Should we allow AI generated text on this forum?

About This Forum
101

same for me in the post below.

Alan posted
“I edited” “beyond pasting relevant parts that illustrate my point.”

“If you read the full comments, I’m confident It will show that what I pasted is” “way misleading.”

See how that’s way different from your original text. Or even from proper quotes like below.

This is an exaggerated example. (i did 3 cuts vs alans 4 cuts. His where over a significantly larger piece of text though) but signifies how copy pasting some-one else their text and cutting things out can become very misleading.

102

<< It feels like arguing against a wall. Not to learn anything myself. Not to help or teach the wall. But just to warn other users to not run into it. (damage control instead of meaningful discussion.)

Sure human users of the forum also sometimes post mistaken information on the forum. But when you respond to that, you have a chance to help or teach another human. Or discover you where wrong and learn something yourself.

This is much more fulfilling than being an unpaid “error check” on AI generated text. >>

I don’t really have a horse in this race. I’ve used ChatGPT a fair amount. Here’re my take-aways:

  1. It is not infallible. It gathers information from on-line sources. Garbage in, garbage out. Moreover, it sometimes makes shit up. This is called “hallucination.” It its a well-recognized problem that may be solved eventually.

  2. It is fast and easy. The two answers I posted above (re whether AI text should/should not be allowed) were generated within seconds after I posed the question. If you want to do an internet search for information, ChatGPT is WAY faster and easier. In general, it gets me good answers in 3 minutes to questions that might have taken me 3 hours or 3 days. No ads!

  3. While the answers are not guaranteed to be correct, they seem to be better than the average of what you or I might find in a casual search. And if you suspect an issue and push back, the AI does not (unlike the average human) argue with you. It politely rechecks its answer and freely admits errors.

  4. ChatGPT seems extremely good at applying known problem-solving methods. For example, if I ask “How many grams of Calcium Chloride do I need to add to a 6.5 gallon batch of mead must to produce a solution that is 150 ppm Calcium?” it gives me a fast and accurate answer. Obviously this is a real world example – been there, done that.

This is still a work in progress. Output must be carefully curated to ensure accuracy and utility. IMO, the user is responsible. Any AI-generated text should be carefully checked. Any published AI-generated text should be clearly attributed.

One final note that will make clear my tentative view: We have all encountered novice growers who appoint themselves the resident experts on a topic. Maybe they like showing off their new knowledge. Maybe they like earning money from ad revenue on FB. If you choose to interact, it can be extremely time-consuming trying to protect other (especially newbie) growers from their misinformation. IMO, output from an AI is likely to be much more benign. If we’re not gonna screen out dumb, misguided humans then we shouldn’t prohibit properly attributed text from an AI.

p.s. This text was NOT generated by an AI. :slight_smile:

3 Likes
103

I agree.

But what do we do than, when users don’t carefully check but still post AI generated text?

What do we do when they don’t clearly attribute the AI generated text? (for example when they don’t mention the prompt. Or put the AI part of the post in quotes.)

some users of this forum have said they use AI to save time. And posted AI generated text on subjects they claim to have no experience or expertise on. How can we than trust them to “carefully check” the text their posting?

This is why i think it would be a good idea to have a forum policy either discouraging the use of AI generated text. Or describing how to go about posting it. (at a minimum mention the prompt) or both.

104

This is part of the problem. It is fast and easy to create large amounts of text. (sometimes filled with errors, sometimes not)

It is than slow and hard for human forum users with actual experience and knowledge. To correct the mistakes in the AI text.

I predict the forum members will grow tired of that quickly. And no longer try to correct anything from AI written texts. This leaves forum members without knowledge on the subject with no reliable way to filter AI texts for actually sound advice.

Another thing i wonder is, what added benefit does posting AI generated text on the forum bring?

People could just ask the AI themselves.

If i where to post all my google searches and corresponding hits that google gave me on the forum. People would comment on me to stop that.

Why should we not do the same for AI asked questions?

I’m not saying you should never use AI generated text tools. Just that you should not copy paste them to the forum. Unless you have the expertise to check them for validity. And properly attribute them. (including the prompt)

This is exactly an illustration of that.
You attributed the ai text including the prompt :heart: thats great!

