OH, NO! AMIR IS UNDER ATTACK!!!

[mhardy6647]

Even better, a null comparison file for cheap and expensive cables/DACs

Robots'd click every time. (i.e., for the perfect null result)
Audiophiles... not so much.

 

[Robbo99999]

Lots of crawlers are active here, and thy feed AI. We might ask for anwers via this way ...

I was curious and fed ChatGPT a load of my posts, and got it to do a role play for me with it trying to impersonate me. It couldn't do it very effectively and it told me one of the reasons is that it has safeguards to prevent impersonation/fraud. There's probably AI you can buy in shady places or even design yourself that doesn't have those safeguards, but I was just curious to experiment with ChatGPT. I'm pleased it has those safeguards.

 

[TTT15]

ChatGPT is programmed with confirmation bias. So if you say ASR is great it will agree and also agree if you say ASR is nonsense. It is not intelligent. Just predictive text based on what it finds on the internet. So it regurgitates what it finds. I’ve found it useful for researching potential HiFi alternative options but it exhibits biases just like any internet source so should be viewed with scepticism like everything else you read anywhere on the internet.

 

[devinplombier]

it exhibits biases just like any internet source so should be viewed with scepticism like everything else you read anywhere on the internet.

Agreed. The threshold is not whether AI makes any errors at all, but whether its average error rate is non-inferior to that of humans. And I'd say it's pretty darn close right now.

 

[ahofer]

 

[EJ3]

It's funny, it really is
View attachment 477627

Food analogies are always there, the same as cars and watches

Can you please come up with a better analogy then? (that most people would understand).
You may have some cheese with your whine, I suppose.
But, if you are going to whine about something, how 'bout offering an alternative?
Or are you just whining because you can't think of something as good or better, so you just whine.
I'm getting tired of many that do this whine but have no alternative to offer.
That seems to indicate that the many of you who whine about it cannot think of something better.
Which seems to me to be ironic.
Whining about something, yet offering no solution:
seems to indicate that the
cooking & automobile analogy's
work better than anything that the whiner's can think of.

 

[ahofer]

Can you please come up with a better analogy then? (that most people would understand).
You may have some cheese with your whine, I suppose.
But, if you are going to whine about something, how 'bout offering an alternative?
Or are you just whining because you can't think of something as good or better, so you just whine.
I'm getting tired of many that do this whine but have no alternative to offer.
That seems to indicate that the many of you who whine about it cannot think of something better.
Which seems to me to be ironic.
Whining about something, yet offering no solution:
seems to indicate that the
cooking & automobile analogy's
work better than anything that the whiner's can think of.

Do we need a solution? I don’t follow this.

 

[EJ3]

Agreed. The threshold is not whether AI makes any errors at all, but whether its average error rate is non-inferior to that of humans. And I'd say it's pretty darn close right now.

On average, I find something wrong with the AI answer about 30% of the time, when I take a deep dive into AI's answer. Granted, human's do worse,
but human's are not supposed to be experts in every field. AI is supposed to be an expert in every field. But it just collate the most popular opinions of humans, with caveats.
So, it is actually just a popularity poll.

 

[egellings]

Knowledge is testable. When found in error, it can be corrected or updated as needed. Belief is read only. It is not to be challenged.

 

[Mort]

Knowledge is testable.

Except that statement is not testable.

 

[egellings]

Except that statement is not testable.

So long as the knowledge is, the statement doesn't have to be.

 

[respice finem]

But knowledge, that has to be corrected, wasn't really knowledge before

 

[egellings]

But knowledge, that has to be corrected, wasn't really knowledge before

True.

 

[respice finem]

There is such a thing as semi-knowledge: correct in the sense of being evidence based, but incomplete.
My former work environment (medicine) is an "ocean" of it.

 

[Robin L]

What's wrong with the UFO thread? It has provided members with a wealth of useful, factual information along with many reasoned and well-considered opinions. Just like the vinyl renaissance thread.

It was a massive Gish-Gallop.

(and I helped!)

 

[Robbo99999]

ChatGPT is programmed with confirmation bias. So if you say ASR is great it will agree and also agree if you say ASR is nonsense. It is not intelligent. Just predictive text based on what it finds on the internet. So it regurgitates what it finds. I’ve found it useful for researching potential HiFi alternative options but it exhibits biases just like any internet source so should be viewed with scepticism like everything else you read anywhere on the internet.

I don't think GPT 5 has as much confirmation bias as the previous version, sycophantic was an often used term to describe the previous one. I think you can sometimes say to Chat GPT "you're not just agreeing with me are you?". But yeah, ChatGPT can make various mistakes, I've caught it making mistakes. I think it's a good tool though, just this year I started using it for anything really, and to see how flexible it is & what it can do.

 

[Westsounds]

Just say the word Amir and we'll 'Order 66' them I'm pretty sure this forum could wipe out their empire

 

[pablolie]

.... AI is supposed to be an expert in every field ...

Totally disagree. AI is very specifically trained for a specific purpose, and knows nothing outside of that. If you train an AI model to recognize a rabbit in a picture, it will not be able nor develop an interest in recognizing anything else other than as "not a rabbit".

 

[mhardy6647]

That's pretty much the way frogs work. If they see motion, it's food. Otherwise, it's irrelevant.
EDIT: Which for some reason got me thinking about Roger Sperry's epochal eye rotation experiments in frogs.

Roger Wolcott Sperry (1913–1994) | Embryo Project Encyclopedia

Roger Wolcott Sperry studied the function of the nervous system in the US during the twentieth century. He studied split-brain patterns in cats and humans that result from separating the two hemispheres of the brain by cutting the corpus callosum, the bridge between the two hemispheres of the...

embryo.asu.edu

Roger Sperry and his chemoaffinity hypothesis - PubMed

In the early 1940s, Roger Sperry performed a series of insightful experiments on the visual system of lower vertebrates that led him to draw two important conclusions: When optic fibers were severed, the regenerating fibers grew back to their original loci in the midbrain tectum to re-establish...

pubmed.ncbi.nlm.nih.gov

 

[Timcognito]

That's pretty much the way frogs work. If they see motion, it's food. Otherwise, it's irrelevant.

And this is how politics works also a natural process