Menu
What's new
By:
Filters

AI - Next steps we can take [SPINOFF for non "next steps" conversations]

BTW I just tried Suno AI. We are totally fu...d.
(If I just arrived late to the party sorry!)

This is the first time I'm hearing something AI musically and sonicaly .... "good" "well produced" ?
The writing, the vocals, the instrumentals. It is, for the first time for my ears, getting there.
It's really hard to admit but it might be usable good very very very soon.

I'm starting to feel like graphic designers felt a year ago more or less ... (what? you are asking what is a graphic designer? It's one of those extinct jobs from the beginning of the millenium.)

I'm not talking about how artistically valid is Suno's output (up for a long heated debate I guess).
I'm talking about how convincing it is.

Not happy. Future bad.
Hope they and every other AI company get sued to hell for copyright issues, the copyright assignment matter doesn't get solved leaving AI music not commercially exploitable, and all this disappears.
 
Last edited:
I've been around long enough to see the music industry change several times. When I was still in training jazz was on the rise again, there were loads of jobs for orchestrators and arrangers, etc. By the time I went into the work force a lot had changed. Just a few years after that sample libraries started to come out and I began receiving less and less calls for studio sessions. Just a few years after that the iPod came around and record sales plummeted. Not too long after Spotify came out and it became hard to sell records even at gigs! Last jazz festival I played, about 10 years ago, I sold zero, but most people were kind enough to say they'd love to buy it, but have nowhere to play it and is it available in Spotify?

My composing career went down a similar path. The point is, things always change, it's the natural process of life. It might not necessarily be technology that disrupts our livelihood, it might be a ban, a dictator, a change of taste, a natural disaster, etc. Humans are good at adapting.

As for AGI, even if it does happen, I doubt it will be anything human-like. It'll probably be more like the computer in Star Trek i.e. very smart, very fast, but not conscious as we understand it. Biological imperative plays too big a role in what we are...
 
The main issue with these things is the input into the models are all music based.

Real musicians take from everything in life. Art, food, design, landscape, sex… dumb but true.
Until music models can digest information beyond a musical dataset, it will never reach human levels.

Yes, it can regurgitate a Frankenstein piece, but can it write music that has something to say? We’ll see…
 
rpaillot said:
It's astonishing...


These AIs not only "borrow" or draw inspiration from the notes but also from the sound itself. It might not sound perfect, yet some of the instruments sound impressively realistic. Like it came from a LSO recording (with a lower sound quality of course)

Who produces the sound? Primarily, the original musicians from the recordings the AI is trained on, but also the sound engineers, and even the acoustics of the venues (credited to the architects who designed these spaces). Additionally, the craftsmanship of those who built the violins and brass instruments plays a role as you can imagine it’s a lot of money for a LSO recording. Did Udio compensate all this ? No.

This differs from previous shifts in the music industry like streaming, synthesizers, etc.
These previous changes didn’t steal anything without permission.

Sample libraries for instance were done during orchestral recording session, musicians were paid and on some librairies musicians are getting residuals.

In a recent orchestral piece on Udio, I heard a horn solo that seemed to mimic Richard Watkins playing the horn.

Did he give permission to be digitized like this ? No..

Its the most substantial steal in human art history, operating on a level unprecedented.

These websites must be shut down asap like the mp3 piracy websites around 2000s… it’s the only way.
Shutting them down, lawsuits. When Udio and Suno are shut down and their futur competitors will try. Shutting them down again.
Click to expand...
This is the thing, people saying it doesn’t sound human are objectively wrong because the underlying sounds clearly have a human source, therefore have authorship tied to them. They’ve basically “sampled” the entirety of human recorded music and refuse to acknowledge it or remunerate a single composer, musician, engineer, or vocalist.

Pretty astounding.
 
Mike Greene said:
Even at its most optimistic, we’re small potatoes compared to what I believe is their real goal of selling zillions of monthly subscriptions to wannabe songwriters around the world.
Click to expand...
Not disagreeing, but I've heard this goal described several times and I've gotta say..I'm having difficulty seeing the market for this. There are already a million ways (many free) to dabble around with music for those who want to.