But did you check the text for validity?
Did you notice 5 out of the 6 points mentioned in the 2e AI generated text. Where straw man arguments? And now that that has been pointed out. What should or shouldn’t you do about it?

If you would have had to spend the time to write that whole text yourself, would you have used the same (straw man) arguments? Or would you have come up with less arguments, or more to the point arguments?

It took you seconds to generate a large part of easy to read text. Full of straw man arguments. With text generative AI. And than took 100s to thousands times more time for a human forum user to point them out.

How does this not lead to a lowering of the quality of information on the forum?

1 Like
105

I’m assuming that the posts will be curated by humans. I definitely DO NOT want automated bots posting without human mediation.

With a human as a mediator, the use of AI-generated text would NOT lead to a lowering of the quality of information unless those humans were irresponsible. In that case, moderators can curtail their privileges. We have exactly the same problem sometimes with some human members who appear to measure their self-worth in words produced. Rather than attack AI, it makes more sense to attack the problem at the source. For example, we can limit each member to N posts per day and/or X words per post and/or Z cumulative words per day. And we can shut off an irresponsible user.

Anyway, other members made the key observation: You’re not going to be able to tell whether text was generated by a human or by an AI. So why fool ourselves?

No, but ONLY because forum members asked to see what the AI would produce. I felt obliged to show them verbatim without intervention or comment.

Note, I’m not expecting the AI to be infallible. We humans make plenty of mistakes too. I would never use it in the manner illustrated – making arguments for and against – except to stimulate my own thinking. Mainly I would use it to ascertain facts or perform calculations. And I would always subject answers to reasonable “quality control.”

3 Likes
106

your not always gonna be able to tell when some-one is lying. Does that mean lying is “okay” or should not be discouraged?

I’m not saying lying is 100% the same as posting AI generated text. (although posting AI generated text without attributing it, is a form of lying)

But this is more to illustrate that other hard to “deal” with problems exist. And that we try to deal with those anyway.

Exactly :sweat_smile: :wink:

Now imagine them being able to do that not by spending hours and hours typing and thinking up what to type. But just copy pasting AI generated text. Seems much worse to me…

I know, and it obviously was in jest. (first an AI generated text was made saying why AI generated text should not be posted on the forum. Then some one commented we should also let AI generate a text from the opposite viewpoint)

I don’t mean to suggest you did anything wrong.

But still it’s an interesting discussion.

I’m curious what you think the poster of AI generated text has as a responsibility in this case? (if the post was not made in jest)

107

So here’s the thing… If you don’t have the time and knowledge to answer someone’s question, then just don’t answer them. If someone wants an AI answer then they could get it faster without going through the forum. They come here to get first hand experience and perspectives.

6 Likes
108

As your definition states, a “straw man” argument refutes an argument not actually made. I don’t see any refutation, so I don’t get your point. Your observation that the AI answer referred not to AI-generated text but to other forms of AI seems a quibble. I’d expect the AI to use diverse resources to generate an answer. AI access to other forms of AI seems a plus.

I’m struggling to imagine the situation where i’d ask the AI to argue with me. For this reason, I’m struggling to imagine the situation where the AI chose to refute an argument that I never made. Can you show me specifically (1) what argument I made, then (2) what OTHER argument my AI refuted?

Let me use a real-world example. On another thread, a forum member asked whether wild leeks can be successfully transplanted. I answered yes but then cautioned that it might not be legal. I was trying to be helpful. I didn’t want to appear to encourage activity that might be illegal. So I used ChatGPT to look up the legality in VT, just as an example. The AI gave what I think is a helpful answer. I posted it, identifying the source as ChatGPT.

Was I not supposed to post the information about environmental regulations just because I didn’t know the answer before I used the AI to learn more?

109

Just for the sake of equality, and my own amusement.

Me:
Why shouldn’t AI-generated text be allowed on an online gardening forum?

ChatGPT:
AI-generated text, while often indistinguishable from human-generated text, lacks the direct experience and expertise that human contributors bring to a forum like an online gardening community. Here are a few reasons why AI-generated text might not be suitable for such a forum:

  • Lack of Authenticity: AI-generated text may lack the personal touch and authenticity that comes from real gardening enthusiasts sharing their experiences, tips, and advice. Human contributions are often based on personal experiences, trial and error, and genuine passion for gardening, which can be more valuable and relatable to other community members.

