• This forum is strictly intended to be used by members of the VS Battles wiki. Please only register if you have an autoconfirmed account there, as otherwise your registration will be rejected. If you have already registered once, do not do so again, and contact Antvasima if you encounter any problems.

    For instructions regarding the exact procedure to sign up to this forum, please click here.
  • We need Patreon donations for this forum to have all of its running costs financially secured.

    Community members who help us out will receive badges that give them several different benefits, including the removal of all advertisements in this forum, but donations from non-members are also extremely appreciated.

    Please click here for further information, or here to directly visit our Patreon donations page.
  • Please click here for information about a large petition to help children in need.

AI and ChatGPT Concerns

Status
Not open for further replies.
Fandom is considering to begin using A.I. of their own, but has not set up specific standard directives for or ways to use it yet, as far as I am aware.

See here for what they have told us so far:


Anyway, I do not think that we should allow people to get lazy and avoid doing research, as the A.I. recurrently gives wrong references and answers.

I am open for sensible input from our staff members regarding whether or not it can be used as a starting point if its references are investigated afterwards.

I am also not aware of how good it is at producing reliable calculations that would be easily overviewed for our calc group members to properly evaluate, but suspect that it would likely not be satisfactory in this regard.
 
Seriously? People are using ChatGPT in debates?

Are there any examples of this on here?
 
Fandom is considering to begin using A.I. of their own, but has not set up specific standards for it yet, as far as I am aware.

See here for what they have told us so far:


Anyway, I do not think that we should allow people to get lazy and avoid doing research, as the A.I. recurrently gives wrong references and answers.

I am open for sensible input from our staff members regarding whether or not it can be used as a starting point if its references are investigated afterwards.

I am also not aware of how good it is at producing reliable calculations that would be easily overviewed for our calc group members to properly evaluate, but suspect that it would likely not be satisfactory in this regard.
Aye, true enough. Maybe a rule against ChatGPT generated references would be good, but like... would such a rule really carry any more penalties than those already presented for falsifying scans/references (that is, a ban)? The simple fact that a chatbot is writing them up for the person doesn't matter much to me. The only instance I can envision where that rule may be relevant is if a person tries to earnestly argue that they thought it was reliable somehow, and at that point they ought to be banned for incompetence.
To address these points, in the OP, I had put in text for an advisory notice against using AI as a reference.

It is generally advised to avoid information obtained from chatbots (ex. ChatGPT) and other AI devices as the info they provide may be unreliable. It is best to research information (ideally from .org, .edu, or .gov sites, or academic libraries like Academia, Science Direct, and Researchgate) or ask for human input when possible.
I generally prefer an advisory approach myself, as anything further to me might make things escalate a little quickly.
 
Seriously? People are using ChatGPT in debates?

Are there any examples of this on here?
Let's see... There's this calc that used ChatGPT as a reference: https://vsbattles.fandom.com/wiki/User_blog:Hagane_no_Saiyajin/Vaporization_of_Quartz_Glass_Value

Then there's this CRT I did which I got Word of God from... But part of Word of God used ChatGPT, so that's a mistake I've made.: https://vsbattles.com/threads/grey-goo-reassessment.146047/

Those are two examples I know of. It's mainly a bunch of backbone-y nonsense.
 
I mean, whether or not the user themselves or an AI they use gets wrong information, it gets peer-reviewed in the end nonetheless.
 
I mean, whether or not the user themselves or an AI they use gets wrong information, it gets peer-reviewed in the end nonetheless.
It's not as though this is foolproof, though, to be absolutely fair. Human error is very real and happens frequently enough that using citations generated by an AI would be worthy of a notice up, somewhere, saying "do not be very very dumb".
 
Yeah... I think we can safely prohibit using ChatGPT as a source. There will be other sources out there on the internet.
It's not as though this is foolproof, though, to be absolutely fair. Human error is very real and happens frequently enough that using citations generated by an AI would be worthy of a notice up, somewhere, saying "do not be very very dumb".
I personally agree with these evaluations.
 
So would a version of the text draft that Flashlight237 wrote above be useful for our purposes here?

