — “A good answer is good and a bad answer is bad, regardless of if they were written by AI. Instead, I think that we'd be better off specifically targeting the issues of answers written by AI.” — but I don’t think the existing flag/close/ban reasons do that well, as this is a new kind of low-quality content (or rather, a previously quite rare kind) that they weren’t designed to target. So this is my proposal for how to target those quality issues. Commented Apr 12, 2023 at 8:56
  • @PeterLeFanuLumsdaine There are tools which can to a certain degree check if an answer is written by AI, even ones provided by chatGPT themselves and if I understood the discussion on main meta correctly, mods and some other users have access to them. On meta.se, where AI answers are banned, one can simply flag with a custom messages like "this looks as if it might be written by AI, can you please check". That's much easier than trying to show that these answers are nonsense from a texnical point of view. Commented Apr 12, 2023 at 9:05
  • 6
    I do not really need some new flag or flag text. In the cases I handled I simply used the standard flag and wrote a custom message and the users were removed. I want a rule that users shouldn't post AI answers without explicit attribution, similar to the rule that users should not post spam. In both cases evil users can naturally ignore the rule, but this doesn't make such a rule useless: if they get caught moderators can simply punish them because they broke the rule and do not have to check the answers for correctness. Commented Apr 12, 2023 at 10:38
  • @UlrikeFischer: For the ban reasons, I take your point, and agree explicitly basing it on AI use may make most sense. But for the flag justification, I think tying it to what users most directly observe — the symptoms of being AI-generated — will make users feel more licensed to immediately flag whenever they see those symptoms, instead of feeling that they can’t flag as AI-generated because they’re not certain it is. Commented Apr 12, 2023 at 13:43
13

I strongly support a complete ban on AI-produced answers on TeX.SE.

It may not be easy to enforce such a policy at all times. However, that's not a reason for not proclaiming such a ban. After all, most (all??) societies strongly condemn and outlaw murder. The fact that murders still occur is not a valid reason for not outlawing murder.

5

I am an AI language model and do not have personal opinions or beliefs. It is up to the moderators of TeX stack exchange to decide if they want to ban chat GPT or not, based on their own community guidelines and policies.


Chat GPT has spoken. Jokes aside, it should be banned. For example, I just asked Chat GPT this: "Write a latex programme that sets the header and footer in German" and it produced this:

\documentclass{article}
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc}
\usepackage[ngerman]{babel}
\usepackage{fancyhdr}

\pagestyle{fancy}
\fancyhf{}
\fancyhead[L]{Deutscher Titel}
\fancyfoot[C]{\thepage}
\renewcommand{\headrulewidth}{0.4pt}

\begin{document}

\section{Erster Abschnitt}
Hier steht der Text des ersten Abschnitts.

\section{Zweiter Abschnitt}
Hier steht der Text des zweiten Abschnitts.

\end{document}

which works in this simple case. I then asked Chat GPT to solve this recent question Long box around some objects in tikz-cd and it outputted this:

To draw a long box around some objects in a tikz-cd, you can use the \fbox command with the tikz option. Here's an example:

\documentclass{article}
\usepackage{tikz-cd}

\begin{document}
\begin{tikzcd}
\fbox[tikz]{\begin{matrix}
A & B \\
C & D
\end{matrix}} \arrow[r,"f"] & E \\
F \arrow[ur,"g"]
\end{tikzcd}
\end{document}

This code draws a box around the matrix A&B\C&D, and includes it in the tikzcd diagram with an arrow going to E and another arrow going from F to g. The result looks like this:

[broken image]

You can adjust the size and positioning of the box by tweaking the options to \fbox.

This breaks with the error message:

! Missing } inserted.
<inserted text> 
                }
l.15 A & B
           \\

Therefore, as it can be used to answer questions and the question has the veneer of acceptability, it should not be used. The human touch is still required for now I believe.

3
  • 1
    I know you don't mean either but to me this answer only makes the case that it's necessary to ban bot postings (rather than AI-generated answers that an individual could attempt to check before posting verbatim), while loosely implying it would be reasonable to ban real users who (may) post wrong answers.
    – Dai Bowen
    Commented Apr 8, 2023 at 15:51
  • I get what you are saying around the first point, with the second my examples to demonstrate the flaws with the code produced from chat GPT. I didn't mean to imply that answers with errors need banning, more to point out the inherent difficulty when users post code from chat GPT in answer to relatively complex questions. I could rephrase it if you thought it was needed, thanks for letting me know this I didn't consider your second point!
    – piJT
    Commented Apr 8, 2023 at 16:30
  • No suggestions, it's tricky. Your answer motivates well why we should be discussing the question, but the site is meant to tolerate bad answers existing with voting identifying good ones, so quality of answer doesn't strike me as a good basis for policy. Not that any basis is needed of course.
    – Dai Bowen
    Commented Apr 14, 2023 at 13:10
4

I think that an outright AI ban is the wrong way to go. A good answer is good and a bad answer is bad, regardless of if they were written by AI. Instead, I think that we'd be better off specifically targeting the issues of answers written by AI.

