This topic is closed.
58 replies1.4k viewsLast reply
I voted NO but as others have mentioned, it’s so widely used by search engines. I do not trust the source of some data because data can be manipulated and influenced by funding criteria.
Great point!
Doesn’t stand for “All Income”? (Or “All Idiots” in some things I’ve seen by AI)
Paul
You’re ignoring this member’s posts.
I like the others’ suggestions of leaving available, but only if it is clearly indicated that it is AI generated information. Banning it would be like returning to the Book Burning days.
Sometimes what AI produces provides a spark of idea for the reader to further explore from more credible sources. (Internet credible? Sometimes, sometimes not. Due diligence is needed.)
On some television ads I’ve seen with people who look completely real have a disclaimer at the bottom stating that the ad is “Performed by AI”. If you ever see an ad for Frndly TV channel with a farmer talking about his cable bill and a woman holding a baby who talks; those are AI “people”. They look real to me!
Or look up Tilly Norwood. She’s fake, but looks real.
Paul
You’re ignoring this member’s posts.
I don’t think it should be completely banned, but I would like people to get away from copying/pasting entire AI-generated responses. I’d rather hear the posters take on the AI output and maybe even how their question/search was posed …
Wise words. They highlight three key issues:
You’re ignoring this member’s posts.
In a PM discussion several years ago with @AbramBaileyAuD, I pointed out to him that incorporating skillful and useful into this forum might help it survive in the long term, rather than outright banning AI for, IMHO, questionable reasons. In initial posts about the current discussion, Abram stated that, under the right conditions and with the right AI, he himself found it useful.
I’m making this post now because I read yesterday that the Washington Post laid off 30% of its workforce, including the entire Sports section. The reasons cited were increased competition and the use of AI by would-be customers, instead of buying a Post subscription. That shocked me.
So, I would predict that, as time goes by, Abram’s website (the Hearing Tracker product review and blog article part) and the Hearing Tracker forum part, are going to suffer the same fate. If Phonak, ReSound, Oticon, Starkey, etc., offer elite, restricted-content AI advisors, tech support, etc., via AI, traffic to Hearing Tracker and this forum may be severely reduced unless it offers something competitive in return. For example, since Google is a search company, Phonak might prefer to work with Google on increasing its Search presence and its Gemini AI presence in the future, rather than buying an ad on the main Hearing Tracker website, where Abram will presumably follow through on his promise and ban AI there, too, so it doesn’t taint human-generated content.
Most of the opinions against AI sound poorly experienced. Yes, AI that relies on Internet content can be quite bad, since the Internet is a big slop pit of human-generated, ill-informed and ill-curated opinions. But AI restricted to authoritative content can reach a Ph.D.-level of knowledge, which goes far beyond the average level of participation on this forum.
So, I’d predict that if Hearing Tracker bans AI completely, the site will eventually be riding off into the sunset, just like the Washington Post.
If Abram really follows through on his promise to ban AI, to do so, he’s going to have to remove the AI-generated thread summaries, too. According to the detractors, you shouldn’t really rely on that stuff, either, because as @MDB asserts, AI is crazy-uncle stuff and he hasn’t seen a single worthwhile AI thingy on Hearing Tracker.
Hey! Just like the non-GMO logo on foods, we can have a non-AI logo for all pages of Hearing Tracker, and blog authors can sign a disclosure that “No AI was used in the generation of this article.”
You’re ignoring this member’s posts.
I’m making this post now because I read yesterday that the Washington Post laid off 30% of its workforce, including the entire Sports section. The reasons cited were increased competition and the use of AI by would-be customers, instead of buying a Post subscription.
The answer shouldn’t be to lower the quality of information here.
People will come here if they can get better answers.
Yes, AI that relies on Internet content can be quite bad, since the Internet is a big slop pit of human-generated, ill-informed and ill-curated opinions. But AI restricted to authoritative content can reach a Ph.D.-level of knowledge, which goes far beyond the average level of participation on this forum.
Exactly the results posted here generally fall into the first category. How would you propose to allow only an “authoritive” AI?
No-one wants to come here to see AI results copied and pasted as an answer. They can do that own there own. The point is that the forum has to add value over AI generated content. Answers from real users and audiologists help do that. Copied and pasted AI answers don’t unfortunately.
You’re ignoring this member’s posts.
