will.i.am: AI Heralds New ‘Renaissance’ for Creatives (YouTube Video Transcript)

(00:00:00) Your YouTube transcript will appear here (00:00:00) It's a renaissance. The same way the (00:00:03) Renaissance, (00:00:04) you know, catapulted, (00:00:07) you know, dimensionalized the creative (00:00:09) industry in the past, the Renaissance, (00:00:12) this new renaissance to do it for this (00:00:15) era that we're in. So, it's it's the age (00:00:19) of the hyper creative. So there's a (00:00:22) copyright law change that the UK (00:00:24) government is thinking about making to (00:00:27) ensure that the models and the AI (00:00:30) companies here can innovate quickly and (00:00:32) use copyrighted data more freely with (00:00:34) less risk of legal (00:00:37) response to that. And so the artists (00:00:40) including Damon and Annie Lennox put out (00:00:42) this silent album in February as a as a (00:00:45) protest to say this is this is going to (00:00:48) this is going to be a problem. (00:00:50) >> Yeah. So those the artists that are (00:00:52) protesting have every right to right (00:00:54) because say for example Coca-Cola wants (00:00:57) to license your music (00:00:58) >> or McDonald's or Starbucks or (00:01:01) >> you know some politician running for (00:01:04) office they want to use your music (00:01:07) >> for their campaign. You have to license (00:01:09) that and the artist has to sign off on (00:01:10) that. So if if artist music are in some (00:01:14) like overall catalog and the artist (00:01:17) didn't say yes, you could (00:01:19) >> use it for however you need to use it, (00:01:20) whether it's to play it, associate it or (00:01:24) train on it. The artists have to should (00:01:27) do that. The thing the artist also has (00:01:30) to know is we are at a point where they (00:01:33) don't have to train on our music. (00:01:36) theory is theory (00:01:38) and the model is going to get so (00:01:40) sophisticated that it's just going to (00:01:42) train on theory (00:01:44) and because it can predict. (00:01:46) >> Yeah. and songs can be broken down to (00:01:49) predictions (00:01:52) um based on theory. (00:01:54) >> But but licensing of copyrighted music (00:01:56) is possibly a solution for these AI (00:01:58) models that are training and so far a (00:02:00) number of them are trained the (00:02:01) allegation is a number of them are (00:02:03) trained on copyrighted music without (00:02:06) compensating those artists and licensing (00:02:08) could be the solution. (00:02:09) >> Licensing is a solution. M (00:02:12) >> um (00:02:14) but what is the value of those (00:02:17) recordings and those copyrights? So (00:02:19) that's the that's the heavy one. We (00:02:21) licensed our music to Spotify. They (00:02:23) don't value it (00:02:26) the way that it should. Then that brings (00:02:28) upon another another conundrum like a (00:02:32) song is worth how much? A stream. So, (00:02:35) you could license it, but let's let's (00:02:38) hope that the folks that are, you know, (00:02:43) divvying up its value and its worth love (00:02:47) it and appreciate it and price it for (00:02:50) what it actually costs (00:02:52) um as far as time. (00:02:55) But the model (00:02:58) is uh it's getting better and better and (00:03:01) better. (00:03:03) It doesn't have to (00:03:06) learn from the past to create. (00:03:10) >> Are there specific guard rails that you (00:03:12) think about that should be implemented (00:03:14) when it comes to protecting creatives in (00:03:17) this time of momentous change as a (00:03:20) result of AI? (00:03:20) >> Yeah. (00:03:22) First off, the reason why we're having (00:03:26) this situation and the conundrum that (00:03:29) we're in is because the data practice (00:03:32) overall, (00:03:34) the data practice, not just for music, (00:03:37) everything needs to be thought and (00:03:39) humans need to be protected. You you (00:03:42) came at it from like, hey, what about (00:03:43) music? What about your job? (00:03:46) How you interview? (00:03:48) How Bloomberg says, you know, we really (00:03:51) like how you do that. We want a avatar (00:03:55) agent to do exact. Do you own how you (00:03:58) ask questions? So right now the easy way (00:04:00) to look at is like you musicians in your (00:04:02) songs like how Yes. Everybody (00:04:06) >> everyone needs to be protected. (00:04:09) the sellers, [clears