AI Digging...

Thought to open a discussion about AI generated music for sampling.

For one, it appears to be copyright free, no human owning it. You just tell it, "make more shit like those MacDermot and Axlerods songs" and BOOF! You get a batch of reinvented songs of the OG songs but as original compositions in themselves. Its crazy.

I think this will be the future of sampling, what do you think?

"Make me a beat in the style of Ill-Green making a beat in the style of David Axelrod".

Seriously, the whole AI thing breaks my head every time I read more about it. The US has said AI content cannot be copyrighted, the UK says the opposite. Yet US courts have ruled that a Mavin Gaye 'groove' is copyrightable. Regardless, there's no guarantee that AI content will not inadvertently infringe an existing copyright.

And as alluded to in my previous post, if it's that easy to generate specific styles of music, who needs artists at all, at any stage of the process?

Record labels have a hard time controlling their "content creators"... AI is great for them... make an artificial brand and persona... and yeah... they don't need flesh and blood that is unpredictable... while an AI might have problems... it won't miss shows... complain about overtime... do heavy drugs...

The question will be, who is all this artificial content for? Will people gravitate towards it, or is there something special about an in person, imperfect, performed show... or are we striving towards something else?

MPC-Tutor wrote:"Make me a beat in the style of Ill-Green making a beat in the style of David Axelrod".

This made me laugh.

But its real talk. I personally see sampling AI music a good thing because you are not infringing on any flesh and blood artist but might be a nightmare for the upcoming and starving artists because now corporations can be done with human content and just generate music from their own offices not even having to search, pay, and communicate with a person who creates music.

Unfortunately it won't do anything against the false copyright strikes people get hit with for their beats.

The generated audio doesn't come from nowhere - it's a result of training models on existing material - so there will be loopholes in copyright law which are more than likely being written now.

It'll only be the super rich artists who fight such cases... the rest of us will carry on making music however it suits. It's quite an exciting time for samplists, with stems technology and AI related music generation.