"It is currently generally prohibited to use information obtained from chatbots (for example ChatGPT) and other A.I. devices as the information they provide may be unreliable. It is best to research information (ideally from .org, .edu, or .gov sites, or academic libraries like Academia, Science Direct, and Researchgate) or ask for human input whenever possible."
 
It's not as though this is foolproof, though, to be absolutely fair. Human error is very real and happens frequently enough that using citations generated by an AI would be worthy of a notice up, somewhere, saying "do not be very very dumb".
Yeah, that's part of the issue; you'll never know who would fall for it and who wouldn't. Then again, I don't think an AI is capable of watching the latest episode of Spongebob or Total Drama or whatever decides to pop up on TV. I would hope nobody resorts to having AI get the receipts for them in that regard; otherwise, call me immature, but I think that would wind up being a skill issue.
 
I could agree to that, after some consideration on the thread, aye. I don't think we ought to take math advice from a bot that can be convinced that 2 + 2 = 5.
 
So would a version of the text draft that Flashlight237 wrote above be useful for our purposes here?

"It is currently generally prohibited to use information obtained from chatbots (for example ChatGPT) and other A.I. devices as the information they provide may be unreliable. It is best to research information (ideally from .org, .edu, or .gov sites, or academic libraries like Academia, Science Direct, and Researchgate) or ask for human input whenever possible."
Looks sufficient to me.
 
In my opinion, it is better to ban the use of ChatGPT in vsbattles, because after all there will be an impact such as calculations and argumentation problems, it will become a dunning Kruger effect and Logical Fallacy, maybe the impact will be big, if not banned. Sorry, I'm just being honest here, I don't want this forum to get messy (even though I'm not a staff member). 🙏
 
In my opinion, it is better to ban the use of ChatGPT in vsbattles, because after all there will be an impact such as calculations and argumentation problems, it will become a dunning Kruger effect and Logical Fallacy, maybe the impact will be big, if not banned. Sorry, I'm just being honest here, I don't want this forum to get messy (even though I'm not a staff member). 🙏
I respect your concerns. Looks like we're going for a reference/debate ban on ChatGPT if the modification s are correct. That's how it seems at least.
 
I can understand not using it as a source for references, but everyone saying that it can’t assist in calculations is just a bit of a silly concept
There’s literally no way to cheat a calculation. If it’s going to get evaluated then it’ll be declared right or wrong.
“It can be wrong sometimes tho” is not an argument in the slightest when it needs to be fact checked before being added to profiles in the first place.
 
I can understand not using it as a source for references, but everyone saying that it can’t assist in calculations is just a bit of a silly concept
There’s literally no way to cheat a calculation. If it’s going to get evaluated then it’ll be declared right or wrong.
“It can be wrong sometimes tho” is not an argument in the slightest when it needs to be fact checked before being added to profiles in the first place.
I agree with this, you mathematically and physically cannot cheat math, therefore i think using it for calcs is fine. However, if you're referencing very specific information (ie: density of a specific star) then you should source where you got it from, which ai like chatGPT can give you the source it used.
 
So would a version of the text draft that Flashlight237 wrote above be useful for our purposes here?

"It is currently generally prohibited to use information obtained from chatbots (for example ChatGPT) and other A.I. devices as the information they provide may be unreliable. It is best to research information (ideally from .org, .edu, or .gov sites, or academic libraries like Academia, Science Direct, and Researchgate) or ask for human input whenever possible."
I'm still more in favour of something like
"If you use any A.I., such as ChatGPT, to obtain information it is your responsibility to check for the correctness of it before presenting it to the community. That includes, but is not limited to, checking whether sources obtained that way exist, are reliable and claim what they should, that any mathematics done is correct, that any formulas or constants obtained actually exist and are used correctly and that any text generated is actually saying the truth."
I feel like even today AI has potentially useful applications and it is steadily advancing. Instead of just prohibiting it, telling people to use it with proper care seems more future proof.
 