throat] (00:04:10) the storytellers, (00:04:12) the lawyers, the finance people, (00:04:15) human resources at the company, (00:04:18) everyone. It's a human like urgent (00:04:24) um matter to where we get we need (00:04:26) governance and regulations ASAP, not to (00:04:30) stifle creativity, but just to make sure (00:04:32) the business model, it's the business (00:04:34) model that's the issue, not the tech. (00:04:36) Bro, (00:04:37) >> in 10 years time, are we listening to (00:04:38) more AI created music than human created (00:04:41) music? How do you think about that? (00:04:44) >> You like fruit? (00:04:45) >> Yeah. (00:04:46) >> Uh, you go to the supermarket, do you (00:04:47) say, "Hey, where's the organic oranges?" (00:04:51) >> Sometimes, (00:04:52) >> but then you just buy oranges. (00:04:54) >> Yeah. (00:04:54) >> So, if you're not buying organic (00:04:55) oranges, then what the were those (00:04:57) oranges that aren't organic? (00:04:59) >> Yeah. (00:05:00) >> Why would you have to call these organic (00:05:02) and those just oranges? (00:05:03) >> Yeah. If those are if these are organic (00:05:05) and that what are those? Only reason why (00:05:08) I'm saying that it's just processed. (00:05:09) >> Yeah. (00:05:10) >> To the point where you need to identify (00:05:12) organic from non. The same will be for (00:05:15) music. The same will be for everything. (00:05:18) >> It'll be on us the consumer to identify (00:05:20) what is synthetic, what is AI made (00:05:22) music, and what is humanmade. Or does it (00:05:24) need to be labeled? (00:05:26) >> I'm trying to go I'm trying to elevate (00:05:28) this conversation to everything. (00:05:31) I'm starting with oranges to organic (00:05:32) oranges to music to organic music to (00:05:36) interviews to organic interviews. (00:05:38) There's going to be a premium in terms (00:05:40) of the human experience in this new AI (00:05:42) world. Things like live experiences, (00:05:44) live concerts, live sport. Is there (00:05:47) going to be a premium for that kind of (00:05:48) entertainment? (00:05:49) >> It's called lived experience. The AI is (00:05:52) not living. So when you go see your Yes, (00:05:56) there's going to be lots of music that (00:05:58) there's gonna be AI bands. That's true. (00:06:02) There's gonna be like an AI freaking (00:06:04) Bruce Springsteen. There'll be an AI (00:06:07) Lana Del Rey. There'll be a AI freaking (00:06:11) um (00:06:13) Adele. Yeah. And it's going to sound (00:06:15) great. People like it. And they're going (00:06:17) to you're going to go see it perform (00:06:19) live. and the team that makes the live (00:06:22) performance gonna be freaking Yep. It's (00:06:24) gonna be dope. And then there's going to (00:06:27) be human lived experience that you're (00:06:29) still going to love. It's not doom and (00:06:32) gloom for the music industry. I'm more (00:06:34) worried about my accountant than I am my (00:06:36) drummer. (00:06:39) I'm more worried about my lawyer than I (00:06:41) am my sing the singers that sing with (00:06:43) us. Cuz that's law, memory, and numbers. (00:06:47) There's a hundred billion dollar (00:06:50) industry, if not more, launching (00:06:53) freaking agents and co-pilots (00:06:56) than there are the record industry (00:06:58) launching AI artist. Get out of here, (00:07:01) bro. Let's break it down to the real (00:07:02) real. What's going on? (00:07:05) So, I'm more concerned about my (00:07:07) assistant than my guitarist. (00:07:09) >> Yeah. And you've worked you've worked (00:07:11) clearly you work in philanthropy in (00:07:13) music in design across all of those (00:07:15) different sectors as well as tech of (00:07:17) course. Of course all those different (00:07:18) sectors where you see where do you (00:07:19) expect to see the most disruption as a (00:07:22) result of AI or is it across the board (00:07:23) the same? (00:07:24) >> The creatives are going to be all right. (00:07:26) Hey we're going to be cool. (00:07:28) They like this. They like this. They (00:07:31) want the heart. And stop chasing the (00:07:34) algorithm on TikTok. Like wrench your (00:07:36) soul out. People want that. We're gonna (00:07:39) be cool, but as cool as we are, we're (00:07:42) going to be writing songs called Where (00:07:44) Did the Humans Go? And then they're (00:07:46) going to come to our concerts to hear (00:07:47) that.