What's new

Aiva - Artificial Intelligence Composition: beta starting today

I envision a boutique market in the near future for * Fully Organic Scores * A novelty for a premium aural experience.
 
intriguing. Will be interesting to see where this goes.



pack it up, boys; time to go home.

Reading the comments is fascinating. I wonder how many would change had it not been mentioned this track was made by an AI.

Oh cool, this is a good first pass. The director has some notes for Aiva--can you conform to the new cut, line up the timpani with the drone sunrise shot, and producers were thinking to make it more "Egyptian". They want the harp to sound more "minor chord".
 
Oh cool, this is a good first pass. The director has some notes for Aiva--can you conform to the new cut, line up the timpani with the drone sunrise shot, and producers were thinking to make it more "Egyptian". They want the harp to sound more "minor chord".

"This cue sounds too much like discovery. Can you make it sound more like foreshadowing?"

"Character A really changed Character B's career forever. We need you to represent that in the music."

When AI can handle matters of abstract storytelling, I think society has bigger changes coming, outside of music.
 
How does this compare to Hexachords Orb Composer, which also touts itself as an AI compositional tool?

I actually quite like the idea of AI as an assistant. I've seen how Algonaut's Atlas can blast away writer's block with the click of a button. Powerful stuff.
 
Do you think once millions of jobs in the transportation industry are lost to self-driving cars and loading/unloading-robots, there will be new political parties campaigning on the promise to outlaw employing AI for jobs that a human could do? A pro-human-workers-party?

Have any of you seriously considered to switch profession to being a programmer if the machines ever take your job? *insert Southpark reference here*

I think I probably have a good decade or two left before they're coming for my job as a freelance artist, but is it smart to try and make the transition at the last possible moment? Probably not...
 
The piece sounds nice indeed, but my first thought was if this kind of AI could possible compose in an even more innovative way than a human being that is bound to his own musical roots? It could maybe try out some crazy things with a mouseclick and helping us composers find some new fresh ideas, who knows?

Maybe it could become a kind of experimenting machine for us composers - does someone know how this works?
 
still new to this, is Aiva a software plugin or is it a cloud service that requires internet?
It's a cloud service - all the processing is done on Aiva's servers. You can expect a ~2 minute piece of music to be generated and rendered in around 5-10 seconds (I'm sure that's liable to change depending on the server load and internal processes).

I envision a boutique market in the near future for * Fully Organic Scores * A novelty for a premium aural experience.
I'm sure I read somewhere that it raises your anti-oxidants and reduces the appearance of aging!

A pro-human-workers-party?
This already happened; they were called Luddites. It may happen again!
 
Well, it's better than what I can compose :(. Would certainly feel weird if I was persuing this as more than just a hobby.

Then you simply need to study more scores. Because that's what these AIs are doing. AIs are pattern recognition machines. They're being fed large amounts of information about existing music, and "learning" the patterns. There is no real creativity here, just regurgitation of patterns that a sufficiently large data set would indicate are acceptable or pleasing.
 
Then you simply need to study more scores. Because that's what these AIs are doing. AIs are pattern recognition machines. They're being fed large amounts of information about existing music, and "learning" the patterns. There is no real creativity here, just regurgitation of patterns that a sufficiently large data set would indicate are acceptable or pleasing.
What is "real" creativity in this sense? Pattern matching + random variation + culling mechanism for deciding what works and what doesn't. And how do we imagine "real" creativity differs? There is perhaps an intention that separates creativity from regurgitation from an indefinitely large data set but how do we define that intention in a rigorous way?
 
Top Bottom