Using AI In Audio Debates & GR Research [Video]

[DualTriode]

Dismissed. You can triple down, to only be dismissed again.

Dismissed by opinion.
That is my point.

 

[antcollinet]

Dismissed by opinion.
That is my point.

None of these posts - neither yours nor any of the replies, including mine, including this one, nor any of the notes in my signature (do you believe BTW that there is a solved mystery somewhere that has turned out to be magic?), nor the dismissal of your posts have been examples of "settled science"

Please feel free to point to any settled science - say a peer reviewed paper accepted in a scientific journal, or pick something closer to home : perhaps the sampling theorem, or the work of Maxwell or anything else along those lines as long as it is actually "settled science" - and demonstrate how it is "no more than strong opinion"

To fully validate your rather pointless statement, in order to demonstrate the "often" part, you'll probably need to pick apart multiple instances of "settled science"

 

[Alan O]

I have used AI with great success. But I am a diligent user.

Yep, I use AI every day in my work. It saves me time, and helps make me more money. But I'd never trust it for important things like choosing audio gear!

 

[Hipster Doofus]

WOW, NICE SHIRT ....

ok now my 2 cents... Just a suggestion

It's clear to me that the world is full of fake news, misguided souls, self-promoters and Marketing. Oh yeah and maybe a little lying.

So, one response could be... How do you set up ASR to continue past yours and our lifespan? Perhaps setting up ASR as a non-profit with standards that could be carried forward may be a better use of time. But who am i to say?

 

[amirm]

Maybe by then AI will be trustworthy and tell consumers the truth about audio!

 

[DualTriode]

Please feel free to point to any settled science - say a peer reviewed paper accepted in a scientific journal, or pick something closer to home : perhaps the sampling theorem, or the work of Maxwell or anything else along those lines as long as it is actually "settled science" - and demonstrate how it is "no more than strong opinion"

I do not need to be the first to cite a peer reviewed paper in this thread.

The hard science researchers in the universities and industry are not likely use the term "Settled Science". You know Science evolves (yes pun).

A couple or 3 examples of "settled Science" come from the not too long ago news cycle:

If you take the vaccine you will not contract the disease or pass the disease to others.
The U.N. Climate people have changed their minds, Miami Florida will not be under water in a minute. I suspect that Al Gore has moved to Miami.
Duke University was fined over 100 million dollars for falsifying medical research data. Fraudulent Science.

Thanks for taking the bait.
My post was intended to be illustrative. Clearly it was.

 

[PristineSound]

I do not need to be the first to cite a peer reviewed paper in this thread.

The hard science researchers in the universities and industry are not likely use the term "Settled Science". You know Science evolves (yes pun).

A couple or 3 examples of "settled Science" come from the not too long ago news cycle:

If you take the vaccine you will not contract the disease or pass the disease to others.
The U.N. Climate people have changed their minds, Miami Florida will not be under water in a minute. I suspect that Al Gore has moved to Miami.
Duke University was fined over 100 million dollars for falsifying medical research data. Fraudulent Science.

Thanks for taking the bait.
My post was intended to be illustrative. Clearly it was.

 

[amirm]

The hard science researchers in the universities and industry are not likely use the term "Settled Science". You know Science evolves (yes pun).

Of course they do. Where do you get stuff like this?

As for rest of your post, you best not continue. Won't warn you again with respect to topic drift and getting into things that can be highly political.

 

[AudioX3]

Danny is the perfect person to start off with and then jump away from. He got me interested in audio science and components. It sounded right and showed discernable differences on charts and measurements. Then ASR came around, and there was a world of engineering knowledge to be had. Now the limus test, the anacdotal arguements. ASR's anacdotal arguements point back to measurements, electrical engineering, scientific method. GR Research points to anacdotal rules of thumb and emotion and scoffing at measurements others make while extoling his measurements. He then does the folksy thing, so that is surely trustworthy. It maybe comes to what tribe you want to belong to. I will say it seems much more fun and mystical to be with the Audiophiles. My pocketbook and ears like the Engineers and it turns out it kind of fun there too.

 

[xnor]

I apologize in advance for the potentially off-topic rant...