Even if the market does take off, it does seem like something that would peak pretty quickly, especially when the new world order of composers discover what we already know: No-one listens to the music you make. ;)

I kind of liken it to my initial love with AI generated imagery ("this is amazing I'm gonna use it everywhere") but the novelty wore off quickly and I went back to using stock art libraries. It's often easier to browse a selection of items rather have to engage to brain every time there's a need for media.

I'm also warmed by a comment earlier today from my teenage daughter when I queried her views on AI music: "Ergh. We all hate it. Music should be made by humans". Though I fully accept that's a survey of one.

All this said, I remember arguing passionately that MP3 would never replace CD. So there you go.
 
Alex Fraser said:
Mike Greene said:
Even at its most optimistic, we’re small potatoes compared to what I believe is their real goal of selling zillions of monthly subscriptions to wannabe songwriters around the world.
Click to expand...
Not disagreeing, but I've heard this goal described several times and I've gotta say..I'm having difficulty seeing the market for this. There are already a million ways (many free) to dabble around with music for those who want to.

Even if the market does take off, it does seem like something that would peak pretty quickly, especially when the new world order of composers discover what we already know: No-one listens to the music you make. ;)

I kind of liken it to my initial love with AI generated imagery ("this is amazing I'm gonna use it everywhere") but the novelty wore off quickly and I went back to using stock art libraries. It's often easier to browse a selection of items rather have to engage to brain every time there's a need for media.
Click to expand...
I agree completely. Their intention is to cash in on the "Look! I made a song!" crowd, but I think they will fail. The vast majority of tech startups fail, and I don't think it will be a shock if these do, too.

Making music is fun largely *because* effort is required. There's no joy in making a song if all you did is type in a few prompts. Heck, we have an entire section here (Members' Compositions) that proves people enjoy, and take in pride in, the work.
 
I would assume the startups competing to show off their AI tech are mostly hoping for an acquisition deal, and not to be the one actually interfacing with the wider market in the long term.

Imagine Windows and Mac having built-in music stations. Turn it on from the taskbar, change the station/genre from a pop-up menu, enjoy your infinite source of background music.
Now, either there will be ads thrown into the mix, or you'll be paying for it with your Apple Music/Microsoft Copilot subscription.

Either way, I think generative music will be quite popular as an alternative to playlists and radio stations for the general consumer. A big way in which music is consumed these days is as a background to an activity, such as studying, working out, programming, dining, cafe music, spring cleaning, dinner party, etc. etc.
For most of these applications, we're not necessarily looking for songs to sing along to, or that are familiar to us. We just want a vibe that we can start on command and then stop it when we're done with it. Hence why spotify puts so much emphasis on their playlist options.
Generative music has shown it is already capable of being used to build that kind of platform, so I'd expect that to be one of the first ways it becomes widely adopted and integrated as a consumer product.
 
At some point could we have a dedicated AI forum so some of us could (if we wanted) mute the whole thing?
Not that it's not important but it is tiring. It's been non-stop in all of the discord servers I'm in lately.
If you look at Perspectives right now literally 49 of the last 50 posts are all about AI.
 
chillbot said:
At some point could we have a dedicated AI forum so some of us could (if we wanted) mute the whole thing?
Not that it's not important but it is tiring. It's been non-stop in all of the discord servers I'm in lately.
If you look at Perspectives right now literally 49 of the last 50 posts are all about AI.
Click to expand...
No. You must remain in the Train Man's Station for the rest of digital time like Neo had to in the Matrix. Also, the irony isn't lost that you just posted about AI. 😂

This is amazing Simon Franglen has posted on VI-Control! Total legend!

And honestly, what the fuck? Why are people worried about this shit? Just write music and love writing music. Writing music is the best fucking thing ever! AI is here to stay and so are we. Well, I am (I can't speak on behalf of the weak 😂) 😍
 
Where I *can* see Udio being used is within existing social media creation tools, such as CapCut. I've just spent a (slightly terrifying) half hour getting Udio to write silly songs to send on to friends. Maybe the "market" isn't so much the traditional music distro, but one where we're flinging endless amounts of social media and content at each other.

That doesn't sound like my idea of a good time, but I'm not 15 anymore.
 
Alex Fraser said:
Where I *can* see Udio being used is within existing social media creation tools, such as CapCut. I've just spent a (slightly terrifying) half hour getting Udio to write silly songs to send on to friends. Maybe the "market" isn't so much the traditional music distro, but one where we're flinging endless amounts of social media and content at each other.

