Skip to content
HearingTracker
Log in

Any chance we could introduce a forum rule about AI content

This topic is closed.

69 replies1.4k viewsLast reply

Jump
Link to this topic

I agree with @AbramBaileyAuD’s view. You guys are a surprising bunch of Luddites. Perhaps you should carry ICE cards in your wallets or purses (ICE formerly stood for IN CASE of EMERGENCY), that say, “In an EMERGENCY, I do not want to be treated with any methodology using or derived from because it may be wrong, and besides that, I just plain don’t like AI.” Put your money where your mouths are, so to speak. When you go to your , tell them you don’t want any fitting recommendations derived from AI, for sure. Otherwise, there’s a big disconnect between your forum view and your own personal hearing care….

This forum is Internet content, and like much of the Internet, it has some forum posts by humans that are just plain wrong, subjectively biased, or emotionally over the top. Hey, but that’s OK. We’re humans, and to err is human. But AI can err?! Holy Cow! That really detracts terribly compared to the plain wrong, subjectively biased, and emotionally over-the-top pure-human-content posts you deal with as you wish, being an inevitable part of the Internet. One is so skilled at dealing with various human posts, but unable to ignore or hide posts or posters using AI you don’t like?

As someone else noted, I think AI is here to stay. One is going to have to grow up and deal with it, if not on this forum, elsewhere.

Are you afraid to use research to help you understand how the technology that helps you hear better works? Seriously!. Learn about what helps you and make certain your devices are programmed correctly.

As I said to @Um_bongo, I’m fully aware of the data corruption problem when the slop becomes the training data. No need for more analogies, but thank you :slight_smile:

In reply to @jim_lewis

I don’t think they are luddites, and let’s not name call, please. There’s plenty of reason to be suspicious of , and we should all take AI content with a mountain of grains of salt :slight_smile:
I personally think the forum should be 99% human generated content, because hey, we can all just ask AI at the end of the day … it’s not really providing much value to the community or the people that visit the forum for advice to provide AI answers. They are here for your views and experiences… so I’m on the fence about this one, and if I had it my way, there’d be some flexibility to handle the grey area. However, if most of the people here are against AI answers making even rare appearances in the forum, then I can easily side with that…

In reply to @jim_lewis

Division and fears don’t help any of us hear better. There is no 1 way to learn, and nothing is perfect. If I have a genetic problem in an ear canal who is anyone to judge that oxygen may not help me hear better?
This forum is full of help. I appreciate all that I learn and use a lot of it. If participants in the forum prefer trauma and fatigue well good luck to you. Thanks to all that help for the sake of hearing.

And yes has value but will never replace human value. Get over your fears and just help here.

Edited

In reply to @jim_lewis

It’s not about being a Luddite, Jim, I use weekly for content creation, imagery and advertising.

It’s about the loss of data quality by iterative destruction of source information.

I’m an administrator on an Anti-Reform group of 90,000 in the U.K. - we have a huge problem with people posting emotive AI content because they use AI like an obedient favourite pupil, or more correctly the AI is taught to reproduce more of the results that the ‘client’ prefers - factual or not.

So the weakness in the model is that the verification loop is policed by humans with inherent biases. I note it’s particularly an issue amongst the idiotic denial of climate change: as I sit here in February in a pair of shorts……

Hi, @Um_bongo

I have to hit the sack, but I asked ChatGPT 5.2 for its opinion of “loss of data quality by iterative destruction of source information.” It gave me a much longer reply, but I asked it to condense it down to one paragraph. I’ll see if I can attach the longer reply as a PDF:

Response from : “I think the concern about low-quality information propagating on the Internet is valid, but that’s fundamentally a human curation problem, not an AI-specific one. AI doesn’t continuously retrain itself by “photocopying” forum posts; what actually matters is how people use it. When prompted casually, AI may summarize consensus-looking noise, but when explicitly constrained to authoritative sources—peer-reviewed audiology journals, professional standards bodies, documentation, or manufacturer white papers—it can just as easily reduce the influence of fringe opinions and anecdote. The real risk isn’t AI attribution; it’s uncritical copying without sourcing. Ironically, banning AI citations doesn’t improve information quality—it removes transparency and makes it harder to evaluate where claims came from in the first place.”

ChatGPT Response to Recycling AI.pdf (93.7 KB)

As the author of the post in question that seems to have prompted this topic’s discussion, I feel I have a responsibility to respond as there are several conflating issues here.