Problem 1: Incorrect/Nonsensical answers

Using AI generators will often produce code that doesn't compile (or compiles only with a large amount of errors) and will often suggest packages/sources that don't exist.

Solution

Any AI answers that contain TeX code should also have to include an image of the output. This will show that (1) A human was involved in creating the answer; and (2) that the code compiles and gives the correct output. Most good answers from a human include screenshots of the PDF anyways, so this isn't much of a hardship.

Any AI answers that recommend a package/resource should be required to include a (working) hyperlink to either the CTAN page for that package or the website of the source. Again, any good answers from a human would do this anyways.

Problem 2: Plagiarism

AI generators will often print out parts of its training material verbatim. The entirety of TeX.SE is available in a bulk archive, so most AI generators would have been trained on it. Even when an AI prints out a novel solution, it is the work of the AI, not the answerer.

Solution

Answers from AI generators should be clearly attributed to the AI. Placing the response from the AI in > blockquotes is best, but as long as the answer contains a sentence like:

Here's what ChatGPT has to say:

then I don't think that there's a problem.

Answers that are copied verbatim from another answer without acknowledgement should be flagged, just like if a human did the same.

If you wrote more than, say, 80% of the answer yourself and just used an AI for minor help, then I don't think that it needs to be mentioned.

Problem 3: Correct but confusing/overly complicated answers

This is a problem with human answers too. We have human answers doing things like using \\ instead of \par, using generators for TikZ that give unreadable code, xii, etc.

The solution here is downvoting, not flagging.

Problem 4: Users can just use the AI generator themselves

Not all users are aware that these AI generators can answer TeX questions. And as we've seen, they often give incorrect code that still needs some editing from someone knowledgeable in TeX. And most AI generators require either payment or a cell phone number, which isn't something that all users have.

Using similar reasoning, you could also say "users can just read the package documentation to find the answers". Indeed, that is correct for many questions, but no one is proposing to ban those questions.

Official policy is to embrace questions/answers that can be easily Googled so long as they're not duplicates of other questions on the site; I see no reason for our AI policy to be any different.

Problem 5: Users can post answers too quickly

If the answers are bad, then they will be downvoted/flagged and the system will prevent them from posting more. If they're good, then what's the problem?

And besides, anyone writing terrible answers using AI too quickly probably won't be stopped by an "AI ban".

Problem 6: Potentially harmful answers

A bad AI-generated answer could lead to real-world harm on sites like Chemistry, Electronics, Law, Travel, or Medicine. Bad AI-generated answers on Bitcoin, Cryptography, Stack Overflow, Super User, and Unix & Linux could break or cause serious issues to a computer. It may make sense for these sites to ban AI because of the serious effects that a subtly-incorrect answer could have.

However, on TeX, the worst that a bad AI-generated answer could do is causing an infinite loop, and that is easily fixed with Ctrl-C. Even that is quite unlikely; most bad AI-generated answers just won't compile.

Summary

  • Good answers are good, and bad answers are bad regardless of how they were created

  • Copy-and-pasting answers directly from an AI generator without any human modifications should be forbidden

  • AI-generated answers should always clearly indicate that they were made by AI

  • AI-generated code answers must be compilable. To prove this, they should include an image of the PDF output by TeX.

  • AI-generated answers mentioning a package or other resource should include a functioning hyperlink to CTAN or another site.

