Radio inspired will.i.am’s career. Now, the musician and entrepreneur is eager to innovate on the dying medium with Radio.fyi, a new interactive AI model. Will shares his philosophy about AI development, the dawn of “intelligent media,” and why he believes Radio.fyi will likely be a big part of radio’s future.

This is an abridged transcript of an interview from Rapid Response, hosted by the former editor-in-chief of Fast Company Bob Safian. From the team behind the Masters of Scale podcast, Rapid Response features candid conversations with today’s top business leaders navigating real-time challenges. Subscribe to Rapid Response wherever you get your podcasts to ensure you never miss an episode.

You’ve been on Masters of Scale as a guest, sharing lessons from your career. For now, I want to ask you right now about how you’re managing in this crazy world of ours. You’ve been on tour with The Black Eyed Peas, but at the same time, you recently launched Radio.fyi to reimagine radio as a media experience. Can you explain what Radio.fyi is? 

Radio.fyi is text to station. So we know what text-to-action is. And in this case, the action in our architecture is the format of radio. So I could ask any question about any topic that is presented to me. Where, you know, this AI host is speaking over this mood music, then throwing to song, then grabbing information as information is unfolding in the internet, and then presenting it back in the form of radio and then throwing to song and then asking you along the way if I have any questions on anything that is presented, and then it could then answer those questions.

It’s part AI assistant and part listening experience. It’s like this blend of information and entertainment? 

It’s what we call the dawn of intelligent media—where the media itself is interactive, engageable media rather than a passive leanback experience. It’s the lean-in, dig-in, dive-deep media experience.

So I don’t have to just read it? I can have a conversation with the DJ, I guess, about the things that are there, right?

Exactly. What we’ve been able to do with our architecture is allow for humanlike banter. If you were to be on a radio station and you could call in and talk to a person, you’re able to talk to the information itself. And that’s pretty profound to be able to say, like, “Yo, what do you mean you’re able to talk to the information? That sounds freaking ludicrous.” Bro, you could talk to the information, bro, you could literally have a conversation with silicone freaking lithium now because of the technology that we’re living in.

So you’ve got your library that you have curated from other places and you’re paying for the rights to that as opposed to through a separate streaming service?

Yeah. We’re not a streaming service like Spotify. It’s radio-like radio. You can’t fast-forward or rewind. Most you can do is pause. It’s not streaming. It’s not à la carte. It’s not on demand like that. It’s like radio.

And because I love radio so much, I want to innovate and help out the realm that changed my life. And what changed my life is radio. What changed Lionel Richie’s life is radio. What changed Stevie Wonder’s life is radio. And so innovating around the space that urgently needs innovation, for our community of music, the artists that make it . . . Radio, radio, radio.

You were the first one to tell me about Udio, which is this amazing AI-based music generator that you’re invested in. You prompt it to create an original song for you. It’s amazing. The record labels are suing it for infringement. How do you think about squaring this sort of awesome tech for creative work and the kind of creative work that underlies it, that trains it? How do you think about those things fitting? 

So, say for example, Spotify had no users, the record companies wouldn’t have sued them. So, the reason why they sued them is because there’s like growth. The reason why they sued Udio is because Udio has growth. So it’s a good thing because that means they’re going to come to some type of agreement faster with artists, and how they trained their stuff on.

It’s going to be resolved, but we’re in tricky grounds now. For example, let’s say, Prince was alive, and you went to Prince and said, “Hey, Prince, who’s your inspiration?” Prince is going to obviously tell you James Brown, Sly and The Family Stone, Jimi Hendrix. If you would ask Michael Jackson, “Hey, Mike, who’s your inspiration?” Michael Jackson’s going to say James Brown, Fred Astaire, Sammy Davis Jr., and Gene Kelly. What they both have in common is James Brown.

And it’s clear that James Brown is their big inspiration. You could see it. So in this case, from human to human, we call it inspiration. Machines call it data set and training, what you trained on. If you would ask like an artist, you’re like, “Yo, who’s your inspiration?” It’s the same shit because we have a neural network, and that neural network was based off of our neural network. So now we’re in touchy ground here because last time I checked, Prince did not give James Brown royalties because he was inspired by him. Like, his pathways were part of his data set. So that’s where it gets blurry, where we are now.

What do most people misunderstand about AI? 

This is the worst it’s ever going to be because it just gets better from here. And that’s being optimistic in every sense of the way: ethically, morally, business practice, legally, compensation-wise . . . This is the worst it’s ever going to be. Being optimistic, let’s be optimistic that the powers that be don’t lead with greed. This is the worst it’s ever going to be . . . That sentence’s it: Let’s be optimistic that the powers that be don’t lead with greed. This is the worst it’s ever going to be.

You’re straddling these sort of multiple careers in music, in tech, you do TV, and I wonder how you manage all that? Like Jack Dorsey was CEO of Twitter and Square at the same time, and he had designated days for each company, right? Like, how do you prioritize all the different things that are part of your world? 

It’s really on the teams you build. The premise is to build a Warriors-type of team. You want a Stephen Curry, but then you need Draymond Green. You need the squad. So the same is here for FYI, where my squad allows me to go out and hunt and find other partners that complement the things that we’re trying to do.

When I’m working with Mercedes, Mercedes needs to know, “We thought Will was just a musician, but he’s more than that. He’s great in our workshops. He always has good ideas.” And then when I’m there, I can’t flood them with my idea faucet. I have to be super strategic and aim it. You could drown somebody with “and then, and then, and then, and then!” You don’t want to ‘and then’ people. So you have to go in there with being super focused.

I worked with someone once who had a ton of ideas and I felt like my job was like to say, “No, no, no, no, no. Yes, that one I like!” Do you do that for yourself where you’re going through your own ideas? You’re writing them down and you’re putting them down and then you’re saying, “No, no, no. Oh yes. That’s the one to strategically push?” Or do you use other people around you to help you do that? 

You have to have a different sounding board. I have a great sounding board outside of FYI because if it’s FYI, a lot of times, like, you could get lost on your own sauce. And what’s grounding, you need an outside team of trusted truth-tellers, and you have to be super you.