@coastal_california Your argument rests on the assumption that the user will always get a better answer from the forum than they would, say, from a dedicated Phonak agent grounded in Phonak technology, all its hearing aid models, Roger performance specs, etc. I’m saying Phonak or another website could potentially provide that much more directly, rapidly, and specifically to an inquiring user than having to wait for and wade through a bunch of forum replies to a query and trying to decide, especially if you’re a newbie to the forum, who the true experts replying to your query really are.
The Washington Post’s decline is very complicated and multifaceted, but, to be a little facetious, why subscribe to the Post for the Sports section and game analysis when you can go online and search or use an AI agent to get a custom-tailored analysis of just your games of interest in seconds? I did it all the time in the past fall college football season. Updated betting odds, too, for games in progress, etc. Illustrates why users turn to AI rather than a website.
To give another more personal example, I had a homeowner’s question about our current Code and Covenant Restrictions (CCRs - I live in a gated HOA). I fed ChatGPT 5.2, our whole CCR document, 4 MB, IIRC. It did an excellent job answering questions about the content and the legal implications of the 29-page document. As a former board member and someone who’s read extensively from texas.gov regulations and Cagle’s 4th edition of Texas Homeowners Association Law (as well as our entire CCR several times over), ChatGPT did an amazing job of explaining our newly revised CCR relative to Texas law. It was NOT dealing with source material written on the Internet by “hallucinating” humans. It involved a grounded document written by our highly paid HOA lawyers and a comparison of sections to cited Texas case law.
So, with the right source material, an AI query can rapidly return fairly accurate, not always infallible information about a topic, particularly if the query is narrowly grounded on authoritative source material. You get much faster instant “gratification” than the scattershot approach of looking for an answer on this forum and hoping a true expert will tune in to your query “real soon now.”
Part of the problem here might be not so much that AI is always inherently bad, but some of the anti-AI folks just never learned how to use it to best advantage. We had $30K of contracting work done on our driveway because of degradation by live oak roots. The AI agent I used obtained construction data and costs for San Antonio from Angie’s list. I could design and revise my driveway on the fly, change tensile strength, concrete thickness, amount of pebbling, rebar type and layout, etc. It would update the projected cost in seconds and was right on the money with its the bid estimates that we ended up getting from a bargain basement contractor vs. a middle-of-the road, and a premium contractor. The things AI told me about driveway construction jived very well with advice from the contractors who showed up to give their job estimates.
So, the information on and hearing loss is probably out there to make a good, focused, grounded AI advice site. But it looks like Hearing Tracker won’t be that site.
You’re ignoring this member’s posts.
Your argument rests on the assumption that the user will always get a better answer from the forum than they would, say, from a dedicated Phonak AI agent grounded in Phonak technology,
No - as I said.
How would you propose to allow only an “authoritive” AI?
So, with the right source material, an AI query can rapidly return fairly accurate, not always infallible information about a topic, particularly if the query is narrowly grounded on authoritative source material. You get much faster instant “gratification” than the scattershot approach of looking for an answer on this forum and hoping a true expert will tune in to your query “real soon now.”
If that was true (but its not with the typical question asked here) just use the and not this forum.
The point is the forum has to add value to exist (and I think it does) without needing people to cut and paste AI answers here
You’re ignoring this member’s posts.
@coastal_california What I’m saying is, rather than just totally banning and saying that’s the best, there’s a better path forward for the forum. There are ways to ground queries in authoritative sources. You could tell the AI to answer a query strictly from Harvey Dillon’s book Hearing Aids or from an introductory college-level audiology text, for example (although Dillon is too antiquated now for many queries).
I’m just saying that clinging to the past isn’t going to work, IMHO. It didn’t work for the Washington Post.
AI has incredible power, but it’s still in the Model-T days. Failing to find a use for that power on your website will come at a cost.
If anyone needs an example of good everyday use, go to Amazon and read the AI-generated summaries of product reviews. Use the Rufus AI agent to ask comparative questions about products. It works amazingly well. Although Amazon owns Ring, the Rufus agent gives better answers about Ring products than the ring.com AI agent, which is a moron.
The Hearing Tracker website could have more accessible, ongoing AI summaries of threads, like Amazon review summaries, so you don’t have to wait eons to get the overall gist of a thread. Abram’s Hearing Tracker reviews could use an AI agent to answer comparative questions about HA models and brands, just like Amazon’s Rufus AI agent. The problem is that all these things cost $$$ to design and run, so Abram’s budget may not make my suggestions affordable.
