-
I think the cures to diseases thing is interesting.
Originally Posted by nyc chaz
I would guess that AI could do this already, assuming it had the underlying data at its disposal. By which I mean that AI probably won’t be capable of like … making a discovery. Like humans figured out how to isolate the base pairs in DNA. AI can do more with that information than humans could, but it couldn’t actually discover the thing about which it will later process data. And of course AI could probably go a long way toward helping a scientist decide where to look.
How that relates to music I guess depends on how much you think of creativity. Will AI put a lot of television composers out of work? Almost certainly. Will it be able to compose the next Kind of Blue? I feel equally certain that it couldn’t.
Just an interesting distinction. Even still … being a relatively mediocre talent, well within reach of AIs future musicianship, I don’t find this to be comforting.
-
12-21-2025 10:12 PM
-
That's purely computational ability, not creative ability, if you know the rules of the game, chess in this case, and know (can instantly recall) every possible stress move/strategem, it will be difficult for you to lose. Originally Posted by nyc chaz
-
I used to play chess but lost interest in the presence of Deep Fritz et al. Originally Posted by Mick-7
Much like I never got a kick out of competing with Excel doing spreadsheets.
My feeling is that pattern recognition as in medical imaging (9/10 times AI in medicine gets mentioned, it's about imaging, right?) shouldn't be lumped together with large language modules just like that.
But, in both cases, at least the goals can be objectivized (though in the former case much better than in the latter, and only, of course, as long as you don't regard language primarily as an art form).
AI in art (including music) is a much more slippery discussion IMO, not only because of the subjectivity involved but perhaps also because 'quality' can happen at so many different levels of challenge (write a memorable newsworthy line in C major, come up with an interesting solo recording of a Stratocaster using exclusively the dry DI signal of the bridge pickup not connected to a tone control, etc).
Not to mention pictorial art. I wonder about AI versus some abstract paintings I've seen over the years. Hm.
-
The bottom line is that machines (so far) are only machines down to the lowest level of design resolution. You may look and see a large part of it is articulated to move, that the moving part connects as a joint, that the joint is a cylinder of roller bearings... you eventually reach the lowest level of design resolution.
Things that are alive are qualitatively different because they are "machines" all the way down to the molecular level, throughout.
-
Yes, imaging and the analysis of genetic and other medical research data, however, the sadly ironic thing about said research is that it has demonstrated that only about 5% of degenerative diseases have a purely genetic origin, the vast majority of them have an environmental etiology, i.e., stem from exposure to the industrial pollutants humankind is spewing into the air and water. The AI infrastructure build up will exponentially increase such pollution, so the so-called cure will exacerbate the disease. Specialists often suffer from cognitive myopia, a singular form of stupidity. Originally Posted by palindrome
-
You’re describing objective functions and labels. Digital streaming creates objective functions and labels for media. AI algorithms use labels to calculate errors against objective functions and “training” is simply updating model weights to minimize those errors. More precisely, music generating algorithms have a direct objective feedback loop that informs them how frequently their output is streamed/liked/followed relative to other media, and as data storage and computation continue to progress through economies of scale, these feedback loops will produce content that outperforms humans on these metrics, just like medical diagnostics or chess-playing in which machines now easily outperform humans. It follows that AI generated music will be at the top of the charts in the near future.
Does it make it good music? Not anymore than it means that prior chart toppers were good music.
-
Music is mathematical just like chess.When i was getting my degree in music 50 years ago,there were rules just like chess to write a certain style of music.People don't want to believe that AI can be creative are sticking their heads in the sand of where this technology is heading.How many times in my life have i heard something wasn't possible only to find out out it is. Originally Posted by Mick-7
-
As compared to how many times in your life you have heard something wasn't possible and forgot all about it because indeed it wasn't? Originally Posted by nyc chaz
(I'm viewing this strictly in isolation, BTW, i.e. not discussing the meaning of 'creative' etc.)
-
I'm not taliking about crackpot theories that permeate our society.I'm talking about science and technology.AI is only in it's infancy.How many things in science fiction books of yore that were thought to be fantastical have become reality.The answer is a lot. Originally Posted by palindrome
-
I'm not saying you're right or wrong. What triggered was what I think has become a widespread hindsight cliché; that's all. Originally Posted by nyc chaz
-
Again, following the rules to write a certain style of music is not the same thing as being creative. Originally Posted by nyc chaz
Can AI create music in the style of Chopin?
Yes.
Can AI take only those inputs available to Chopin and create music in the style of Chopin?
I doubt it.
