AI Music Generators - Uses?

I have been messing around with suno a bit, try and see what the tech is like.

Technilogy wise its pretty insane how it can work like it does, you dont need to give it much and it makes something. The feat in itself without thinking about artistic implications, is mighty impressive!

Artistic implications though, can it be used, is that a bad thing?

For instance I have made a couple of things, where Im like thats a good riff or solo, id like to learn that. Whats the views on whether it can be used as a writing tool? Lyrics also it can give a base that you can modify or change to suit etc.

Im not totally sure how I feel about that, but I can see its application, but sure it muddies the waters out there in the artistic marketplace.

Here is an example of something it made, where I thought I need to figure that out to play along

https://suno.com/song/0caf0bd1-9b25-4f02-9fb2-0142890f6f47

Anyone got thought on. orals/ethics of it all? Or any comments in general on the technology?

theres a specific topic for AI related stuff

Thats my bad then, if someone could move it.

Ona note, I have tried to look for the best place to put this, never saw an Artificial Intelligence area. This forum format is incredibly difficult to navigate to find different sections!!!

its in the community hub

Done, Jamie

If you suspect there may be a category or that there should be one, you can always try a search.

And if not sure, feel free to drop one of the Mods (myself, @Richard_close2u or @LievenDV) a DM and we’ll lend a hand.

I’m really torn on this. On the one hand, AI is unavoidable going forward - in pretty much all parts of our lives. I hope there will still be a want and need for real music, performed by real people.

From a technical point of view, the AI generated tracks that I’ve heard are quite impressive. But it’s hard for me to get excited about something that is inherently “fake” (IMO).

Where I think it might be useful is for idea generation. Perhaps have an AI come up with some songs or riffs, then take those in a real band, with real people (replacing all parts) and tweak them until you have an actual original song. But I don’t want to hear any trace of AI in the final product :wink: (maybe that’s naive, but hey…)

1 Like

You’ll never do that with Generative AI. It’s, fundamentally, based on copying existing work.

All you will ever get with this technology is “derivative” material. Any originality on top will need to come from humans. And, IMO, that will need to be considerably more than “tweaking” it a bit.

Cheers,

Keith

2 Likes

What are your thoughts about Go Move 37? I think AI might be able to generate unique and ‘creative’ solutions to some problems, depending on how you define ‘creative’ of course. Whether this can be applied to music I don’t know , but I wouldn’t rule it out.

1 Like

For context, great documentary on this topic. :slight_smile:

I will expand on this further when I’m not travelling…

Firstly, Alphago isn’t Generative AI.

But, also the task of playing a game with,at any turn, a specific and discrete set of rules and possible moves and a clear deterministic end-goal is very different from generating art from a prompt.

Cheers,

Keith

3 Likes

Ah, I thought that might be the case.

I did think about this, music has some rules and parameters of course.
I do wonder how much human creativity is just leaning on past influences with the odd ‘mistake’ or random element that turns out to be serendipitous.

Don’t get be wrong, I’m not a fan of AI music at all, but it does get me thinking about how we define ‘creativity’.

1 Like

More specifically:

Games like Go and Chess have very specific and distinct rules that apply at every step, and there are a very limited set of options.

In Go, there’s, at most, 362 options at each move. That’s a lot of options for a human, but it nothing for a computer.

AlphaGo used a Monte Carlo Tree Search which is an algorithm that is now, around 20 years old. Really the difference between AlphaGo and earlier Go playing applications is the quality of the machine learning, and the speed of the computer running it, which allowed it to consider many, many more options, and in far more detail than it could have in 2006.

Consider that, when humans are taught complex games like Go, chess, or even Poker, they are not taught to analyse the game in the same way as computers are. They aren’t taught to analyse every possible path in the game and weight that path based on the likelihood of success. They are also not taught to apply random factors.

Humans are taught human strategies and to play against other people employing human strategies.

And, yes, applications like AlphaGo are trained on human games, so they will be weighted towards chosing moves based on that, that is only one factor.

Personally, I don’t see a computer playing Go program choosing a move that a human wouldn’t have made to be at all surprising, yet alone “original” or any of the other breathless words used to describe it. Frankly, that is why we use computers for such tasks.

I actually worked on a system several years ago which wasn’t that dissimilar to some of the analytics used by AlphaGo: it used a set of rules and some AI technology to automate and optimise the planning and placement of connections in complex telecommunications networks. I spent over a year working on a project with BT to help the system learn the network topology, circuit placement rules, equipment characteristics and limitations, etc. to drive the model.

It regularly came up with solutions a human wouldn’t have chosen. That was the point.

And this is the crux of it, IMO.