Another potential use of AI on HT: ask the HT search engine to “Please find posts that best answer this specific question,” rather than getting a long list of hits that kinda match your search term but don’t really provide the answer you’re looking for and, thus, waste your time.
There are quite a few people, including AI enthusiasts, who think AI is way overextended and that the bubble will burst, with devastating economic consequences worse than the dot-com bust in the 2000s. IIRC, Amazon just announced it’s going to spend $200 billion on AI in the coming year, and its stock dropped 10% (Amazon is the world’s leading provider of web services in AWS).
You’re ignoring this member’s posts.
There are quite a few people, including AI enthusiasts, who think AI is way overextended and that the bubble will burst,
Yes this unfortunately is 2001 all over again.
You just have to drive up 101 from San Jose to San Francisco and see all the billboards advertising the latest and greatest Agent which can all do similar things but just better than the other guy…
You’re ignoring this member’s posts.
OK. Here’s are several ways you can use to better get information from the HT site, and it won’t show and annoy people who don’t like AI, and can still be used even if AI were banned from the site. The following is all intended as a demonstration that it’s short-sighted to ban AI use and impractical to enforce anyway because people can disguise the source of their information.
First, @AbramBaileyAuD tried to provide AI summaries of threads, but it doesn’t work at any reasonable speed for me. Instead, you can provide a link to a thread to ChatGPT 5.2 (or similarly powerful agent like Gemini) and tell it how long a summary you want: 300 words? 1000 words? It will oblige in seconds, and if you decide the thread summary is not long enough, just tell it to expand the summary to whatever you want, 1000 words, 2000 words, etc.
Example of a 500-word summary of this thread up to this post:
Query: Can you summarize the following thread and note whether you think any particular posts by me (jim_lewis) are off-topic? (a minor consideration). How about a 500-word summary of thread, with the off-topic commentary being extra?
Example of a HT Thread Summary by AI.pdf (108.8 KB)
Secondly, the HT Discourse forum search engine is very powerful, but it can only do a syntactical search for words that match your search criteria. It can’t search for meaning. You can have an AI agent wade through search hits or wade through all the posts in a thread to find the post(s) that best match, in its “opinion,” what you’re looking for. Like the previous “summarize this thread” suggestion, only you will see, deal with, and judge the results.
As an example, ChatGPT suggested you could use the following template to ground the AI results returned.
Test AI Query Search of HT Forum.pdf (142.9 KB)
Find HearingTracker forum posts that address: (your question)
Then: (stipulations to be included in your query, as per your preference)
• Prefer technically detailed posts
• Prefer posts consistent with mainstream audiology textbook models (e.g., Dillon-style reasoning)
• Quote key lines if helpful
• Flag disagreement between users
You can obviously make up your own set of criteria and tune it as you go. If you find you’re heading in the right direction and already know a good deal about hearing aids and hearing loss, you may be better than the AI agent at further narrowing your search down amongst the best hits it could find. If you feed the agent specific thread links that you think might answer your question, that will work better than asking it to search the whole forum. You can chat with it to tune your approach.
As an example of how finding the best posts explaining something can work, I’ve attached the output to my query, “Find HearingTracker forum posts (and threads) that address: (the pros and cons of an open fitting and what sort of hearing loss it works best for)”
An AI agent does not have access to the number of LIKES a post or a thread gets, nor can it count the number of links to a particular post or thread. An AI agent that had API access to the HT Discourse website could probably do a better job of finding the “best” hits by considering number of LIKES or number of links to a post or thread.
You’re ignoring this member’s posts.
I’m here to read other users’ experiences. The value of this forum lies in learning from others. I came here to get away from manufacturers’ bells and whistles. is very useful, and I can do my own AI research. AI-generated content should be flagged or tagged as AI.
You’re ignoring this member’s posts.
I know very little about but when I hear about it I always think of the movie Terminator. What I like about this forum is the human interaction. I’ve learned a lot here and like reading from the regulars I see who post here.
You’re ignoring this member’s posts.
This seemed pertinent to the discussion. Not sure it paywall will be involved. In short the gist is that people come to forums to ask questions to get an answer from a human. The Tech Friend from The Washington Post8
You’re ignoring this member’s posts.
