Any chance we could introduce a forum rule about AI content

You claim lack of time but you took the time to go to an AI. Why wouldn’t you write a reply in your own words? If you, the expert, can answer the question in your own words then you wouldn’t need an AI to begin with.
I’m certainly not going to give money to an AI system.

I think I could agree with a rule saying a disclaimer must be added when the text is from an AI. If lacking and found out then the post gets removed altogether.

Um Bongo’s initial post already has 18 likes in 5 hrs. I’d fully support a complete ban. Much energy has already been wasted trying to correct incorrect AI posts.

@AbramBaileyAuD said:

I query AI, get decent information (I know because I am an expert in a couple of fields, like audiology and web development)

I think this is key. AI can be useful if you can discern its output.

I’d be OK with a rule where ‘registered providers’ (and @Bimodal_user) are allowed to post AI here, but the rest of us plebs are not.

Not sure what you mean … Simply put, AI is just better at thinking of all the nuances that it’s easy (even as an expert) to forget in the moment.. That’s why you see doctor’s utilizing LLMs in practice… you can be an expert and still benefit from the comprehensiveness and speed of an AI.. but it’s helpful to be able to quickly read the content and know, as an expert, whether it’s right or wrong… That whole process just takes far less time, and tends to be more accurate, than me trying to pull out a comprehensive cited explanation myself.

Guys, - whether you like it or not AI is not going to go away… In fact over time it is, and will be, getting better. Yes it produces wrong answers at present - but they are becoming less and less.

I agree that pasting many tens (if not hundreds) of lines from AI is not useful - however posting the AI prompt you made might be useful for people to see the response themselves - and maybe improve upon it.

AI mistakes are no worse than misinformed humans that post rubbish unintentionally - just recall a whole slew of posts about oxygen being necessary for Spheres….

Having a rule will not change this behavior whether from AI or not… How many people read and adhere to forum rules?

@Raudrive

I agree that he has a way with words! Well, he is from Wales, and Dylan Thomas comes to mind. You can’t get a better “lineage” than that!

On the AI content—ugh!

For an analogy, look at self driving cars (real ones with LiDAR)… they still mess up, but at a rate that is far less than mere humans

Once the models get really good, humans will only be able to add NEW information to the dialogue. New experiences, thoughts, feelings, creative ideas, etc… All of which will still have a value and see forums like these still remain relevant… But as far as I’m concerned, AI will soon surpass us as a source of concise / factual information. Some would argue that it already has (model dependent)

PS. That’s just my personal opinion, but ultimately want to steer the forum in the direction that feels right to the community here.

That’s not a good analogy. You seem to be saying self-driving car messes up and human takes over to avoid a crash. Applying that here, that mistake should then be reported (posted) via the humans so that more cars error. But the human correction is what needs to be reported and it’s a lot of work for humans if the former is what occurs because then every car errors until some human is distracted and there is a crash. At least in the car scenario humans can save some effort while the car drives. Forums do not have that task offloading process.

What people here are saying is they do not want the former. They want this to remain a source of human supplied truth even if there are sometimes human errors. If the human has nothing to add without AI then just read and lurk. They do not want to have to process or correct AI. We do not have the ability to make it clear to AI which fact is correct. And they do not want the AI reinforcing its errors and degrading the site information. Those who want AI answers can get them elsewhere. If humans keep driving the car there is not the battle of how to manage the AI errors.

Whether things improve longer term doesn’t seem to inform the present. The site will be scraped and we’re training the AI but that is passive and does not require increased workload.

People likely will not listen anyway so a ban will be only a partial block. What are other niche forums doing?

@user490 said:

People likely will not listen anyway so a ban will be only a partial block. What are other niche forums doing?

Personally, don’t really care, but as I said above, I’ll go with the results of the poll.

I didn’t follow your explanation about the self driving car analogy.

@AbramBaileyAuD, I like your jab at Tesla

I just wanted to add that there are many credible experts that say that current LLM technology is ultimately a dead end. I think the jury is still out whether current models will progress much further.

And even if they do, the rate of progress seems to be slowing down already, while the costs are rising exponentially. Kinda like chip development…

Thanks for posting the survey!

I fully agree with you. Thumbs up!

My concern lies more with individuals who pose as experts in conversations by copying and pasting responses from AI tools without verifying their accuracy and, like you said, feed into further chaos.

These individuals should, at least, provide a disclaimer: “I’m not an expert, I don’t know the answer so I asked ChatGPT, and this is the response I received. However, I haven’t validated this response, so I can’t confirm its accuracy.”

The hole in this strategy is of course that it’s an honor system and ultimately there’s no way to catch all AI generated content. People will likely continue posting it—without any disclaimer in the case of a total ban.

People also post factually incorrect material they really believe they “know” to be true. (I’ve done it, and was reasonably gently corrected.) There isn’t much to prevent that either. I wonder how AI will be able to bypass these kinds of things.

WH

@Hooby said:

post rubbish unintentionally - just recall a whole slew of posts about oxygen being necessary for Spheres….

I remember that.

It went on and on. They kept justifying it was correct because blah blah AI said so.