• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

The future of AI in music

Phoney

Senior Member
Joined
Nov 1, 2021
Messages
370
Likes
236

The world of AI is making its way into the music business.. The tools that producers and writers will have access to will be very advanced and can make their job a whole lot easier. There are possibilities of making very good remasters of old songs or even creating new tracks using vocals from retired artists or artists who have passed away. This is just the start, and the music business is facing some real challenges already. As someone who doesn't really care all that much about modern music, I can't really see how it will affect me much personally, but isn't this a cause of concern for up and coming musicians (for example producers)? Imagine spending like 10+ years of your life learning how to produce music (which is complicated) and then AI happens..
 

robwpdx

Active Member
Joined
Aug 8, 2020
Messages
268
Likes
370
AI in music, or more properly machine learning, has been around for some time. Hit Science, and the Music Genome Project are examples. John Cage was noted for using random processes in composition. I'm sure other composers have used machine learning since.

A lot of music has been digitized and it is likely machine learning was used in the audio feature extraction for pattern matching. It is used by Shazam and for copyright enforcement by YouTube and others. They haven't got video feature extraction to that level, so video copyright enforcement is usually based on the audio.
 
OP
Phoney

Phoney

Senior Member
Joined
Nov 1, 2021
Messages
370
Likes
236
AI in music, or more properly machine learning, has been around for some time. Hit Science, and the Music Genome Project are examples. John Cage was noted for using random processes in composition. I'm sure other composers have used machine learning since.

A lot of music has been digitized and it is likely machine learning was used in the audio feature extraction for pattern matching. It is used by Shazam and for copyright enforcement by YouTube and others. They haven't got video feature extraction to that level, so video copyright enforcement is usually based on the audio.

I didn't know this, thanks. But it seems like it's seriously picking up pace now, the tools are getting more advanced very quickly, and I assume the availability of such tools are also going up quickly. It seems to me that up until recently it hasn't really been that much of a problem for the music business? I'm trying to imagine just how easy and quick it can be to generate new tracks in the future, and just how good they can turn out to be..


Not that famous modern musicians usually write their own music and doesn't use ghost producers anyways.
 
Last edited:

Dunring

Major Contributor
Forum Donor
Joined
Feb 7, 2021
Messages
1,148
Likes
1,235
Location
Florida
On that note, Google just opened up MusicLM for signups. It's lets you describe music and turns it into listenable tracks. I just signed up, so don't know how long it takes to be accepted. I clicked yes to contact me and other boxes to increase the chance of getting on the list (even if you post on sites, clicking you're a journalist will increase chances).


There's some examples of it you can play in a browser here:
 
Last edited:
Top Bottom