  • Limited Understanding: While AI models can generate text based on vast amounts of data, they may lack a deep understanding of the nuances of gardening, such as local climate conditions, soil types, pest management strategies, and plant-specific care requirements. Human contributors can provide contextually relevant information and personalized recommendations based on their firsthand knowledge and expertise.

  • Potential Misinformation: Without proper oversight and verification, AI-generated text may inadvertently disseminate inaccurate or misleading information, especially in a domain like gardening where precise advice is crucial for successful plant care. Human moderators can help ensure the accuracy and reliability of information shared within the community.

  • Community Engagement: One of the primary purposes of online forums is to foster community engagement and interaction among like-minded individuals. While AI-generated text can contribute to discussions, it may not actively engage with other members, respond to questions, or participate in meaningful dialogue in the same way that human contributors do.

Overall, while AI-generated text can be useful for generating ideas or providing supplementary information, it may not fully replace the value of human contributions in an online gardening forum. Encouraging authentic human participation helps maintain the quality, trustworthiness, and sense of community within the forum.

:slightly_smiling_face:

1 Like
110

Yes, if you don’t know what environmental regulations apply, then don’t pretend to. It’s fine to just admit that you suspect there may be some regulations and recommend that everyone look into local laws before proceeding. If AI actually helped you learn any relevant regulations, then you could have simply linked to the official regulations as published since they would be far more useful than any AI rephrasing of them.

By the way, ramps can be transplanted and can be sourced legally in one way or another regardless of local harvest regulations.

2 Likes
111

LOL – @Oscar, here’s a great example of a straw man, and it didn’t come from an AI.

@JohannsGarden – I didn’t pretend to know the regulations. I made it quite clear that I had used ChatGPT to learn about them. So you’re attacking me for something I didn’t do.

FWIW, I didn’t feel the obligation to find the link the the official regs. I figured that the forum member who asked the question could do some of the work. :slight_smile: At any rate, you have taken this conversation far away from AI and into basic forum etiquette. Personally I think it is perfectly reasonable to help when you can. The help does not have to be perfect. As they say, the perfect is the enemy off the good.

And yes, of course ramps can be sourced legally. But that’s cold comfort to a person who sources them illegally, which is the outcome I was hoping to avoid.

Why are you being so snotty?

112

I thought we were having a civilized conversation and I didn’t accuse you of not disclosing that you got your answer from ChatGPT. I am trying to convey to you that it is okay to not have an answer for everything. If you aren’t in the know about the regulations, then you’re not the one to speak on them unless you’re willing to familiarize with them first. While this may relate to forum etiquette, my point was about the use of AI as a substitute for people actually processing matters with their own mind.

4 Likes
113

The argument that your AI text is refuting, is that “AI generated text should not be allowed on the forum” It is doing this by your prompt off

And the AI text than gives arguments for why AI generated text should be allowed on the forum. But the arguments are not about AI generated text. But other forms of AI.

It is arguing about something else than the argument that’s actually under discussion. (AI generated text is the actual subject. Not all AI in general)
Hence straw man argument.

This might seem to you like a quibble. But it’s an important distinction.
It is what makes it a straw man argument. (the arguments subject not being the subject that’s under discussion)

an analogy would be the following.

Subject under discussion.
“you should eat less fries, because they are unhealthy for you”

refuting that statement by saying. “research has shown that broccoli is healthy and good for you”

Fries (potato) and broccoli are both vegetables. But their not the same.
Just like Text generating AI is not the same as translation AI or the AI in your thermostat.

you could also say that there is an (assumed) false equivalency.
(since translation AI is good for the forum. Than text generating AI is also good for the forum)

  1. you did not make an argument. Your whole post apart from “me” and the question prompt was AI written. The arguments where from the AI written text.

  2. The AI generated text tried to refute the argument of " AI generated text should not be allowed on an online gardening forum"
    (it is giving arguments for why it should be allowed)

This is an example of the subject being “AI generated text” But the argument for using “AI generated text on the forum” is arguing about using “translation AI” It is acting as if the subject is different than what it actually is.

