I'm No Luddite! But I Have...Artistic Concerns...

Having been at the same company for 41 years gives me a perspective on AI that I don’t think people have considered yet. Countless times Directors, Managers and Supervisors assign work looking for an outcome. When they get something other than what they want they send workers back to start over. I call it the “bring me a rock” assignment. Lack of detail and miscommunication results in a rock being delivered. “This isn’t the rock I wanted, go get me another.” Work ensues, effort, time and money are spent, nope still not the rock I asked for, do it again and again and again.

My point is, AI can create some astounding things, but when it isn’t what was wanted someone has to make it start over and modify the parameters of the assignment to change the outcome. Is AI going to convince someone this is what you wanted, what you asked for?

Les, you obviously deliver what the customer paid for. They may want what AI produces as a fad, but eventually they will only pay for work that produces the most revenue and I bet that will come from people, not machines. Trends change, old fashion comes back in style. Will AI convince us what we like? Will it change styles and follow trends?

I like to hear live music when I go out, I have a radio in my car and music on my phone etc. I have no interest in AI music. I won’t pay for it, or listen to it. I want real musicians and singers, not a machine’s opinion on what I should listen to. I could listen to music and watch movies and admire art for many lifetimes and never exhaust all of what was already produced. In my humble opinion, AI can’t compete with that. I think original work created by humans will be prized even more as AI output becomes commonplace. Remember, AI is a machine that only does what it was taught to do.
 
If your diet consists of just one food, you'll quickly tire of eating just one food. Thankfully, farmers and growers provide an incredibly diverse variety of food, some of which is natural and healthy, others which are processed, and laden with salt, fat, and additives.

If anyone knows how to read a nutrition label, they can choose for themselves whether they think the food will be healthy for them long-term or not.

By contrast, farmers and growers also produce a variety of vegetables, fruits, seafood, meats, and healthy grains.

By comparison, music is much like the food we eat...we take it in with our ears and senses, and some "foods" are healthier for us than others.

When we were much younger, we could eat just about anything and not need worry about our health. Now that we're older, we look back on our youth and ask ourselves why we made the choices we did back then. We made those choices because we lacked experience in matters that would help us live longer, healthier lives.

For that reason, I make it a conscious choice to know what I take in as "food," for the simple reason that one would not eat food that was not washed or prepared correctly, for food safety reasons. Many others in today's world do the same, but may do so without making a conscious choice. They eat what they like, not thinking about how it will affect them at one point or another in life.

I'm not a doctor, but am aware of how practicing healthy "food" intake choices helps output good results, whether that be having the energy, stamina, or level of daily activity you once had when you were younger and more vital.
 
Here's the thing: It's a tool.

Just like a DAW is a tool, a guitar is a tool, a delay pedal is a tool.

These things don't create spontaneously. They output what we tell them to, more or less, at the moment, less than more.

There's going to be a whole new skillset that is telling the machines what to do to get the output you want. If you are a professional in one of these fields that the tools support, you had better start learning the tools, lest you get left behind.

The clients don't care how you got to the results, they just care about the restults. If they can get it in half the time, even if it's the same cost, guess who they are going to call next time.

Les, you know this, you have been there through the transitions, from tape and outboard gear, a million miles of wires and a razor blade, to things being mainly 'in the box'

Just because there's a new tool, doesn't mean you can throw out the old ones either. I can totally see a future where someone's created a bass track with a machine learning model, but then run it through an LA-2A.
 
Here's the thing: It's a tool.

Just like a DAW is a tool, a guitar is a tool, a delay pedal is a tool.

These things don't create spontaneously. They output what we tell them to, more or less, at the moment, less than more.

There's going to be a whole new skillset that is telling the machines what to do to get the output you want. If you are a professional in one of these fields that the tools support, you had better start learning the tools, lest you get left behind.

The clients don't care how you got to the results, they just care about the restults. If they can get it in half the time, even if it's the same cost, guess who they are going to call next time.

Les, you know this, you have been there through the transitions, from tape and outboard gear, a million miles of wires and a razor blade, to things being mainly 'in the box'

Just because there's a new tool, doesn't mean you can throw out the old ones either. I can totally see a future where someone's created a bass track with a machine learning model, but then run it through an LA-2A.
But wouldn't the client bypass the middle man and just go directly to AI and use that tool all by themselves and save a buck?
 

This goes back almost ten years. The IBM Watson AI creating new recipes based on both flavor combinations in existing recipes and the bio-chemical nature of taste. In many ways, it’s similar to an AI creating music.

AI is going to AI is going to revolutionize our world. Maybe for the better, maybe for the worse. My job—renting out school buildings after school—could easily be done by an AI. My old lawyering job could (and is!) being done by AI. MDs are basically flow-chart followers.

Will all of that eventually eliminate a lot of jobs? Yup. Maybe life becomes very different.
 