Issue One: Lengthy Posts. That is entirely my fault. My brain is wired to be very detailed, technical and very thorough in ensuring I understand all aspects of my research. So, I have always struggled to get to the point because my brain goes 100 miles an hour with so many thoughts. I have learnt to be much better with clients by giving them a high level overview, but when it comes to technical discussions like my original topic, I struggle to condense. I do humbly apologise unreservedly - but that is not the fault of , that is my fault! My brain is just not wired for short and sweet and I know it irritates people when I write novels.

Issue Two: The post in question was never intended to be an in depth discussion of my specific hearing loss. It was actually intended to serve the purpose of me providing a review of a new audiologist that I will be seeing in a couple of weeks time and second to get feedback from more experienced people with regards to some of the services that she offers, that I have never heard of and which 99% of audios don’t seem to provide as a service. The topic took a life of it’s own and ended up being an in depth discussion of my hearing loss, specifically why I struggle to hear soft speech properly and no audio has ever been able to resolve it properly. Again, not an AI issue - because the topic started going down a rabbit hole long before I introduced AI into the topic.

Issue Three: I was the one that introduced AI into the topic in question and this is very unusual for me. I don’t bring in AI, because I (as an IT Technician) am more aware than most people that the computer does not know best! I will never irritate people by saying to experts but Chatgpt says you are wrong - because I know Chatgpt will never be as smart as a human - because it cannot understand nuance without life experience. I specifically introduced AI in that topic to discuss a very specific point regarding one of the services and how it would help, because @Neville one of the people in the topic said that he could not see the value of that specific service - so I copied and pasted what Chatgpt said to broaden the discussion on that point - but I did make it clear that I was by no means an expert in hearing - I just wear hearing aids! It is unusual for me to introduce AI into any discussion (in fact I don’t think I have ever done it before) and it was done only to broaden discussion on one of the hearing services the new audio is doing.

Conclusion

That topic ended up going down a rabbit and I myself realised it, which is why I encouraged ending the discussion because while interesting, it was taking up a lot of people’s time. I am going to see a brand new audiologist who will hopefully be able to scientifically explain I have the specific hearing issues that I do, which made the discussion of my hearing a bit moot prior to the visit as I want to give the new audio a proper chance! I realised that I was annoying some people with lengthy posts, so I gracefully let that topic come to a close because it was going down a massive rabbit hole - long past Alice in Wonderland!

I am going to vote No on the AI Poll - We should not allow AI into forums (even though I “broke” that rule) because AI is not ready to be used as a reliable source. There is far too much AI slop on the Internet ready. The single biggest problem with AI is poor quality data in results in poor quality data out. In addition, AI really struggles with nuance because it has no life experience. While I have used AI extensively in my discussions about my hearing loss amongst other things - I know that I can’t trust the accuracy of that data. I use it more to gain insights not to be able to say to the audio - do this, do that!

Let’s ban AI on the forum until the quality of it’s data has improved to the point that it is equivalent to a trained professional. It may never reach that point, because AI will never understand nuance as that requires life experience which is one thing that AI can never have obviously. Thanks

Maybe I will AI to condense my posts :grinning_face:

For me there are several issues. The biggest is when people quote an response (often huge) and then insist that it’s a reliable source because it’s a premium AI that they pay good money for and of course it is blatantly wrong.

There’s also the issue of AI training itself on AI responses which leads to GIGO.

A ban on quoting AI doesn’t mean you can’t still use AI heavily, but run the results through one’s brain first and put it in your own words instead of deluging us with a huge quote. Preferable would be that you’d also try to do some verification. I know that a simple ban paints with too broad of a brush but trying to specify which uses are appopriate gets tricky fast. I haven’t seen one AI post on here that I think would be a great loss to the forum if it wasn’t but many that I think are detrimental. That’s my frame of mind about the issue.

The main problem here is that people expect god-like capabilities from a new tech whose development will eat up some 500 billion USD this year (projected). What we are looking at is a big bubble for a modestly useful new tool. Like I said, is handy for rapidly retrieving technical information (like: “what kind of processor sits in an 11th gen iPad?”). Problems start when people also use it to ask if Phonak is better than Oticon, for instance. Not helping there. Overall, I don’t see it as a real threat for a forum like this.

Maybe put the ban in for a period of time, and reexamine this issue at the end? Look at the issue once a year until we decide it is in our best interests to like before it kills us all? :wink:

WH

In reply to @AbramBaileyAuD

I think an summation of information with a link to the source information would be very nice addition, but the forum software needs to be changed to have an operation for adding AI information to any post, and it would – in a stanardized format – surround paragraph(s) on all sides in a box formed by a line that has this flag
—– AI ——

Posting AI material without using this format would be verboten. It could be added as a new option surfaced by the circled plus sign in the header of the standard input box. This would allow AI summaries to be identified, and also identify the source in a standard way for all uses of it.

Active discussions

What the forum has been talking about this week.