Is choosing heads when most people would have chosen tails “original” or “creative”? That is how I see AlphaGo’s move 37.

As you say, there is, in music, a set of rules and, just like for Go or Chess, those rules can be programmed into computers. And, for the last 40 years or so, computer generated music has been a thing and it’s been getting better and better.

But, in reality, if you look in terms of music (at least Western music) we all mostly use the same notes, the same chords, the same chord sequences the same time signatures, the same rhythms, etc. These form the basic rules of music and is part of what makes it recognisable and enjoyable. In that respect, there’s been almost nothing “original” in music for, maybe, 100 years.

What we consider as “originality” in music isn’t deviation from those rules, but more about how we use them. What instrumentation we use, what expression, the lyrics, the emotion, how the song is arranged and mixed, how we apply effects, dynamics, ornamentation and other expression.

Often it’s just “the sound” which, to be original, often could just be in the song’s production.

Generative AI is, at it’s basis, a tool which is based on mimicking patterns it is fed, whether that is musical patterns, speech, images, or text. A large part of that is literal copying.

Often, with AI, when it does do something remarkable, it’s because there’s actually humans behind the scenes steering it.

That’s not to say that AI isn’t capable of originality, to some degree: witness some of the strange and, often, creepy, grotesque, or nightmarish images AI has produced.

But that’s where we come back to how we, as humans, define creativity. Having 6 fingers on one hand, a body that doesn’t join up with itself, or 3 eyes are the sort of “hallucinations” that AI image creation frequently does and these are, certainly, not the sort of decisions a human artist would normally make, unless it was with specific intent and purpose.

But AI doesn’t do these things with intent and purpose. It does them specifically because it has no intent and purpose, only data. It does them because it has no real understanding of what it’s doing.

Unlike a game, in art there is no specific measure of “good” which can be used to drive the algorithms.

AI can mimic human works, if they exist already. But if it does produce something truly original that is seen to have value, that would be entirely by accident and random.

It’s basically like Shakespeare typing monkeys.

Cheers,

Keith

7 Likes

Thanks Keith - that’s very insightful. Back in the early 90s I wrote a simple checkers game that just evaluated a position using a few simple rules, played every possible move and every counter move then reevaluated the position. I think I went two levels deep (after that the IBM AT ran out of steam!). I don’t think I ever beat it - I wasn’t good at checkers, but I loved computing at the time!

1 Like

Some interesting thoughts, Keith.
When any new tech emerges, there’s often a response of “That’s a poor imitation of what we can already do!” :face_with_monocle:
Look at artists protesting against photography, claiming could never capture the ‘essence’ of a portrait.
Early film was not seen as an art form. It was “vulgar” and “lowbrow” cheap entertainment.
As you say, the vast majority of music we listen to (‘in the West’) is highly formulaic and has been done to death millions of times, each song ‘unique’ of course, because that one person did it slightly differently, at that moment in time.
If I write a song, I simply select a well-used chord progression, choose a topic I’d like to write about, and try to make amusing rhymes with some interesting or controversial imagery. My execution aims for mediocrity. Computers already do that much better and at a much faster rate than me and are improving as I type.
Imagine the pleasure a grandparent might get, who has never learned any musical instrument or done any creative work, being able to use a computer program to ‘write’ a personalised birthday song for their grandchild. That in itself is not only worthwhile, but is also likely to encourage them to pursue other creative activity.
You’re also spot on pointing out how it’s often humans steering the ‘AI’ for spectacular effects. It’s a cool tool, although not nearly as well-developed as the even cooler computer we have beneath our thick skulls :wink:
Now if only there was some way of unlocking that potential… :thinking:
Here are two pieces of human-directed AI art that give me great pleasure

Interesting commentary and examples Brian.

I think AI has a lot of potential, but also a lot of hazards.

Behind every grandparent wanting to write a personalised song for their grand-kids is a greedy tech-bro looking at how to steal content from actual musicians and monetise it by wrapping it up in an AI service they sell to those grandparents.

And, personally, if a grandparent who has spent their whole life and not learned a musical instrument (even how to sing badly) is, suddenly, disappointed because they cannot quickly create a personalised song for their grand-kids, then I’m not going to shed a tear over that.

And consider, if the grand-child is old enough to appreciate the song, they will also be old enough to understand that it was produced by AI with relatively minimal effort from their grandparent.

The skill and gift of music and creation of a special song will simply become another commodity offering from Hallmark, and just as fleetingly appreciated and discarded.

My personal view is that that is not “Art” (big “A”) and that real Art is much bigger than casual use of AI. That is content creation, not Art.

Of course, AI can be used to create Art, as you have shown, In that context it is a tool. But I believe Art is as much about intent and context as it is about the actual content.

