Harvey Mason Jr.
π€ SpeakerAppearances Over Time
Podcast Appearances
For all the reasons I just said, the importance of music and the ability for us to tell stories and change hearts and minds, I think the human component to that is really, really dang important. So a little nervous that we haven't got it sorted out, but I'm also optimistic, to be honest, Nilay, because... Human creators are not like computers.
We take the chaos and the uncertainty in life and the stuff that AI hates, and we make incredible art from it. And we are able to dig down deep into some of our most creative spaces and pull out the next amazing thing and make great art that I don't think any computer is going to match our head.
As much as we're nervous and we worry about it, I don't think you can tell me that AI can create songs in the key of life or Nevermind or Illmatic. I don't see it happening. So I want to make sure we're able to use AI and I'm not an AI hater. I think it's got great potential. I've been using it for eight or nine years in different forms. I've always been an early adopter to new tech.
So I'm with it. I get it. But we have to make sure human creativity is protected. And we have a chance to make sure we're remunerated properly. We have proper approvals and make sure it's credited properly. Those are the things that are really important to me.
Same way that we let music that have samples be eligible or we let music that has synthesizers or auto-tune or pro tools be eligible. It's a technology, an evolution that has allowed people to do more, create differently, think differently, make sounds we've never heard before.
So for us to draw a line in the sand and say, if you used artificial intelligence, you are ineligible, would be, I think, short-sighted. And I think it would also cut down on a lot of the music that's being created and submitted. Also, where would you draw the line?
I mean, there's AI and so much of the software we use now for even analyzing and doing mixes and sound design, not even just the generative AI that's making music, but The finer point is that we'll allow AI to be utilized, but we're not going to honor AI in the sense that if AI is performing a song, it's in our rules. We will not give the performance an award.
If AI is writing the song, we will not give an award for the songwriting component. For example, if you wrote a song, it was a beautiful composition, it was the best lyrics, best music, best chord progressions, and you had AI sing it, you could submit it. It's not going to win for singing. It could win for songwriting.
Conversely, if you had AI just write a song as a great song, but some vocalists sang it or rapped on it and they performed the heck out of it, I'm not going to penalize the human creativity that went into that. I'm not going to give it an award for the songwriting, but I will give it an award for the performance. That's the way our rules are currently. I'm sure it's going to change.
This stuff is moving so quickly. But for now, that's how the academy is moving.
I believe there needs to be an understanding of what these models are training on. And I'm not sure exactly to what level it will come down, whether there's compensation payment or crediting. But I do think something has to change. And I don't believe that people's personally copywritten material should just be used or accessed by everyone to do anything they want with.
So we have to come to a bottom line and understand there's fairly trained models out there, people that are licensing groups of music or catalogs to train AI on, and I think that's a good place to start. But there's a lot to talk about. That's probably a whole other show that we could dive a little deeper into.
You know, it's a complicated subject. So I think there's some real talks that need to happen around that. Should it be compensated on the training side? At least we need to know what it's training on, how much of it's being used. So there's a lot of nuance to that question.
There's no perfect solution or magic bullet to any of this stuff, especially the speed at which it's moving. We are really proud of the legislation that's been introduced and passed in a couple of different states, but now we're pushing for federal legislation with the No Fakes, No Frauds Act in the House and the Senate.
That's not the intention. That's not our desire.
Again, there's a lot of nuance. There's no perfect bill. None of these bills are exact. Everyone is trying to compensate and accommodate the needs of a lot of people that have concerns and fears.
Of course, we don't want to prevent someone from impersonating Elvis, but we do want to prevent people from impersonating artists, singers, and using their voice without any form of payment or approval whatsoever. or the right crediting. And these bills are starts. I'm sure they'll be revised. I'm sure there'll be new bills, new things enacted.
But right now, we've got bipartisan, bicameral support. There needs to be some legislation that supports and protects human creativity, human artistry. So for us, it's a first step.
It does, but also laws protect certain usage of other people's voice even if it's another human doing it. You can't pretend to be an artist and then go monetize that in certain ways. So there's laws in the books that already prevent that from happening.
It's all going to continue to be a mess until we get a story out because, yes, it's difficult.