1 Like
114

Now it’s your turn to explain why it is a straw man argument. (and what the argument is. because it is not clear to me)
Where is the example?

disclaimer: I don’t know the post where you posted the AI generated text. All i know about it is what you said in this topic. (maybe link the post to avoid confusion/misunderstandings in this discussion?)

If you didn’t know the regulations. And did not have the expertise or wanted to do the work of getting the expertise to check the validity of the AI generated text you posted. Why than post the AI generated text?

Wouldn’t it be better? to just say : " I’m not sure it is legal to harvest them from the wild though. Check for yourself"

By talking about the regulations and posting AI generated text (even with disclaimer) it can suggest to people that that information is trustworthy.

And if the information is not trustworthy, does it really help them to post it?

The problem is that the AI generated text tends to seem really well written. Like it was written by a trustworthy expert. While sometimes being complete BS.

I think your intentions where/are good. But posting AI generated text can be misleading. Even if you don’t intend it to be.

Did you post the prompt that you gave the AI? Knowing the prompt of the AI is essential to judging the accuracy of the AI generated text.

The AI generated text can be different from the following prompts.

“how to harvest ramps legally?”
“is it legal to harvest ramps?”

The AI generated text tends to try and comply with the prompt. So trying to comply with the “how to harvest” part of the prompt could give different text from the
“is it legal” part of the other prompt

An excellent example of the AI generated text being complete bullshit because it tried to comply with the prompt.

Is given in this post of this topic. (click the :arrow_up: in the quote/link to see the post)

Since you brought up forum etiquette. Could you remove ?

115

We are going to temporarily close this thread until things cool down. Please remember to focus on facts/issues, not the personalities here you are interacting with. Thanks!

117

What about generating text as a human who consumed AI generated text as a secondary source?

What about AI as a fact/grammar checker or reformatter to make the posts more concise or readily consumable?

What about AI generated ASCII art?

1 Like
118

You raise a very interesting point here.

If you remember that the source is AI text. You should attribute it. (i read in an AI text bla bla bla. It was from a promp: …)

And like always be honest what your actual experience/knowledge is.

Don’t use AI generated text to misrepresent your actual experience and knowledge.
would be an excellent and short point to keep in mind when using AI

However at some point things you read in AI generate text will seem like common knowledge to you. I am curious what AI generated “hallucinations” or falsehoods will be most wide spread in society in a few years.

AI grammar checker is a different form of AI. Not really the subject of discussion in this topic. (just like AI in your thermostat is not the subject here)
AI grammer checker to me seems fine.

AI as a fact checker is an interesting point.
since AI still makes a lot of mistakes. Im not sure how useful it would be as a fact checker…

I think if you don’t know enough about the subject to contribute without the need for AI generated text or fact checking. It is better to not post.

Or be very honest about your limited knowledge. The problem is that AI generated text, can make it seem like your an expert while actually being full of errors.

Reformatter is an interesting point again.
Personally i think it depends on the degree it will reformat. If it suggest to me a better sentence structure and i can change that. (just like spellcheck suggest the correct spelling)
i have little problems with it.

If it reformats or adds so many words that it changes the meaning, your pretty close or already at AI generated text. (just with a larger/more complex prompt)

I think a lot of AI generated content violates

Post Only Your Own Stuff

You may not post anything digital that belongs to someone else without permission.

Since it’s a regurgitation of information on the internet. Where the original writers did not give premission. I’m sure there are lawsuits about this right now. I’m curious how those will turn out.

i love how your mind works. You thought of really interesting “cases”!

Although AI generated ASCII does contain letters etc. It’s not “text” It is closer to an AI generated image. I don’t think it would be a problem on the forum. I also think AI generated text for art or amusement purposes in the lounge would not be a problem. (bring on the AI generated fruit growing meme’s or short stories)

3 Likes
119

Votes ‘other’ to see the results.
Personally, I have no idea. I wish AI could grow my fruit!

2 Likes
120

I agree with your points.

I think referencing AI generated text may be similar to existing posts that may start with something like, “Another person I knew had an experience with XYZ, and they foo’ed when they bar’ed”

I don’t have an issue someone giving me info from AI. Perhaps the AI’s words might be of higher quality, relevance, or truthfulness than from a real person. I’d rather have the truth and be able to understand more effectively as someone learning all the time.

2 Likes