AI is useful in fields like software development because the field is based on logic and maths and you can't fake incorrect software working correctly in a widespread manner.
Naturally, the results will be of much higher quality.

But in areas like audiophilia, AI models inevitably will be trained on mountains of mixed subjectivist "facts" and marketing BS, which are often indistinguishable.
This literally poisons the AI.

So it will "lie" straight to your face based on completely made up nonsense and half truths, because of the utterly warped reality it was trained on.
Unless you know this, you'll fall into a confirmation bias rabbit hole. Just like people have been "educated" over the past decade by carefully curated FYPs on social media platforms. /s

Every prompt should request a breakdown to underlying EE principles and actual psychoacoustic research, and even then you have to be skeptical.

 

[AdrianusG]

My best case for using AI
View attachment 532860

Love your living room and equipment , that is, if it is yours of course

 

[xnor]

After watching the video: @amirm is absolutely right. Even when asking for references to back up the synthesized responses, AI usually will not give you the actual references used to come to its "conclusion"... because it didn't in the first place.
So instead it will come up with references that are related or use web search to find websites that confirm its own answers after the fact! This is confirmation bias all over again.

Now there have been some advancements in recent months in the "thinking" models, which basically repeatedly insert hidden skeptical questions into the prompt, but since these AI models aren't capable of reasoning they may still take a wrong turn or make up nonsense.


A very important related life lesson:
Just like we learn through mistakes, information is gained through contradiction.
You don't learn that "all swans are white" by looking for more white swans. Sure, it helps a bit to have observed a million white swans vs a thousand, but it helps a whole lot more to find just a single black swan.

And to prove my point, I philosophized with Gemini's latest "extended thinking" model about this and it came up with this gem:

If theory T is true, then we should observe prediction P (T ⇒ P).
If we observe P, we cannot logically conclude that T is true. To do so is a logical fallacy known as affirming the consequent. A thousand green swans do not prove the rule "all swans are white."

Where's the puke emoji?

 

[Mort]

Thanks for taking the bait.

So you concede you are trolling.

 

[Hapo]

...trolling is not necessarily a bad thing, like being a troll...

 

[RickS]

...trolling is not necessarily a bad thing, like being a troll...

only if members like wasting staff time, getting warnings, and potential bans!

 

[Hapo]

...I will try to avoid wasting your time in the future and certainly meant no disrespect...I can see that DualTriode is getting to be a troll here...

...I finally got up to speed on this Danny thing...I had not really paid enough attention to it because I figgered even a cave man could see it was a load of horse shit...

...I thought this was just about AI...

 

[RickS]

...I will try to avoid wasting your time in the future and certainly meant no disrespect...I can see that DualTriode is getting to be a troll here...

...I finally got up to speed on this Danny thing...I had not really paid enough attention to it because I figgered even a cave man could see it was a load of horse shit...

...I thought this was just about AI...

Np, your comment gave me an opening and my response should have been less directed. I have fixed.

 

[Hapo]

...I thought that AI stuff I posted backed up Amir(ASR)...also I thought it was funny...

...one can get AI to support almost any viewpoint...

...even with careful prompts and instructions to not post bloat it trips out...

...I still find it very useful in many aspects of audio science...it will manipulate my opinions given the chance...

 

[Koeitje]

If you take the vaccine you will not contract the disease or pass the disease to others.

Take your strawman arguments someplace else, because this was never stated as such and you damn well know it.

Anyway, anyone actually using LLMs at work will know that you really need to think about a good prompt and that without a thorough understanding of the subject yourself you can easily be lead astray. It tends to poop out a lot of nonsense that a non-subject expert will not spot.

 

[xray]

Anyway, anyone actually using LLMs at work will know that you really need to think about a good prompt and that without a thorough understanding of the subject yourself you can easily be lead astray. It tends to poop out a lot of nonsense that a non-subject expert will not spot.

People who only use the free AI services for casual tasks have no clue what they are up against. For AI to be useful one has to know where they are going and be able to recognize when they get there. It also involves carefully prompting to make sure you don't take a contorted route that doesn't end up wasting a lot of time. A lot of times telling it what you don't want us just as important and that requires some knowledge.