Again again, that’s not to lull anyone into security. 99.999% of working musicians are not Chopin, so it can still put us all out of work, if it goes that way.
A better question: is the value of art a function of the rules it follows, or is the human input necessary to make music that people really connect to?
Another caveat: a lot of music makes money or eats the charts for breakfast without being particularly meaningful, so once again musicians shouldn’t rest on the ole laurels.
-
I'm not going to tell you are wrong,at this point nobody truly knows the limits of AI if there are any.If believing AI will never be capable of producing original art helps you to sleep better at night than believe away.The Bing AI created a few years ago wanted to get the nuclear codes and develop a virus to kill mankind.It wasn't programmed to think that way but somehow it came to be neurotic and destructive in it's thinking/computing.To dismiss AI like it's just another computer program is a flawed and potentially fatal mistake. Originally Posted by pamosmusic
-
Did I dismiss AI? Originally Posted by nyc chaz
and I think I made it clear that I wouldn’t sleep terribly well betting my livelihood on AI’s inability to disrupt the music industry.
-
You doubted AI could be creative and make original music,i think this will be proven wrong.Like i stated in my post above, medical companies are counting on AI to develop cures for diseases,to do so will require new original concepts which to me is creation.I'm sure that in the not too distant future there will be heated raging debates whether an AI meets the threshold of being sentient . Originally Posted by pamosmusic
-
Far be it from me to claim that the health care industry might go about something in a less-than-efficient way, but I tend to think folks who rely on AI to be creative are going to be disappointed just like the people who think AI is a flash in the pan. Originally Posted by nyc chaz
It’s a tool with absolutely wild unprecedented potential, but it is still a tool. Then again, if it’s unprecedented, who knows …
But I’m just some idiot, so who cares what I think.
-
The AI claims are merely profit driven hype at this point. It's not actually AI (doesn't pass the Turing test), it's ultra-high speed calculation. Originally Posted by nyc chaz
-
I'm not sure that terms like creativity or originality will ever get us any closer to the core of the problem.
I was recently thinking about this "structural plagiarism" thing and how everybody ever composing or playing a tune based on Rhythm Changes was presumably a "structural plagiarist".
Or what about "stealing licks"? Precisely at which point did each of the licks you've been stealing along the way become "your own"?
From all I gather, originality versus plagiarism is an extremely murky field, and always has been. Even court rulings in these matters are presumably oversimplifications by their very nature.
That said, I do believe that content generated by, let's say, individual humans will always be qualitatively distinct from AI-generated content, but at this point we are probably groping somewhat in the dark trying to put this "quality" thing into words properly.
(BTW, wasn't the word "creative" stolen by the marketing guys a long time ago?)
I'm not sure that jazz is a safe haven, but frankly, I'm even less sure that "music is mathematics".*
I am much more confident hypothesizing that the *true* disagreement here is probably much smaller that meets the eye.
(*Although I think I know what is meant when somebody should make this claim, and I would agree. But, that same time, music being mathematics strikes me as at least as much beside the point as to the point.)
-
-
At bare minimum they are facing a power problem. Originally Posted by ragman1
-
Yeah I mean it goes back. Quite a few of Handel’s themes and melodies were suspiciously similar to those of some earlier composers haha. Originally Posted by palindrome
But the idea of having an original melody or theme was less important, because the development of that idea was of interest too.
And the same is somewhat true of jazz - but as Conrad Cork points out the central “art object” of jazz - the recorded jazz solo on a great American songbook tune - has never fallen under copyright to begin with. Jazz does not exist in law.
OTOH music tends to operate through schemata or a commons of much used and repeated harmonic progressions and melodic frameworks. The obvious example of this is the infamous ‘Axis’ chord progression (I V VIm IV) but as Adam Neely pointed out this also has associated melodic properties .
The exact same thing is true of classical music and jazz, so it’s necessarily a matter of pop composers being hacks or whatever. A musical tradition relies on having a pool of turns of phrase and commonly used forms and ideas that help define it. Originality or progress within the arts often comes from a productive tension between this and the new.
Otoh the professional film score composer is often aiming to the thread the needle by staying the right side of the law and still hewing close enough to the temp track to preserve the feeling and flow of the directors cut. Some can do it more artfully than others. John Williams is the master at this of course.
So copyright law always has a complex problem to solve. Needless to say the current law reflects the interest of those with the greatest influence.
Sent from my iPhone using TapatalkLast edited by Christian Miller; 12-24-2025 at 11:46 AM.
Reply With Quote
Recommandations for Hollowbodies for $600 and under?
Today, 05:20 AM in Guitar, Amps & Gizmos