I had an interesting conversation with chatgpt 4 about tuning my hearing aid going by my Audiogram. Very detailed instruction on what to tweak and why. I only need Genie 2 download now so I can try the recommended settings.
Of course I will save my current settings before doing so. Any way, really keen to give it a try and when I do Ill report back.
I have had chats with ChatGPT and Grok3 with wildly differing results. I had Grok3 recommend settings for Adaptive Phonak digital, HAL-HL2 and DSL v5a, The recommended settings were nowhere near my actual settings and I think they are unusable but reading the analysis was a real education. I have not analyzed the ChatGPT results but they look much more reasonable. Anyway using these tools could be addictive and fun 
Yeah I think chatgpt is still ahead in terms of correct proper results. I ask what makes you qualified to adjust my hearing aid. After the response I got, I am sold haha. Anyway it did recommend i save my previous settings in case I am not satisfied. Once I get Noah link next week I’ll give it a crack and post my results.
I had a conversation with Chatgpt about fine tuning m’y spheres for a more natural sound for music. The results were quite detailed. I hadn’t thought of submitting my audiogram but will do so. I will report back eventually.
Looking forward to hearing about it 
Chatgpt will analyse your audiogram and make the necessary adjustments and will work through it with you in fine tuning.
Yes, it gave me precise settings for each frequency to enhance music listening. Not sure how these compare to the original settings but will have a look later today.
I’m assuming it went by your Audiogram?
I ask GPT what makes you qualified to tune my hearing, And it went on about knowing everything there is to know about the tuning and that it is here for me until we get it right. I told it, “sold” 
Yes, it went by my audiogram and came up with the following settings for gain at each frequency for moderate sound (G65)
| Frequency | Suggested Gain (dB) | Notes |
|---|---|---|
| 250 Hz | +10 to +15 | Adds warmth & body to music (bass, drums) |
| 500 Hz | +15 | Support vocals & rhythm |
| 750 Hz | +20 | Helps vocal richness |
| 1000 Hz | +25 | Critical for vocal clarity |
| 1500 Hz | +28 | Smooth transition into upper mids |
| 2000 Hz | +25 | Reduce slightly if music sounds sharp |
| 3000 Hz | +20 | Moderate—too much sounds harsh |
| 4000 Hz | +15 | Controls brightness, watch for distortion |
| 6000 Hz | +10 | Keep modest due to your high-frequency loss |
| 8000 Hz | +8 (or less) | Go easy—avoid harshness or feedback |
It then guided me as to how to adjust these frequencies. However, I am struggling to do so. Each time I move a given frequency up or down, it affects all the other frequencies. I asked Chatgpt to help me with this but I had reached my message limit. Guess I will have to subscribe to Chatgpt Plus. For interest, these are the settings that I had:
| Frequency | Suggested Gain (dB) | Notes |
|---|---|---|
| 250 Hz | +5 | |
| 500 Hz | +20 | |
| 750 Hz | +32 | |
| 1000 Hz | +39 | |
| 1500 Hz | +40 | |
| 2000 Hz | +32 | |
| 3000 Hz | +25 | |
| 4000 Hz | +19 | |
| 6000 Hz | +17 | |
| 8000 Hz | +17 | |
| So quite a difference. I’d love to program this in and see what difference it makes. |
I asked for target input amounts for the three Adaptive Phonak, NHL-NL2 and DSL v5a and it took me three sessions to complete the analysisl (I timed out twice and had to continue the following day).
Do you use Chatgpt free?
I have been using the free version so far. I guess I would need to get the Plus version if I don’t want to be patient and continue the inquiry/analysis the next day.
Yes. I’m going to sign up for plus today. Using it for DIY on my HAs is a clincher but I have a complex legal situation to manage for which it will also be very helpful. My son has Pro and uses it for all manner of things.
So I got Chatgpt+ and set to work on adjusting gain levels. After I set the levels that it recommended I listened to a couple of tracks. Bass was good but I could hardly hear vocals. It’s recommending me to increase gain at 2khz and 3khz 1db at a time and then test.
After 4 or 5 sessions I am getting somewhere. I have changed the gain for moderate sound at frequencies from 250hz to 4khz following chatgpt’s suggestions. It also suggested suitable tracks to listen to. Will do some more on this tomorrow if the wet weather here in south west UK continues. It will be interesting to talk this through with my audiologist at our upcoming remote appointment.
Seems a bit of nonsense really, appreciate your trying something different, but unfortunately AI has no way of knowing how you like to hear, so it’s just experimenting which you could do better yourself DIY, it’d be best to stick with the software and your brain to find a good fitting your happy with, DIY on the fly in different environments really helps as well, real world testing.
I think using GPT plus which I have just downloaded will give you all the time you need to help perfect the best listening experience you can possibly attain from your hearing aid. Unlike going to a specialist which give you limited time per visit, you can spend hours with good guidance from GPT. As tenkan said, you can do it on your own… which I considered but to be honest I am a little in the dark and have no experience in tuning. I have though learnt lots I had no idea about thanks to GPT.
Keep at it Teewens until you get the best possible result. Good luck and please keep us informed
I did the same. Paid for the Plus $20 aud per month. It is very good. At least do for one month to get the settings right for your hearing aid and learn heaps along the way
Looks like it lowered all the settings except the 250Hz
Yes Steve, I think the main thing is that you learn a lot just by interacting with it. I have learned how to quickly adjust gains across the frequencies, apply them and test. Also what the impact for music listening will be with different frequencies. Chatgpt has also been useful in suggesting tracks to use to test changes. I think I have made good progress towards my goal of making music sound more natural but I am not there yet. As you say, Steve, it’s like having an audiologist on tap for as long and as often as you like. I’m not sure about its skill level and suspect that a good audiologist, which I have, would be much more efficient. But I plan to combine the two. Tomorrow I will send my audiologist an email detailing the changes that I have made as recommended by Chatgpt and then discuss them with him on Tuesday. This will be a remote session but he’ll be able to tweak the gain levels and other settings. After that I’ll report back to Chatgpt and take it forward from there.
Glad to hear you are making progress. I will start tweaking once Noahlink arrives. My music listening experience sounds completely alien to me… not good.
Hopefully your Audiologist will be open to you using GPT, I suspect some may see it as a threat to their livelihood. In any case it’s nice to have the power in our hands to make adjustments.