This seemed pertinent to the discussion.
OTH, getting an opinion from a Washington Post tech journalist who just lost her column slot because of (in part) is not likely to get you an unbiased view:
The Tech Friend newsletter will be on hiatus after today’s edition.
I think the columnist overlooks the fact that if you have crazy, foolish, or stupid things you’d like to ask, AI is not the special, unique place where you can do this. The problem with the Internet in general is that there are plenty of places for humans to do this all by themselves. Places where humans may foolishly encourage or outrageously abuse each other, even encourage someone to commit suicide.
But we’re making progress in 2026. We’ve moved on from the turn-of-the-century view that the Internet is the end of the world to now, AI is the end of the world.
You’re ignoring this member’s posts.
In my opinion Ai and Ai Agents should not be banned but if quoted, source material should always be referenced and Ai posts should be identified clearly as such. I use the HT forum for personal experience of forum members and speaking for myself the community has helped me immeasurably since my first ’s. My may be similar to someone else but how we process hearing with our respective HA’s may be different. How the HT site is able to integrate Ai tools could influence how we interact with each other and our respective experiences.
You’re ignoring this member’s posts.
And just for fun, I configured gemini 3 pro as our summary model and asked it to summarize this thread…
You’re ignoring this member’s posts.
How the HT site is able to integrate Ai tools could influence how we interact with each other and our respective experiences.
The use of in the background to search the forum might help you find past posts by others who’ve had similar experiences to you and find the similarities or differences they experienced relative to your hearing loss. If Abram’s budget allowed, he could incorporate AI-enabled Search in the background (as an option), but as I pointed out earlier in this thread, a user can employ a powerful agent like ChatGPT, Gemini, or ? to “read” the forum and look for posts that match the semantic intent of your search. Such searches are private and don’t expose any other members to AI. Sort of an “AI can help you find humans who match up to your experience” sort of thing:
OK. Here’s are several ways you can use AI to better get information from the HT site, and it won’t show and annoy people who don’t like AI, and can still be used even if AI were banned from the site. The following is all intended as a demonstration that it’s short-sighted to ban AI use and impractical to enforce anyway because people can disguise the source of their information. First, @AbramBaileyAuD tried to provide AI summaries of threads, but it doesn’t work at any reasonable speed for …
Disclaimer: I use a paid $20/mo ChatGPT Plus subscription. The AI can digest much more “context,” i.e., forum posts and threads, in one go than the more limited free version of ChatGPT. So, I may get better search results in one go than a non-subscriber would. ChatGPT thinks its free version would work well enough, but a user might have to use “chunking.” So, more steps, more thinking and typing are involved.
When searching across threads, chunking usually means:
Break the search space into manageable sets
i.e., if there are several different things you want to match up to other users, make each characteristic a separate search set to begin with.
You’re ignoring this member’s posts.
And yet the apparently can’t make use of the poll numbers. Poll numbers are currently 80 to 37. In any kind of election that would truly be seen as a landslide.
You’re ignoring this member’s posts.
Edit_Update: I see. The button with the icon
at the bottom of the thread is now the Gemini “Summarize This Thread” button.
Where does “our summary model” show up on the forum? I noticed last night that the “Summarize This Thread” button disappeared from the bottom of the thread and now I see a Top Replies button instead, which is good because it allows you to more easily find most-liked posts in a thread. But why not have both options at the bottom of a thread?
Also, relative to any AI results on the forum, I think it’s good for the poster to provide the query that generated the results. It helps readers learn how to or how not to use AI in searches. For instance, your Gemini results cite user names and their individual opinions, whereas a ChatGPT summary that I did grouped users into categories and summarized as a group, the prevalent opinions in each category. So, I’m interested in why Gemini behaved differently from ChatGPT.
You’re ignoring this member’s posts.
Thanks all. Going to lock this post for the time being, take some time to digest, and come back with a solution. As promised I will honor the poll 
Edit: I would appreciate if we could all move on from this topic for the time being as I know some of you would like to get back to the primary focus of this hearing aid forum!
PS. Thanks for all the helpful suggestions.
You’re ignoring this member’s posts.
Forum Support69 replies
Fitting Software Requests1.1k replies
Fitting Software Requests513 replies
Fitting Software Requests419 replies
Hearing Aid Self-Fitting (DIY)301 replies
Auckland, New Zealand
Colorado