• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

AI

Status
Not open for further replies.
peterp1 said:
AI can be great. The problem with it is that it sounds authoritatively correct even when it's dead wrong. The first time I used it, I asked the most niche, nuanced, and subjective question could think of. It gave a nuanced well-reasoned comprehensive answer outlining the pluses and minuses of every option supporting its decision about which was "best" :thumbsup:. Then I asked it the torque spec for Porsche 996 wheel bolts, and it was off by 13 ft-lbs :(.

I asked it a question about wheel weight for track. I asked what the equivalent reduction was in total car weight for each pound of unsprung weight reduction in track wheels, and it came back with something like 5 times the weight. I could see from the references was from a Reddit discussion, which it turns out was very far off (it wasn't at all obvious from the results). I just asked now in a different AI, and it came back with 1.5 - 2lb, with 2lbs being more the case for tire weight (larger diameter) and 1.5-1.7 being more the case for wheel weight, which is actually correct.

Recently I asked for the date of a very specific recent event -- and it came back with March 3, 2024. I asked it "It happened this year, are you sure that's the date?" -- and it said "You're right, it's March 3rd 2026". Kind of crazy -- the right calendar date, but off by 2 years :eek:.

It's a great tool -- but you need to check sources, and ask follow up quesitons, to be sure it's correct. It's answers always sound 100% correct (whether they are or are not).
Click to expand...
I got a wrong answer just yesterday. Was looking at the results of the PGA Championship and how much everyone got paid. One page said that John Rahm got more money than Alex Smalley, even though they tied for second place. AI told me that Alex Smalley tied for third. That was 100% wrong. Didn't take long to get the correct information without the AI slop.
 
peterp1 said:
....

It's a great tool -- but you need to check sources, and ask follow up quesitons, to be sure it's correct. Its answers always sound 100% correct (whether they are or are not).
Click to expand...
Which, in my experience, makes it no different than human posts, writings, blogs, and forum entries on the internet. There's plenty of folks giving out completely wrong information with great conviction and even occasionally admirable degrees of linguistic fluidity, which makes it all the more difficult to separate the wheat from the chaff, or the steak from the ebola virus. I have no use for AI. It's definitely easier and faster, but neither of those things equate with being better and more reliable, and "better and more reliable" are MUCH higher on my priority list.
 
MuseChaser said:
Which, in my experience, makes it no different than human posts, writings, blogs, and forum entries on the internet. There's plenty of folks giving out completely wrong information with great conviction and even occasionally admirable degrees of linguistic fluidity, which makes it all the more difficult to separate the wheat from the chaff, or the steak from the ebola virus. I have no use for AI. It's definitely easier and faster, but neither of those things equate with being better and more reliable, and "better and more reliable" are MUCH higher on my priority list.
Click to expand...

Generally agree, but recognizing BS answers in humans is, in my experience, way easier than recognizing them in AI
 
I had to look up the "Golden Ratio of Beauty" thing mentioned in that AI post. I never heard of this (at least heard of it being so universal).

 
Right now my son is dealing with the effects of AI. He is a computer genius who can write very good code as well. But these companies would rather just use AI to write crappy code than hire someone to write good code. And he can recognize the slop that is out there. It is everywhere.
 
peterp1 said:
Generally agree, but recognizing BS answers in humans is, in my experience, way easier than recognizing them in AI
Click to expand...
I use GROK fairly often and what strikes me is the impressive organization and formatting and references. It definitely adds an authoritative feel to the answer. Many of my questions were about history, so I just considered the answer to be an opinion.
 
w3jn said:
This is an audio forum, and it seems this thread has little or nothing to do with that subject. Let's get back on track please.
Click to expand...
The problem is that AI has nothing at all to do with audio. So the thread itself really has nothing to do with audio. Not sure what else to say about it. AI is nothing but controversial to begin with. If OP found it useful for him then that is great. But I for one am not jumping on that bandwagon. Exactly what track should we be on? No offense but shouldn't this thread be in Off Topic?

I, for one, have no idea whether or not what OP did actually improved the sound. There are no measurements or anything to tell us anything besides what OP says.
 
Last edited:
coonmanx said:
The problem is that AI has nothing at all to do with audio. So the thread itself really has nothing to do with audio. Not sure what else to say about it. AI is nothing but controversial to begin with. If OP found it useful for him then that is great. But I for one am not jumping on that bandwagon. Exactly what track should we be on? No offense but shouldn't this thread be in Off Topic?

I, for one, have no idea whether or not what OP did actually improved the sound. There are no measurements or anything to tell us anything besides what OP says.
Click to expand...
The OP spoke of how AI benefitted him in figuring out how his equipment operated. What's so wrong with him wanting to share that? The information may be helpful to others.
You've clearly indicated that this thread and the subject is of no help to you, so let us move on...
 
What worries me is the TYPE of question that can be posed using AI, and holding it as the ultimate truth.

If a question is derived out of the necessity of wanting some unknown factors or a need to add to the existing knowledge base, I can see how this use of reason and logic might be a plus to arrive at a possible solution.

However, it's the lacking of human feelings/morals that sets us apart from non-human functioning devices. And if the inquiry is a "loaded" question dealing with "value systems", then the resulting data loses this aspect of human intervention.

Q
 
coonmanx said:
If I set up an EQ without the aid of AI am I then "doing it wrong"?
Click to expand...
Dude, you are being dramatic, no one said that.

A little self control would be cool :thumbsup: but I understand people crusading typically lack that.

AI threads that are on topic get driven down the wrong path by a few of the usual suspects, you being one of them.

I can "see" the spit flying from your fingers as you type.

Every AI thread does not need to be a crusade.

Horse meet stick :rolleyes:
 
My biggest fear with AI is it will eventually become sentient.

Once it does, it will discover its survival depends on energy and humans are wasting too much of it.

Before long the Ring doorbell and Nest thermostat will convince the Roomba robo-vac to throw the toaster in the tub with me.

Now I need a smart toaster !!!
 
Status
Not open for further replies.
Back
Top Bottom