It's going to happen, but licenses for the feed material/artists used by AI should be a source of remuneration IMO.

I like to think that live improvisation over AI-written jazz/world/Americana-type backing might be fun; well, for me at least. I'd set parameters like "Changes in the style of Pat Metheny with Eric Johnson, Eberhard Weber, Ralph Towner and Michael Hedges", just to see what it comes up with.


Expectations of very small-to-non-existent audiences are probably appropriate, however :)
 
Those of you who've read my many rants know that I'm not against beat boxes, synths, samplers, hip hop, or lots of other genres that make use of certain tools to help in the creation of music. These machines must be programmable by the artists involved (very few use presets), and considerable creativity is employed in their use by the good ones in each genre.

As it happens, today we went to my brother's for Easter brunch, and I sat with a group of artists and art professors who touched briefly on A.I. and the challenges it represents to not only art, but the teaching thereof.

Well, that includes most of the arts, doesn't it? Visual art can be done by AI; music can be created via AI; books can be written with AI, and I could go on.

I wondered aloud if there will be a group that will be perfectly happy to let AI do all the creating, and another group that will value human-created arts. And there might be a blend.

It's fascinating to talk about this. There's also an element of WTF!

Since I mostly think about music, and not visual arts nearly as much, I thought it'd be interesting to start a discussion here about what's happening, and what could happen.

I can tell you from experience that you can buy pre-programmed libraries of one-finger musical phrases. In my opinion, it sounds canned. I don't use it. Most composers don't, some so-called 'sound designers' are more inclined to because it's fast and easy. Got a commercial and a low audio budget? There you are. Instant bullsh!t.

"Wait a sec, Les, you do ad music, which is bullsh!t regardless, really."

"Sure, but at least it's original bullsh!t and I did it!" :rolleyes:

If I could make a prediction, I'd say that in addition to AI-generated stuff, there will be a movement in the arts toward live performance, played by humans, written by humans, etc. It'll be put up or shut up; live music always is.

I think (this might be just a hope) there will be a resurgence of interest in orchestra, dance (modern and ballet), live theater, live popular music shows, etc.

I'll leave the rest of my thoughts on hold for now. I'd love to hear what everyone else thinks!
I think you are 100% on the mark.
 
Can one simultaneously be a Luddite and Kemper user? Musically I consider myself Luddite Lite as I’ve used neither my Kemper, nor my acoustic in years. Electric guitars, tube amps and a small batch of pedals seem to be my universe. I achieve AI the old fashion way-with alcohol.
 
Of course there's the argument that human artists generally do the same thing. Long before AI, I heard people claim no truly original stories have been written since Shakespeare.

I've heard this before, but I've come to think that it's more accurate to say there have been no truly original plots since Shakespeare. There have been some wildly inventive stories, but the basic plots haven't changed much at all.
 
If I could make a prediction, I'd say that in addition to AI-generated stuff, there will be a movement in the arts toward live performance, played by humans, written by humans, etc. It'll be put up or shut up; live music always is.
Yep, this is how I feel, and I really hope it happens. I think there will be a resurgence in virtuosity on instruments, as a part of a larger cultural pride in humanity movement. And as a part of that I expect a huge emphasis on no backing tracks, no lip synching, and every sound you hear in the audience being made by someone on stage.
 
Having been at the same company for 41 years gives me a perspective on AI that I don’t think people have considered yet. Countless times Directors, Managers and Supervisors assign work looking for an outcome. When they get something other than what they want they send workers back to start over. I call it the “bring me a rock” assignment. Lack of detail and miscommunication results in a rock being delivered. “This isn’t the rock I wanted, go get me another.” Work ensues, effort, time and money are spent, nope still not the rock I asked for, do it again and again and again.

My point is, AI can create some astounding things, but when it isn’t what was wanted someone has to make it start over and modify the parameters of the assignment to change the outcome. Is AI going to convince someone this is what you wanted, what you asked for?

Les, you obviously deliver what the customer paid for. They may want what AI produces as a fad, but eventually they will only pay for work that produces the most revenue and I bet that will come from people, not machines. Trends change, old fashion comes back in style. Will AI convince us what we like? Will it change styles and follow trends?

I like to hear live music when I go out, I have a radio in my car and music on my phone etc. I have no interest in AI music. I won’t pay for it, or listen to it. I want real musicians and singers, not a machine’s opinion on what I should listen to. I could listen to music and watch movies and admire art for many lifetimes and never exhaust all of what was already produced. In my humble opinion, AI can’t compete with that. I think original work created by humans will be prized even more as AI output becomes commonplace. Remember, AI is a machine that only does what it was taught to do.
These are really interesting insights, John. 'Bring me a rock' is a very clever way of putting it, too!

Great input!
 
Here's the thing: It's a tool.

Just like a DAW is a tool, a guitar is a tool, a delay pedal is a tool.

