• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

OpenAI ChatGPT has some interesting insights about audio

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,287
Likes
2,759
Location
Salvador-Bahia-Brasil
it is expected that it gives wron or confusing answers. it is still learning.
my complaint though is that you can't force it to learn by simple logic:

1674906421879.png


it should be possible to guide it to correct itself by logical conlusion
 

TonyJZX

Major Contributor
Joined
Aug 20, 2021
Messages
1,974
Likes
1,920
well that whole spiel on cables has all the hallmarks of the danny ritchies, paul mcgowans and guttenbergs of the world...
 

ahofer

Master Contributor
Forum Donor
Joined
Jun 3, 2019
Messages
5,023
Likes
9,071
Location
New York City
Quoting a friend in another forum:

The thing about ChatGPT is that it is a large language model trained to provide answers that people (specifically the group of people who provided feedback into it;s ground-truth training dataset) found pleasing. It's important to keep that in mind -- it's not designed to be correct, or truthful, or insightful; it's designed to be pleasing.

Think of it as a state-of-the-art, automated bullshitter. That's actually a real engineering accomplishment, and it has some useful potential applications, but I wouldn't be worried about the robots seizing control of SkyNet anytime soon.
 

computer-audiophile

Major Contributor
Joined
Dec 12, 2022
Messages
2,565
Likes
2,880
Location
Germany
I don't care for how AI is described, but for me this isn't it yet. it is just a data analyser.

ChatGPT is simply very literate. I have already used it to sift through philosophical texts for me. It was very effective. For me it isn't 'it', but just another tool so far.
 

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,633
Likes
6,020
Location
Melbourne, Australia
AI is only as good as the data it is fed. I don't know what algorithm they used, but the potential that it may put out misleading information worries me. That information given about DAC's sounds plausible but more detail / evidence is required, for example what measurements would give a more complete picture. Are there things that are audible that has no way of being measured. And so on. At the moment what it said is just pure assertion.

I wonder if it has an opinion on subjects that might be dangerous to you if the advice was followed - bitcoin scams, alternative medicine, fringe politics ...
 

TonyJZX

Major Contributor
Joined
Aug 20, 2021
Messages
1,974
Likes
1,920
oh the racist Microsoft chat bot you mean? :D
the 'problem' i have with that is that an AI is not racist from the outset as much as a Casio FX100 is 'racist'... the programming can be 'malleable' to certain extents but its a mirror of what its exposed to... the AI went ugly because the people it learnt from is... whatever they were and are...
 

computer-audiophile

Major Contributor
Joined
Dec 12, 2022
Messages
2,565
Likes
2,880
Location
Germany
I wonder if it has an opinion on subjects that might be dangerous to you if the advice was followed - bitcoin scams, alternative medicine, fringe politics ...

This problem is real. The same goes for human advisors or influencers.
 

Plcamp

Addicted to Fun and Learning
Joined
Jul 6, 2020
Messages
860
Likes
1,318
Location
Ottawa
Ok, would it’s answers to questions from ASR members change if it read the full content of ASR? Do we know if it has or has not?
 

computer-audiophile

Major Contributor
Joined
Dec 12, 2022
Messages
2,565
Likes
2,880
Location
Germany
Ok, would it’s answers to questions from ASR members change if it read the full content of ASR? Do we know if it has or has not?
I asked ChatGPT about this. You can ask anything. It says it can't browse internet forums by itself.
 

computer-audiophile

Major Contributor
Joined
Dec 12, 2022
Messages
2,565
Likes
2,880
Location
Germany
The first computer dialog program I remember well was ELIZA by Joseph Weizenbaum. It seemed "intelligent" to some people at the time, but it was more of a joke. With ChatGPT, I would no longer call it a joke. It's more like a breakthrough that will change the world.

Unfortunately, the level of knowledge is not yet up to date. One example: Angela Merkel is still Chancellor of Germany for the program.
 
Last edited:

pjn

Active Member
Joined
Aug 28, 2021
Messages
176
Likes
167
Just set up an account with them to see if it can clean up my writing instead of making up answers. I was pleasantly surprised that it can do that. That said, it did change the meaning a bit and took out some specifics. Let me know whether you think my version is better or not. It is the conclusion in my iPurifier3 review:

My writing:
It seems that the iPurifier does what it says it does: clean up the signal/power lines over USB and with it help devices that need it. Then again every DAC I recommend of which there are probably 50 to 100, already perform their own filtering and don't need any help. After all, I test them all on my desktop workstation without any such filters. Adding an active device to the USB connection can reduce available power and potentially decrease measured performance as show in the case of Schiit Modi 3+. Seeing how a better DAC costs less than iPurifier3 anyway, there is no reason to deploy such a device.

ChatGPT:
The iPurifier is a device that claims to clean up the signal and power lines over USB, which can improve the performance of certain devices that may be sensitive to electrical noise. However, many high-quality digital-to-analog converters (DACs) already have built-in filtering and do not require additional help. Additionally, adding an active device like the iPurifier to the USB connection can decrease available power and potentially decrease overall performance. Given that a better DAC can often be obtained for less cost than the iPurifier, there may be little reason to use such a device.
If it was a human, I would say that it is hedging it's answer to avoid legal consequences!
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,632
Likes
6,232
Location
.de, DE, DEU

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,287
Likes
2,759
Location
Salvador-Bahia-Brasil
so, the other day I asked it about nasal tobaco beeing less harmfull then smoking and it wouldn't confirm this. I logicly tried to make it understand that this is fact, but it wouldn't learn.
BUT today the answer is correct.
I can't post the dialog becaus it is in portuguese.
I conclude that it in fact learns, but it takes a little
 

computer-audiophile

Major Contributor
Joined
Dec 12, 2022
Messages
2,565
Likes
2,880
Location
Germany
Also, neulich fragte ich es, ob Nasentabak weniger schädlich sei als das Rauchen, und es würde dies nicht bestätigen. Ich habe logischerweise versucht, ihm verständlich zu machen, dass dies eine Tatsache ist, aber er wollte es nicht lernen.
ABER heute ist die Antwort richtig.
Ich kann den Dialog nicht posten, weil er auf Portugiesisch ist.
Ich schließe daraus, dass es tatsächlich lernt, aber es dauert ein wenig
I would be surprised if it simply accepts opinions from individuals. Unfortunately, I know too little about how it works.
 
Top Bottom