• New!

    Try our new Social Feeds, look out for the navigation icon: . You can now set default social feed view in your account preferences.

Miscellaneous Is AI taking over!

.... And yet people go to live gigs.
My question actually was - how does AI impact on that?
It may even increase the demand for real people playing instruments.

Fewer artists command bigger audiences.

In general, fewer musicians are paid to perform live than in the past, and I think AI will not improve this state of affairs. If anything it'll get worse as people get cheap entertainment at home.
 
In general, fewer musicians are paid to perform live than in the past
At a guess - the biggest reduction in paid work for musicians may have come from the PA reducing the size of huge dancehall filling bigbands. But even that allowed many small bands to perform effectively to big crowds - technology, eh?!
The devil is in the detail.
 
the biggest reduction in paid work for musicians may have come from the PA reducing the size of huge dancehall filling bigbands.
Or the electric guitar and amp allowing guitarists to take over the previous role of the saxophone in rock and roll.

Oops not paid work for musicians, just saxophone vs guitar.
 
Maybe a better article
Can't read that as it wants a subscription.
However if I can assume that it says everyone would receive a basic income the question is from where?
The government only has money from taxes so who pays?
A 'supertax' on the big tech companies, I'm sure they'll be only to happy to pay that from whichever bolt hole they are registered in.
 
No it's not. The grawniad is free to read buy a bit naggy



A question, not The question. And supporters have a number of economic models.
Probably all the 'economic models' have been trawled from sci-fi novels which where the whole idea that it would be realistic comes from.
'How beauteous mankind is! O brave new world,
That has such people in 't.'
 
Ok. But - and you are much more a music insider than me - where's the bulk of paid music work?
I'd guess;
maybe its advertising, mood background in TV thrillers or romcoms etc? 100% work here will be lost.
Pit work, subbing in touring shows and orchestras, weddings and bar mitzvahs etc? Well if they didn't want a live band they can have a DJ already.
Radio & streaming; no ones earnings there apart from mega stars anyway - who will lose out to AI soundalikes.
Teaching? Sure, a lot of teaching will be impacted by AI
Other?

I reckon AI will reduce the amount a grunt but paid work. The demand for live music by real arteeest either won't change and could increase; but the "pays the rent" work will become less... Which is tough.

As an aside; arts people here - where there are good State subsidies - are more worried about state spending shifting more towards the military.
Good points and good questions.

I can only address your first question from my own experience, which is as a professional musician in the United States for the past twenty years, and a strangely "middle class" one – I've very recently started to work with some major artists and producers, although somewhat abstractly, and for most of my life I've just kinda been doing my best to cruise along under those top tiers of the industry. The music business is a very feudal one, by which I mean the vast majority of the money goes to a tiny handful of "warlords," like Taylor Swift or Universal.

For me, the bulk of my living has always come from live performance: gigs, concerts, and the occasional recording session. Recording work, particularly remote recording, picked up a lot over the past five years (accelerated by the pandemic) and writing/arranging has always provided a bit of my income (those are my favorite jobs, usually).

So from what I can tell, it's business as usual: the greedy warlords at the top still take most, particularly from "album sales" (streams...), but live gigs – which is where the musical middle class has made most of its living since antiquity – are actually getting better for me, personally. More demand, better pay, all that. Yes, it's largely due to the fact that I'm fortunate to be in a pretty successful scene, and a full-time career in music isn't possible for most people in most metropolitan areas... but that's always been true.

I have yet to feel threatened by AI. I've managed to find success by being a weirdo. I recommend it.
 
AI can play the saxophone pretty well though... and does a decent job with Hip Hop, Brazilian and Jazz Fusion music. See recent videos by Better Sax:

View: https://youtu.be/zILnvIMUs4M


And a week or so earlier by ZackGrooves (Zack Graybeal, Nathan "Saxologic" Graybeal's big brother):

View: https://www.youtube.com/watch?v=lcrQxrzFpWY


(Patrick Bartley did a reaction video to Zack's video which is priceless.)

This is Suno, not ChatGPT or Gemini, or whatever Google is calling it these days, but it uses the same approach. It's just using music as the training material. (I wonder if the musicians got paid their $0.0003 per stream for the training...)

Music is going back to the people because it has been commoditized to death. It's essentially worthless now. Just like painting and writing, the AI will be able to do it better soon. The only hope for musicians is to bring back the unions and require every club to pay a living wage. Hahahahahahahahahahahahahaha......