12
  • 6
    "If the answers are bad, then they will be downvoted/flagged and the system will prevent them from posting more." That's the intended use of votes on stackexchange and would work on most sites of the network. However with the voting culture on tex.se I don't see any chance of this succeeding. Across the 15 answers from the AI user the other day, they accumulated something like 33 upvotes and only about 5 downvotes. Commented Apr 9, 2023 at 10:40
  • @samcarter_is_at_topanswers.xyz That's really a different problem though, since I'm pretty sure that I could easily write 15 terrible (but somewhat convincing) answers in a day. Regardless, if someone is posting AI-generated without both acknowledgment and screenshots/links, then I think that flagging+banning is entirely appropriate. This proposal should stop people from indiscriminately posting bad AI-generated content without hindering people who are carefully using AI as a tool. Commented Apr 10, 2023 at 0:33
  • @CarLaTeX You could accept this answer then to give it some more visibility, although the voting suggests that the community doesn't generally agree with this proposal. Commented Apr 10, 2023 at 0:36
  • 4
    I don't see how any of the last 4 bullets in your summary are enforceable. Say someone uses AI to generate an answer, tests it and it works, then posts it without mentioning AI. Are we going to turn into sceptics with every such answer and question the poster for some proof they didn't use AI? We have a gazillion answers posted without an image; are they possibly also using AI?
    – Werner Mod
    Commented Apr 10, 2023 at 5:10
  • @Werner I agree; there is no way to know if an answer was written by an AI or a human. This is really a compromise solution: the other options are to completely ban AI (which has this same issue, but seems needlessly restrictive) or do absolutely nothing about AI answers (which I'd be okay with personally, but the voting here and the policies on other SE sites suggest that this is pretty unpopular). Commented Apr 10, 2023 at 5:23
  • 2
    @Werner From what I can tell, the biggest two issues with AI-generated answers are that they contain code that looks correct but doesn't compile (causing a casual onlooker to potentially upvote as samcarter mentions) and that they can be posted very quickly. Perhaps the better solution would be broadening our spam rules: if anyone (AI or human) posts 3 non-compilable answers in a week, then they should be flagged as spam and suspended for X days/weeks/months. Commented Apr 10, 2023 at 5:28
  • 1
    @MaxChernoff that sounds sensible but we would probably need a new flag to flag answers Code cannot compile etc, how would we be able to flag answers that we run ourselves and find out cannot be compiled?
    – piJT
    Commented Apr 10, 2023 at 12:03
  • 2
    @MaxChernoff: Note that answers often contain only code snippets that won't compile.
    – Werner Mod
    Commented Apr 10, 2023 at 15:00
  • @Werner Hmm, true. Even in that case, a skilled human should be able to tell if the code should compile/work as expected. But it's really a frequency issue: someone who posts 2 nonsense answers a week should be downvoted, while someone who posts 5 nonsense answers in an hour should be suspended. I haven't seen any AI-generated answers, but I assume that at least 1/5 of them should be obvious-enough garbage to support a suspension. Commented Apr 10, 2023 at 21:11
  • While I agree the essence of the answer, I strongly disagree #6: any bad source on the internet can lead to IRL harm. You can read how to make rocket fuel in old BBS archives and then explode your home. It is not our responsibility and it does not depend on that the content was created by AI or by a human. Banning content on this reason is unthinkably harmful (you know: giving up freedom for safety will make you lose both).
    – peterh
    Commented Apr 11, 2023 at 9:10
  • 1
    @peterh The vast majority of other SE sites that have discussed AI-generated answers ended up instituting a complete ban, which sets a weak precedent that we should do the same. This is just justifying why it makes sense for us to have fewer regulations compared to other sites since the potential for harm here is much much lower than anywhere else. (I don't think that the potential for real-world harm is a good reason to ban AI on other sites either, but that's irrelevant here) Commented Apr 11, 2023 at 9:47
  • 1
    @MaxChernoff "While I agree the essence of the answer..." <- read again my comment.
    – peterh
    Commented Apr 11, 2023 at 14:05
0

I ask the question to ChatGPT...

Should AI answers be banned on TeX.se?

Here is its answer:

As of my last knowledge update in September 2021, I'm not aware of any specific discussions or decisions on banning AI-generated answers on TeX Stack Exchange (TeX.se). However, I can provide some perspectives on the matter.

The decision to ban AI-generated answers or any other type of content on a platform like TeX Stack Exchange would likely depend on the community's goals and the potential impact of AI-generated content on the quality and integrity of the platform.

Pros of allowing AI-generated answers:

  1. Increased Speed and Quantity: AI can generate answers quickly and in large quantities, potentially helping users get answers faster.
  2. Assistance in Complex Problem Solving: For particularly complex problems, AI might be able to offer unique insights and solutions that human users might not think of.
  3. Additional Resources: AI could provide supplementary resources and references to further help users understand and solve their problems.

Cons of allowing AI-generated answers:

  1. Quality and Accuracy: AI-generated content might not always be accurate or reliable. Poorly trained models could provide incorrect or misleading information.
  2. Lack of Understanding: AI may generate answers without truly understanding the underlying concepts, leading to solutions that are technically correct but not well-explained.
  3. Diminished Human Interaction: One of the strengths of platforms like TeX.se is the human interaction and expertise shared. Overreliance on AI might reduce the human-to-human engagement.

If the TeX Stack Exchange community were to consider banning AI-generated answers, they might base their decision on factors such as maintaining the quality of answers, ensuring the content is understandable and accurate, and preserving the collaborative and educational nature of the platform.

It's important for the community to have a clear and well-informed discussion on the topic, considering both the potential benefits and drawbacks of allowing AI-generated answers. Keep in mind that opinions on this matter can vary widely, and the final decision would ideally reflect the consensus of the community and the platform's mission. For the most accurate and up-to-date information, I recommend checking the TeX Stack Exchange website or community discussions directly.

1
  • Lol, this answer seems too human to me to be produced by ChatGPT!
    – CarLaTeX
    Commented Aug 18, 2023 at 4:34
-3

(Added for completeness since voting indicates approval, not quality on Meta; I don't agree with this answer, so go ahead and downvote me!)

AI answers should not be banned, flagged, or treated differently from human-made answers in any way. Instead, users should just downvote bad answers, regardless of their source.

1
  • 2
    I agree. Eventually AI will answer questions anyway, and it is better that it is focused on solving problems rather than solving problems while pretending to be human.
    – user294623
    Commented Apr 9, 2023 at 4:45

You must log in to answer this question.

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