That doesn't sound like my idea of a good time, but I'm not 15 anymore.
Click to expand...
I’m quite certain social media is one of the main markets for all AI.

Innocent fun, yes. Silly birthday message songs, comical pictures, fantastic videos etc. Delight will be had by all!

More nefariously, “flood the zone with shit” disinformation. Has there ever been a technology so well designed to create bullshit by the truckload? We’ll see soon enough.
 
Pier said:
Click to expand...

Yeah, agreed. Except this particular issue is easily handled through licensing. So while I continue to believe copyright is an important, maybe the most important leg in this battle, it can’t be the only one.
 
Pier said:
I don't think whoever owns the rights of The Beatles will be licensing their music for AI gen.
Click to expand...
Maybe not. But the problem is solved by licensing is the point. And there’s always a price that whoever controls the copyrights (master rights) could be convinced to license. Or someone who wants to license it could just buy the masters at whatever inflated price so they control them. So I would say it can’t be counted on as a fire wall. It’s a leg of the stool but only a leg.
 
Jdiggity1 said:
Either way, I think generative music will be quite popular as an alternative to playlists and radio stations for the general consumer. A big way in which music is consumed these days is as a background to an activity, such as studying, working out, programming, dining, cafe music, spring cleaning, dinner party, etc. etc.
For most of these applications, we're not necessarily looking for songs to sing along to, or that are familiar to us. We just want a vibe that we can start on command and then stop it when we're done with it. Hence why spotify puts so much emphasis on their playlist options.
Click to expand...
I remember Spotify got into a bit of hot water when it turned out a lot of their playlists like ‘Jazz for Cooking’ didn’t have a single notable name. It wasn’t Chet Baker, Miles Davis, Bill Evans. It was stuff like “The John Smith Trio.” It turned out Spotify had created that music to collect royalties from the millions of streams from people playing it in the background. These no names had more streams than classic jazz artists. I think they stopped it when an article dropped about it.

I wonder how Spotify feels about AI music — I wouldn’t be surprised if they bought Udio to be honest. They’ll wanna crush any subscription-based music generating/listening platform.
 
SimonFranglen said:
The problem is that there are genuinely super talented creative AI engineers who believe that their role is to transform humanity into the passengers on the Wall-E Space Ship, giving us "more time for leisure and creative activities" while destroying those very things.
Click to expand...
Oh yeah I’m from an era pre computer and it was forecast that they would create more leisure time and less work. All they ended up doing was making better slaves of us all so corporations could increase their profits and a select few would amass more wealth than is imaginable.
 
This eloquently written article really nails the core issues for AI and creative use:

www.muddycolors.com

The A.I. Lie

I recently had a job which did not go well. Eight weeks in, I was mired in sketch revisions and being asked to start over for the fourth time and so I decided to cut my losses and walk. When I shared the experience, most replies I received were from folks commiserating that, unfortunately, that’s
www.muddycolors.com www.muddycolors.com

a few quotes:

"It has become standard to describe A.I. as a tool. I argue that this framing is incorrect. It does not aid in the completion of a task. It completes the task for you. A.I. is a service. You cede control and decisions to an A.I. in the way you might to an independent contractor hired to do a job that you do not want to or are unable to do. This is important to how using A.I. in a creative workflow will influence your end result. You are, at best, taking on a collaborator. And this collaborator happens to be a mindless average aggregate of data."

"Though they might dramatically speed up or replace parts of a workflow, the short and long term costs are appalling to me. It is removing my own hand, the single most valuable asset I possess, from the creation of my work. Even used for prelim work or “inspiration” as I’ve heard other folks occasionally say, I see it as contracting out something crucially important to the lowest bidder. The early stages set in motion everything that follows. They are what the entire creative work is built around. To hire that out to a robot is to value the robot’s decisions above my own, existing only to paint-by-numbers a design that I did not create. And I believe outsourcing segments of the workflow also degrades one’s abilities, making you more dependent on the service. So if you value your mind, spirit, and vision at all as important components to your work, this should be a non-starter. And if you don’t value those things, you might consider another line of work altogether, because that is what makes an artist’s career possible."
 
Last edited:
You must log in or register to reply here.
Top Bottom