These things don't create spontaneously. They output what we tell them to, more or less, at the moment, less than more.

There's going to be a whole new skillset that is telling the machines what to do to get the output you want. If you are a professional in one of these fields that the tools support, you had better start learning the tools, lest you get left behind.

The clients don't care how you got to the results, they just care about the restults. If they can get it in half the time, even if it's the same cost, guess who they are going to call next time.

Les, you know this, you have been there through the transitions, from tape and outboard gear, a million miles of wires and a razor blade, to things being mainly 'in the box'

Just because there's a new tool, doesn't mean you can throw out the old ones either. I can totally see a future where someone's created a bass track with a machine learning model, but then run it through an LA-2A.
Yes, I can see people using this tool in a creative way, for example, having it 'listen' to a track and come up with, say, a drum part, if they aren't drummers. etc.

On the other hand, I could record a good session drummer and get another human perspective, and a nuanced, original performance, or perform a part myself with a drum sound library, which doesn't take long, if there isn't budget for a live session.

Same with a violin section or any other instrument I don't personally play, where I might perform the part on keyboard with a sample library.

However...while using AI might work just fine for the client, playing at least the pitched instrument parts in myself informs the decisions and choices I'm going to make writing the rest of the piece.

Is that important? Yeah, I think it is, at least for me.
 
Here's a thought I'll throw out for discussion:

Client calls, and says, "Use AI to have virtual Miles Davis play a trumpet part using his Bitches Brew era style."

I could reply:

(a) OK; or,

(b) How about if I hire a great trumpet player so you'll have a more creative, better-sounding result, plus I can give direction to a human being; or,

(c) Why don't I negotiate a license to use something - maybe a recorded split - from Bitches Brew so you'll have the real deal, and Miles' estate benefits from his creative output (one of the things I do is help clients negotiate licenses to music)?

There are practical, creative and ethical choices here (for purposes of this discussion I'll avoid the legal minefield of 'soundalikes'),. What's your answer?
 
Last edited:
There is a Neil Young quote which I cannot find right now so I'll paraphrase. He's specifically speaking to the type of band he likes to play with and says he wants to play with people who play with real soul, fire and emotion, who play bad on a bad night and great on a great night, not someone who is technically proficient and plays great on a bad night and great +1 on a good night ( "and I think that's what people like too").

For those of us who love music, particularly LIVE music, I think this HUMAN element is essential. AI may evolve some aspects of art, but on a level of appreciation, I think there is something uniquely human that isn't entirely replicable and perhaps that lies in the human characteristic of making mistakes graceful.

While I'm sure AI will continue to invade the artistic, I think it is in our nature to appreciate the work of the human hand and mind.
 
There is a Neil Young quote which I cannot find right now so I'll paraphrase. He's specifically speaking to the type of band he likes to play with and says he wants to play with people who play with real soul, fire and emotion, who play bad on a bad night and great on a great night, not someone who is technically proficient and plays great on a bad night and great +1 on a good night ( "and I think that's what people like too").

For those of us who love music, particularly LIVE music, I think this HUMAN element is essential. AI may evolve some aspects of art, but on a level of appreciation, I think there is something uniquely human that isn't entirely replicable and perhaps that lies in the human characteristic of making mistakes graceful.

While I'm sure AI will continue to invade the artistic, I think it is in our nature to appreciate the work of the human hand and mind.
I tend to feel the same way.
 
Here's a thought I'll throw out for discussion:

Client calls, and says, "Use AI to have virtual Miles Davis play a trumpet part using his Bitches Brew era style."

I could reply:

(a) OK; or,

(b) How about if I hire a great trumpet player so you'll have a more creative, better-sounding result, plus I can give direction to a human being; or,

(c) Why don't I negotiate a license to use something - maybe a recorded split - from Bitches Brew so you'll have the real deal, and Miles' estate benefits from his creative output (one of the things I do is help clients negotiate licenses to music)?

There are practical, creative and ethical choices here (for purposes of this discussion I'll avoid the legal minefield of 'soundalikes'),. What's your answer?

It all depends on the requestor's motives and what they view as most important.

Do they:

(1) Really admire Miles Davis and really want an emulation of that period of his output?

(b) Want to brag to their friends/coworkers that they got an AI to do their new ad music way cheaper and so under budget that they're going to get a huge bonus for saving so much money?
 
The @Moondog Wily thread of AI generated art scares me.

My pithy comment about never having to pay a graphic designer is rooted in fear, because I would actually use AI “art” instead of paying someone.

If we follow the conclusions that eventually AI can and will do everything that humans do, and one by one, we become a workless society… I think I just assumed we’d all be freed up to follow creative pursuits. But if AI generated “art” is taking down the human component of creativity first before taking down monotonous labor then… yeesh. What are we gonna do with our lives?

I guess I just thought AI was gonna come for clerical jobs and stuff like that first.
 
Back
Top