I'm still more in favour of something like

I feel like even today AI has potentially useful applications and it is steadily advancing. Instead of just prohibiting it, telling people to use it with proper care seems more future proof.
I agree FRA
 
I agree with this, you mathematically and physically cannot cheat math, therefore i think using it for calcs is fine. However, if you're referencing very specific information (ie: density of a specific star) then you should source where you got it from, which ai like chatGPT can give you the source it used.
Regarding the calc, tools such as ChatGPT is somewhat reliable if you give proper prompt. However, if you does not give proper prompt, such tools have no better than average calculator because of tools' misunderstanding like other misunderstanding of prompt by AI tools.
I feel like even today AI has potentially useful applications and it is steadily advancing. Instead of just prohibiting it, telling people to use it with proper care seems more future proof.
I agree with this. AI is a tool to assist human and it actually makes texting work more efficient. Too bad that more and more people will become unemployed because of the AI and it really reduces my passion of adding more contents in VS Battles Wiki somewhat.
 
Banning AI for stuff like art, editing summaries or the like I think is fine.

But for a ban on math I don't get. Since it's just a worse Wolfram Alpha
I was under the impression Wolfram Alpha was actually pretty good, but its even stuff that it can't do
 
Like, actually imagine if ChatGPT was used to try and downplay Goku just for the hell of it.
I'm imagining it; sounds pretty cool. Give me more DB supporter tears.

AHEM, moving on from that though, the problem with using AI for art is that it just wouldn't be specific enough for the purpose suggested in this thread by multiple users. Characters with low quality or simply nonexistent art would probably be obscure enough that AI wouldn't have omnomnom'd enough data to create anything resembling good rendition of them. So you'd have to rely on getting it just right by trying multiple times, and even then you're probably not gonna get exactly what you were looking for. Also, and this might be just me, but AI art looks uncanny as **** a lot of the time, even when it is technically apt.

As for debating and translations, those should be outright barred in accordance with what people have already said about them. Math help? I'm neutral on it. Thems be my two cents.
 
Also I'd just like to point out that there are multiple normal users commenting on a staff thread without requesting permission, and not all of them are providing solid info. So stop that, otherwise my ratchety bones will become crankier than usual.
 
I was under the impression Wolfram Alpha was actually pretty good, but its even stuff that it can't do
Wolfram is awesome, but my point was that ChatGPT would just be a worse version of that. It giving you like, possible math tips is fine. You just ha e to show the ability to do the formulas and calculations given for a calc.
 
Using GPTChat or AI's for debating purposes sounds pretty iffy. And I basically agree with what KLOL said above.
 
Also I'd just like to point out that there are multiple normal users commenting on a staff thread without requesting permission, and not all of them are providing solid info. So stop that, otherwise my ratchety bones will become crankier than usual.
Earlier in the thread a staff member did say that there was no problem with it
but anyways if that’s how it is then I’d like to request permission to speak here more, since I still have things to say.
 
Earlier in the thread a staff member did say that there was no problem with it
but anyways if that’s how it is then I’d like to request permission to speak here more, since I still have things to say.
If you're talking me, that was a specific person who had a good point of view.
Using GPTChat or AI's for debating purposes sounds pretty iffy. And I basically agree with what KLOL said above.
Yeah. Indian Youtube math tutorials exist for a reason.
 
Earlier in the thread a staff member did say that there was no problem with it
That goes for one person. Not every person.
but anyways if that’s how it is then I’d like to request permission to speak here more, since I still have things to say.
You haven't really provided anything I'd call crucial in this thread. And that's kinda the impetus for allowing anyone that isn't staff to comment here, at all.
 
I'm still more in favour of something like

I feel like even today AI has potentially useful applications and it is steadily advancing. Instead of just prohibiting it, telling people to use it with proper care seems more future proof.
This suggestion seems perfectly fine to me.
 
On one hand, I can understand DT's POV, but on the other hand, my point on having people gather info from reliable sources (ex. .edu sites and Researchgate) should still stand if we want people to do their research.
 
On one hand, I can understand DT's POV, but on the other hand, my point on having people gather info from reliable sources (ex. .edu sites and Researchgate) should still stand if we want people to do their research.
I'm ok with listing some guidelines on reliable sources somewhere. In fact, we could probably add a section for gathering useful ones on some calculation page. Like, maybe restructuring our section on calculators and utilities.
 
my point on having people gather info from reliable sources (ex. .edu sites and Researchgate) should still stand if we want people to do their research.
The problem with that though is that not many people know which websites to get their research from, and often times those websites will need you to pay money or be part of an institution to get the needed research
 
Status
Not open for further replies.
Back
Top