Skip to content
HearingTracker
Log in

Random thoughts on AI

This topic is closed.

11 replies998 viewsLast reply

Link to this topic

Found reading this amusing: Samsung says it adds fake detail to moon photos via “reference” photos - Ars Technica1

In essence when the camera detects you are trying to take a picture of the moon, it has stored information about the moon and enhances the image.

Got me thinking about how “” might try to enhance speech in noise. I think what we all we want is to make the speech stand out in some way from the noise and be understandable. One way to think of it would be to remove the noise. What if the “AI” instead “figured out” what was trying to be said and then played it back in a simulated voice? Considering the fun errors Siri makes I can imagine all kinds of errors beyond the ones we make ourselves. I’m sure I’ve got technical details wrong, but my point is that using AI in hearing aids could have unintended consequences.

6 people in this conversation6 min readJump to latest

I read the same article and was thinking the same thoughts relative to speech recognition with ’s.

I think the moral issue is what the OEM claims is being done and whether the user is fully aware of what’s going on, what the consequences might be, and has the opportunity to accept, reject, or tune the “AI” effects.

As far as Samsung’s moon “photos” go, that’s clearly immoral as it’s leading the buyer to think that the lens has the ability at 100x zoom to capture a higher image quality than it really does. Too bad for them to have a lens system that works pretty well and then trash its reputation by exceeding the limits of what’s acceptable.

I’m less sure where the “line” is. Smartphone photography has been “enhanced” for some time. Colors “pop,” blemishes disappear, sharpness is enhanced. The low light photography is amazing, but things look much brighter than they did in reality.

@Dusty said:

but this guy explains a nasty consequence of AI that will be very destructive in the near future,

I think a lot of journalism these days is in the “man bites dog” category. Who’s going to read your blog or watch your YouTube video if you’re just giving a Thumbs Up! :+1: and piling on positively to the general wave on something? OTH, if you can find the worm in the apple and warn people we are about to get expelled from the Garden of Eden, people will click on your article link or stream your YouTube video. Nuclear weapons can already destroy the human race (“nuclear winter,” etc.). Mass killings are a common occurrence in the U.S. these days. But everyone’s largely habituated or at least resigned to the status quo on these things. Same with climate change, the next pandemic, etc.

So, I’m not too ready to get terribly pushed out of shape by . Like any technological advance that led us to all of the above, it can be abused and be a threat. It would be stupid not to take advantage of AI to improve life and civilization. Also, stupid not to be on guard to make sure its use is not abused and unfair. I don’t think the fault of knowledge is in the ability to generate new stuff. It’s how knowledge is misused, and the fault for misusing it is in human brains, emotions, etc.

In reply to @Dusty

If I understand correctly, we already face this issue now as people aren’t very good at discerning what is real and what isn’t. won’t help, but it’s already pretty bad. Hmm. Wonder if they could create an AI tool that critiques posts for accuracy? :>)

@MDB said:

Wonder if they could create an AI tool that critiques posts for accuracy? :>)

I predict that there will be waves of things like that, except they won’t necessarily be truthful. Whatever gets the clicks. The Tesla phone thing has been going for at least a couple of years. One would think that the market for an easy to debunk assertion would subside quickly, and yet that one is still on the upswing apparently. Combined with ignorance and hope for magic, such crap could suck out all the oxygen, maybe even displace cat videos and porn eventually.

My take?

is good thing but I don’t want AI to automatically adjust thing for me. Think automatically reduce gains, increase gains, in and out hoom hoom etc. I don’t like that.

I want AI to do calculating for the best practice parameter then set it fixed for my hearing, yes. But not during on the use.

It would be cool though if AI can do the jobs like Audiologist in place. Talk with AI what went wrong, why I don’t like this or I want that etc. But leave my hearing aids program alone :stuck_out_tongue:

I read an article in The ASHRAE Journal recently. It was a rebuttal of traditional methods and measurement to determine indoor air quality.

Now we have inexpensive instruments that measure 7 essential factors and report “air quality”. The writer said that doing so 20 years ago the instruments would have cost the same as a car! Now ammazon sells them in the US for about $45.

Times are changing…

Are they practical instruments? Not yet…but I find it fascinating that this leap ahead will force necessary change in research. Old Standards will die…

@DaveL said:

Now we have inexpensive instruments that measure 7 essential factors and report “air quality”. The writer said that doing so 20 years ago the instruments would have cost the same as a car! Now ammazon sells them in the US for about $45.

Sometimes referred to as “party detectors.” https://www.youtube.com/watch?v=D2Wq4Yg52No The example in that tear down is a simple version, but many others have 8 functions and retail for less than the cost of a single sensor they purport to contain. The ruse is convincing until you use hand sanitizer nearby. :slight_smile:

Thanks Dusty.
I’ll wait until they have “air quality monitors” on the iWatch then. :slight_smile:

In reply to @jim_lewis

All marketing departments seem to want to jump on the ChatGPT bandwagon. Well, IMO, saying “we use AI to improve your hearing in noise” is rather vague. Reminds me of when JC Whitney promised “up to 25% improvement in gas mileage” for a whirly-go-round device under the carburetor. 0% is up to 25%.

We have been testing some new video AI software still under development. We are using it in conjunction with surveillance camera software (Network Video Recorder). The camera sends video streams to the NVR and the NVR (based on detection criteria) sends images to the AI software for evaluation. There are various “models” of images containing objects to be used for identification. These files get quite large. Some have animals, some have license plates, some have people and vehicles, etc. We use a model that contains “persons” and “vehicles”. We can define vehicles if we want to, such as cars, trucks, forklifts, etc. We also tell it criteria such as % confidence, how many samples to evaluate and how often (like 100 milliseconds) between samples. It is very complex. Using the video processor in the NVR, recognition times are in the 60-80 msec range if it recognizes the image. It makes mistakes, sometimes it thinks our metal stairway is a truck, depending on shadows. Very few false alerts. Without the AI package, vehicle headlights cause many false triggers, with AI almost none. One user claims that his cat is mad because the AI says it is a dog . Fine for surveillance, but I don’t want the same performance in health care.

We all want better hearing in noise. AI sounds like a perfect solution until you consider that a lot of that noise is from voices that we don’t want to hear versus voices we do want to hear. I suppose you could provide a model of the voices that you do want to hear, for AI to examine. But, what if it introduces delays in the half second range?

My Omnia’s are doing quite well with forward focus, (up to 100%).

Related discussions