1

I am noticing more and more purely AI-generated replies in various threads on Spiceworks. In most cases, those replies look like noise and, in my opinion, are killing all the community’s fun and purpose.

So I have a question for you guys - what will we do with those?

Is AI a death to communities and forums as such?

@Rod-IT @jonahzona @PatrickFarrell @adrianyong4136 @kevinhsieh @bob2213 @da-schmoo

22 Spice ups
2

In my opinion, no it’s not ok. I’ve used chatgpt to write some scripts for things I need, and they often don’t work. While it’s useful to give me a point to what I actually need to do, it’s not an answer.

I do understand that a lot of the replies to questions here are also “not THE answer”, but pointers towards it, but they usually come from people’s prior experience, and are therefore “more use” than a chatbot’s idea of what’s right.

Perhaps there should be a marker to say whether the reply is chatbot generated or not, and we can hope to rely on people’s honesty to show that?

5 Spice ups
3

I’ve seen a few replies that where not even close to being correct, and obviously never verified by the respondent. That said, I’m using AI more and more, and Spiceworks less and less for questions/answers. I will always love Spiceworks for the community aspect.

4

The AI said it’s replies are far superior to any human generated replies, and therefore it is perfectly acceptable, and highly recommended, to use AI replies on the forums in the this community.

9b2f5cff-536f-44d5-bb8a-79e99be6fb7d-Untitled_(Small).jpg

4 Spice ups
5

We haven’t banned it outright… some people use it in a useful way, and acknowledge that it is from ChatGPT. But I have seen a lot of answers that are not acknowledged as well, and answers from spammers trying to establish accounts.

–First of all, if it is bad information, report it.

–If you suspect a spammer, report it.

–If it is off-topic report it.

–If you aren’t sure, go ahead and report it. We’re happy to take a look.

If you use an AI to answer a question, tell us where the answer is coming from, and remember that you are responsible for the accuracy of your posts, no matter where you got the information.

Z

16 Spice ups
6

At first ChatGPT said:

"Using AI-generated replies on technical forums can be acceptable in some circumstances, but it’s important to consider the context and potential limitations of the AI-generated response.

AI-generated replies can be useful for providing quick answers to common technical questions. For example, if a forum user is asking about a specific error message, an AI-generated response might be able to quickly provide a solution based on past instances of that error.

However, it’s important to keep in mind that AI-generated responses are not always perfect or comprehensive. AI algorithms are trained on large datasets of information, but they may not be able to account for every unique scenario or edge case. As a result, using AI-generated responses exclusively on a technical forum may not always be the best approach.

Additionally, it’s important to be transparent about the use of AI-generated responses. If a user is receiving an AI-generated response, they should be made aware of it so that they can make an informed decision about whether to trust the information provided.

Overall, using AI-generated responses on technical forums can be a helpful tool for providing quick solutions to common problems, but it’s important to consider

the context and limitations of the technology before relying on it exclusively."

But later I sort of heard this little sound that sounded like Exterminate.

7 Spice ups
7

IF it is clearly reckognizable, I’ll say it’s ok.

in my supermarket, I almost never go to the self-service check-out, because I love the personal interaction with the cashier.

similar in chats, I love the interaction with persons.

4 Spice ups
8

I think it is not right to use an AI to answer a question in a community like this. If the questioner wants an answer from an AI, he can ask it himself. If he asks a question in the community, he wants an answer from a human.

9 Spice ups
9

Usually, when I quote a ChatGPT response (I’m 99.9% I cite the source when I do), it’s because I was curious as to what the response would be.

2 Spice ups
10

With ChatGPT, Bing and Bard, all giving different responses to code questions, I will run whatever question I have thru all 3 now.

Will I post an answer from any of them here verbatim? nope. Most answers can only be used as a starting point.

I will say that if you generally run a working script thru the AI and ask for improvements, you get an improved working script.

Starting from scratch, not so much.

Without AI, my most recent GUI script would have taken twice as long to complete since I am in no way a Powershell expert.

1 Spice up
11

Many times, a person can just Google their question instead of posting, but IMHO they’re too lazy.ni can’t tell you how many times I have taken their question, essentially verbatim, put it into Google and then replied with the top link (after verifying the content).