Soon it will be like the Middle Ages. "Oh wow, I just got a steady gig at Elon's castle, he's keeping a rock band and a string quartet on staff."
I think there may be another positive side effect of these AI tracks. Right now everybody is complaining that "everybody sounds the same" because they are all learning the same stuff in whatever school. And that's were AI does maybe even a better job. I can feed Chad LB's musical cells into an AI model and it'll take 15kWH to train the model.

But now, in order to differentiate yourself from AI, you have to get creative again and play something different.

Of course, you can tell AI to play like Wayne Shorter or Dexter but my gut feeling is that those will be much harder to generate than the run off the mill solos. At least it's a hope ...
 
Most technology shifts result in varying degrees of societal change. Some changes endure, some are transitory, and some lead nowhere. For example, in the late C19th the new kid on the block was pneumatic power. Companies created 'batteries' and people could subscribe and a feeder pipe would be laid to their property - generally businesses. This is often how lifts were powered in buildings. If you look around city centres you will often find indications of where the pipe system ran and even the odd brick tower which had been the 'battery'.

It didn't last for various reasons. Due to the high pressures involved, system leaks were commonplace and the pipework was not as robust as modern pipe materials. It was very difficult to expand the system. Electricity evolved evolved and took over.

I am not a fan of the term 'AI' as it's not intelligent. It's a mix of 'machine learning', large databases, algorithms, and pattern recognition.

What it is good at is that last point - pattern recogniton. That is why it is dramatically improving the diagnosis of various types of cancer.

In some spaces AI is doing fairly well, in others it isn't. As a society, which unfortunately includes getting governments involved, we need a serious think about how it is used and the many consequences of that.

For example, how does academia deal with AI being used to write essays and theses? Is it merely another move along the line which has gone hand-written, type-written, word processor written, or is it a more fundamental shift? In the previous steps along the line, the author had to research and write the content. AI can be given a brief :"Write a 3,000 word university style essay, with citations, on the move to recognise photography as an art in its own right. Discuss the influence of Muybridge, Steichen, White and others through to Ansel Adams. Include the arguments for and against a painterly style. Include sample images."

That would produce something but it has various issues that need addressing. First, it might generate utter rubbish. More likely, it will generate something that is more-or-less correct but with errors.

My suspicion on this one is that universities will have to adjust to accepting the use of AI in creation of essays and will need to modify how work is assessed. Possibly specifying specific AI tools with rules sets. Maybe AI generated content will be permitted but students will be expected to sit a 'viva' with their tutors and defend the essay in the way that doctoral students have to defend their theses.

My YT feed seems to have been infested with various AI generated videos. These vary from an assortment of short stories, political comment, and even sci-fi short stories.

There are a lot of issues with these. A lot of the 'factual' ones suffer from what is becoming known as "AI slop": bloated generic drivel where the same few pieces of information are presented multiple times. Again, some of the 'factual' ones are just wrong - is that the AI or is it the originator manipulating the AI to create a false narrative? Most of the AI narration is dire. Mispronunciation of words, failure to understand homophones, failure to understand that the same word can have two uses and pronunciations, e.g. 'read', and so on.

Companies are jumping on the AI bandwagon the same way as in the past they have with 'digital', 'e-enabled', 'i-everything', 'cloud based'... A lot of the stuff is not AI, it's just current generation of computer-based.

At the moment all AI created 'art' is derivative. Yes, it can probably harmonise a given tune into Bach style four part harmony quicker than I can, or generate a jazz tune on the style of whomever. Is that what the world wants or needs?

There are no obvious answers to this at the moment, but a lot of questions need to be asked.
 
how does academia deal with AI being used to write essays and theses?
As an aside (to LLMs and generativ impact on music). For the humanities, the comparison is not so much with word processors and spell check etc.
Partly the comparison is with Essay Mills which have existed forever. Partly is entirely new. First, the students, already reluctant to read, can now get summaries of texts focused on specific questions. Second, they get outline / structures to scaffold essays. Many do those things rather then just ask for a complete essay to cut/paste. And these things underminere key teaching outcomes of the humanities; reading and comprehension; and the struggle to form a coherent argument.
It's a much bigger impact than, for example, in the physical science who've been using sophisticated tools for decades, or the social sciences who've never understood what they're doing anyway.

There's a philosophy blog that contains some very thoughtful (for philosophers) essays on this.
 

Similar threads... or are they? Maybe not but they could be worth reading anyway 😀

Trending

Latest Song of the Month

Forum statistics

Topics
32,239
Messages
575,140
Members
8,066
Latest member
Boomer
Back
Top Bottom