What's new

Aiva - Artificial Intelligence Composition: beta starting today

I am just scared we will all be extremely bored in the future.

What will be the point of learning & trying to do anything, when you know a machine can do it quicker and better?
 
I am just scared we will all be extremely bored in the future.

What will be the point of learning & trying to do anything, when you know a machine can do it quicker and better?

Because as Greg intimates above, it is for you, from you and for your own benefit and growth - fuck the machines.
 
The funniest thing here is that unless I'm misunderstanding something, it's fed with music from other sources and spits out a new piece - didn't they do a Star Wars demo?

They say human "intervention" but the starting ingredients, listening to some demos a while back, was clearly existing pieces, so there's no human intervention when you hit go....but there's huge human intervention at the outset, and I'm frankly suspicious that the orchestration is being helped somewhat in the demos.

But no matter....the moment this is used to create something that's clearly heavily inspired by another piece - because that's what it will be used for because that's what people will want from it ("this meets that") - that entity will sue, and in discovery will subpoena the "ingredients" put in to make the piece as well as depose the individuals involved.

And at that point it will become the easiest copyright case to win in history, because you fulfill instantly the intent part, proof that the original work had been heard, proof that it is derivative, and they're quite literally toast at that point. Unless it's only fed with public domain sources, they will be sued by the first person who spots a similarity and because they apparently retain ownership of the music, they'll be liable. Instantly.

Also doesn't this violate most virtual instrument licenses, which rarely license it to a computer and only to a person?
 
Interesting stuff.
I am surprised though that nobody has mentioned that this is nothing new. It has been part of academia since the 70's. It's just got the 'AI' tag nowadays.

Curtis Roads is the man to read on this. I cannot recommend this book enough.. because it completely debunks (in passing, it's not the focus of the book) the whole "robots are taking our jobs (in creativity) nonsense.

https://global.oup.com/us/companion.websites/9780195373240/
 
Honestly, I really hope that they fail miserably. It's bad enough that people can't sit down for a meal without putting their damn phones on the table. Stay the hell away from the arts for the love of god.
 
For the lot of you shouting against the "replacement" of creative human art, consider you have already egregiously took part in this if you have used sample libraries in your productions.

If you have found the need to record live sessions despite the advances of realism and expression in sample libraries, there is proof enough that middle-high tiers of successful automation still have not eradicated the desire for higher creative work to be done.

I feel like it really should be acknowledged that a vast swath of composing work is creating very rudimentary derivations of reference tracks that clients ask for. It seems a bit dishonest to imply that every score a professional composer writes is an unique artistic opus, and it's rare I meet a composer who hasn't had to do a lot of boring/tedious gigs to get by.

The reality is a lot of the work in the industry is mundane. It seems further dishonest then to imply that an AI can't do this job or that it's suddenly an issue that it can "derive" from existing work, considering there is an entire sub-industry of commercial music (called "sound-alikes") literally focusing on getting away with copying as much as possible from reference music.

This is the same way we don't feel the need to record a string family if it suffices to play some block chords in a Kontakt string patch. It's a mundane part and we won't really get anything out of spending the money.

We'll record the strings when it's important to get the expression right. The same goes for composing. It won't go away, it's just going to raise the floor line between what's considered valuable work and what's something mundane and functional enough that it can be produced cheaply through the use of computers.
 
Beta is kind of up, I only have to wait 22 days for my version (or I must invite three friends on social media......I think this shows a lot about it.....), as I am 1200 or so in the list, it seems the interest (from whoever) is big (or it is a good marketing gag).
 
Position 12940 in the queue but can jump to the front if I invite 3 friends :rolleyes:

So because of high demand we are put in a queue. But if we add 3 people to the queue, we get to skip the queue? :confused:
 
Beta is kind of up, I only have to wait 22 days for my version (or I must invite three friends on social media......I think this shows a lot about it.....), as I am 1200 or so in the list, it seems the interest (from whoever) is big (or it is a good marketing gag).

The closed beta has been up for quite some time now. Early beta users needed to be onboarded which is time consuming for the team at AIVA.

I believe there are more people being brought into the beta at the moment, it's worth the wait.

Since I joined in the beta there have been several improvements, new musical presets and new functionality however some of the functionality is still being built to scale with a large number of users.

Most of the demos of AIVA on Youtube were done in a contained environment, things are different when you need to scale the service to thousands of users. Features such as Upload an Influence or Search for an Influence are not ready yet.

The current presets are for Modern Cinematic, 20th Century Cinematic, Tango, Sea Shanties and Chinese Music. Rock/Pop as a preset is expected later this month.

Right now there is a working Piano Roll, the ability to export compositions to Audio (only 192kbps mp3 for now), orchestrated midi or piano reduction midi.

I don't see AIVA replacing professional composers in a hurry but it is a very useful tool for generating musical ideas that can be built into amazing compositions.
 
You can NOT program human emotion into a machine. Our minds are far more complex and advanced than any computer, and given that emotions and feelings are often tied to our experiences, how can a machine have experienced love? it can be programmed to simulate love, sure, but no way, no how will this ever replace a human...at least not a competent one. Why would any of us support this technology anyway? The only fear I see, is that whoever would use such a service is looking for a cheap alternative. I can however see corporate videos, porn, and infomercials using this technology...but no way will a decent director with any integrity, even think about it. My 2 cents, like it or not, this is my stand.
 
I don't see AIVA replacing professional composers in a hurry but it is a very useful tool for generating musical ideas that can be built into amazing compositions.

that's like cheating no? Getting a computer to give you inspiration?...sounds like if this is what you need, then you should not be composing. Music is a form of communication, it's own language...it is life, and experiences, not 1s and 0s. And before you say that computers speak their own language...it's still needs to be input by a human. In the end, it's all a simulation, nothing that AI spits out musically will ever come from a place of emotion, passion, love or experience.
 
Top Bottom