Is a 5 year old child’s simple finger painting of their family Art? I believe so: it has deep meaning for the child that produced it, and was made with intent. And the parents who receive it and put it on their fridge will treasure it.

Imagine we now say to kids: no need to break out the paints, just say something into this computer and it will give you something pretty we can print out for your parent’s fridge door.

I also believe that true Art comes with some effort and pain and some uniqueness that comes from the personality of the creator. To quote Elena Rodriquez: Art is the mirror of the soul.

Your own music is a great example: you could, instead of learning to play the guitar, just downloaded backing tracks and put your words to them. This would still have been better Art than prompt engineering an AI to write a song for you, including the words, based on a theme you provided.

But you did write those words and, moreover, you did learn to play two instruments ( the guitar, and your voice) to make those songs.

And, regardless of whether they will win any awards or not, the intent, the effort, and the context is what makes those songs Art. At least in my book.

Another example is bell ringing which is a hobby you and I share. So often I am asked “why don’t we just program a computer to ring the bells for us”, which completely misses the point of why we do it.

The problem for me is not the AI itself which, used well, is a useful tool. It’s the way the industry is trying to deceive us by telling us that Art and Artists have no value whilst, in the same breath, they drive us to create more and more “content” in a form of digital sharecropping which enriches them, and makes the rest of us poorer (financially and culturally).

The first few minutes of this video by Tantacrul (the designer behind MuseScore and Audacity) sums it up pretty well for me.

Cheers,

Keith

3 Likes

Well put Keith @Majik love the opening of the video.

I know this thread is about AI generated Music. But
I think what most people don’t think about is AI isn’t just for generating Art or Music.
It is getting to the point of replacing Doctors, Accountants, Engineers and a very large section of High paying jobs (taxpayers) and funneling that money to a very few people.

There was a News Story the other day saying about 30% of Americans already use AI instead of going to their doctor. This could cut down on wait times to see a doctor and save the user a few bucks, but according to the report AI was gave bad advice and some times out right wrong advice. Which could be Fatal

Even Dr are using AI to save time Why family doctors across Canada are turning to AI scribes — and what it means for patients | CBC News. Is this a good or bad thing only time will tell.

The more people use AI the more people will become comfortable with it and the more Professionals will loose their jobs. The poorer society become.

Just my 2c

1 Like

I think we agree on most of what we’re talking about, Keith, although we might have slightly different views on ‘art’. I’ve always loved visual art and had many discussions on the subject when my son studied it in Brighton.
I mentioned in a different thread- Over a century ago, Marcel Duchamp found a urinal and declared it a piece of art. It caused outrage at the time. Now it’s exhibited in the Liverpool Tate Museum, and I love it…. Art is what humans perceive as such (either the artist or the viewer).

Your 5-yr old’s painting may well be a masterpiece in his eyes, as well as his doting parents, but show someone the picture without any context and they are just as likely to say: That’s not art. It looks like a 5-year old did it! (or maybe even a computer)

When punk arrived in the UK in the 70s, many were shocked at the idea people thought they could make ‘music’ without knowing how to play instruments or how music ‘worked’. Some still feel the same way.

You’re quite right in comparing AI music to a Hallmark card, but that company has been providing joy to millions for over a century. So much garbage, but I’ve had some superb cards in my time. Sure, a hand-drawn personalised card will carry more meaning, but you’d want to be a pretty miserable individual to deride someone for sending you a store-bought, mass-produced card, let alone suggest that it should be discouraged (apart from environmental reasons).
I love the band ‘The Burning Hell’. They struggle to make a living from their music and for a while, offered personalised songs. I’ve commissioned a few, partly to support the band, but also because I believed the recipients would enjoy being thought of and having a song tailored to them.It’s not the same as me writing a song for them, but definitely a step up from gifting a cd or song from Bandcamp.Ordering a personalised song (where you have to at least put some thought/effort into the prompts) form an AI company, is least a step up from clicking the ‘Happy Birthday’ prompt from Fb

Don’t take any of the above as support for the evils of corporations and capitalism, and the business models they run on. If I had my way, we’d have quite radical changes in how the world was run, but that’s a whole other thread…

AI taking away jobs from humans is a red herring in my view.
Technology replacing human labour has been a constant in human development. It frees up time and resources for us to do the things we want to (like making music).
Almost all human misery is caused by unequal distribution, rather than lack of resources/employment.

Except that, generally, is not what happens.

Which is a shame because I would be in favour of a society where that could happen.

Cheers,

Keith

2 Likes

As I said in another thread:

Cheers,

Keith

Cheers,

Keith

1 Like