Remember HLMGTFY.com? I have thought about an equivalent for ChatGPT. I asked ChatGPT what it would take to build such a site, but it was too vague, and didn’t provide any code.

I think that the AI answers are better structured than most human generated answers, and more complete, for most situations. They should be labeled, IMHO.

I posted an answer on another similar site to SW, clearly labeled the response as ChatGPT, and it was flagged. Was I being somewhat lazy, instead of taking maybe 5-10 minutes to write it all out. The OP was lazy too, as most questions on these sites have already been answered somewhere on the Interwebs.

When the AI posts responses that build community, let me know.

2 Spice ups
12

How does everyone feel about answers that are basically copy/pastes from Google results?

I believe those have existed in not insignificant numbers for years.

If you’re ok with the Google answers, what difference, if any, do you see between those and AI answers?

1 Spice up
13

I come here to talk to IT professionals not chat bots. If you’re taking peoples questions, throwing them into ChatGPT and then posting the reply then what is the point of you being here? There isn’t a point. Leave. Spiceworks could come up with an Ask ChatGPT interface and make that a group to itself. Then you know what you are getting and you know what is providing the answer.

Someone might as well take your post, put it on stack overflow or reddit, and then take the top reply and post that back instead and try and take credit for it. If we caught someone doing that, they wouldn’t be here anymore and yet that’s exactly what they are doing if you use ChatGPT to answer questions.

If someone wants to use ChatGPT to ask a question they are more than welcome to go do that on their own. Nobody is saying they can’t. What you can not do is pretend to be a community member making a contribution when someone/something else is doing the work. That’s called lying.

So the short of it is:

If you are doing this and being upfront, go do it somewhere else. I want to talk to you, not use you as a ChatGPT front end.

If you are doing this and not being completely up front, you should close your account, or your account should be closed for you.

5 Spice ups
14

Uh, no Patrick I think I will stay.

I found ChatGPT to be incredibly useful when writing regex. If I see a regex question, I will likely use it, and give it credit. The main goal is to show people that these tools are out there and hope that they use them next time instead of posting to a forum.

AI is great for questions that are black and white and have direct answers. It is the questions that have more nuance to them that we need real people with experience for.

15

Why not re-jigger the site to send all questions to ChatGPT first. That might weed out all the questions like, “I am trying to connect my PC to my router. How do I do that?”

1 Spice up
16

I’m tempted to write a regex exemption to that statement because regex. I get your sentiment I think we’re probably aiming at different targets here. Pulling a one off here or there on something like that vs full verbose replies generated by ChatGPT as well as doing that on a regular basis.

I stand by the statement of if you’re just being a front end for ChatGPT, you shouldn’t be here. In your example, that’s not the case.

17

I think “AI”-generated responses could be acceptable as long as they’re clearly labeled but I don’t know how helpful they will be on average (especially when people may be looking for a certain level of personal experience to back up such responses). ChatGPT and the like are great at scouring through voluminous amounts of content and giving a response based on their respective learning models but what they cannot do (and possibly may never be able to do) is to actually assess the validity of the information or give reasoned judgements about them. The way they are designed, the response they provide is likely the response currently deemed as optimal to a particular interrogative statement rather than providing the specific information that is being requested for a particular context. Those two things could wind up being the same response, but it is not necessarily such that they will be.

1 Spice up
18

But how u know I not a bot ??

Also how I know the OP (or person who posted) are not bots ??

Like @kevinhsieh ​ said…so many qn can get answers from Google, Bing or AltaVista ?? Maybe thats why I like to answer PowerShell Qns with “why not use this software,” coz literally 95% of me solving PS or Batch or Linux cmds daily is copying off Internet… even switches for robocopy…too many to remember.

Then if a person cannot take the lighter side of some replies and get that serious…bots can cause some mental injury…

19

Everyone googles stuff for code, but I would hope they aren’t just googling and pasting an answer without having an understanding of what they are pasting and knowing if it’s going to work or not.

3 Spice ups
20

AltaVista … lol! DogPile used to be a decent search engine before Google took over the world.

3 Spice ups