00:04:15.420using AI-generated versions of rappers who have died
00:04:18.540so that those more established rappers
00:04:21.160can be on tracks with up-and-coming artists.
00:04:23.860Obviously, this kind of thing has gotten a lot of negative attention as being disrespectful or just plain creepy.
00:04:30.940But Timbaland told Forbes that he feels the industry and consumers need to see a more serious, well-intentioned, and transparent effort to integrate AI technology into hip-hop, saying,
00:04:41.360I don't want to be afraid of what's going on. I want to be the guy that figures out a solution.
00:04:45.140But on the other hand, you have rappers like Ice Cube, who said that AI will make artists lazier and less creative and threaten to sue any platform or person who promotes an artificial intelligence-generated version of his voice or likeness.
00:04:59.140And then there's Lil Wayne, who I think had the best response to the rise of AI in hip-hop, which is basically, bring it on.
00:05:34.300we're basically seeing artists have that very same conversation they were having about autotune
00:05:38.800in the 2000s now with AI. And the story of autotune is one where I think something that
00:05:44.340was once perceived as being a gimmick at best, or at worst, a way to manipulate human artistry
00:05:50.680in a way that kind of creates this vibe that like, oh, like humans are expected to sound
00:05:57.360like perfect machines. I think that was like a big concern that folks were grappling with
00:06:02.080around autotune. There was this big concern that autotune was going to de-incentivize artists
00:06:08.660kind of needing to be good or needing to work on their craft or perfect their craft because you
00:06:12.920could just hit a button on a computer. But today, nobody thinks of autotune as just a gimmick.
00:06:17.780Nobody thinks of autotune as something that, you know, de-incentivizes artists to be creative or
00:06:23.840perfect their craft. It's just something that is totally commonplace in audio today. It's used in
00:06:29.140all kinds of interesting and creative ways, because in the end, creative artists were able
00:06:33.200to make something cool and innovative to create a cultural shift. And so I think that story,
00:06:38.400as we talk about things like the intersection of music and AI, that story seems even more
00:06:43.900relevant right now. So let's talk about how Autotune came to be. Autotune was originally
00:06:49.540developed by Dr. Andy Hildebrand, a research engineer. And weirdly enough, he was not working
00:06:54.880in the music industry when he came up with the idea for Autotune. He was working in the gas
00:06:58.980industry. He worked for Exxon and created a complex set of algorithms to interpret sonar
00:07:04.020generated data to locate oil deposits deep underground. But he was also a musician, a flute
00:07:09.300player, and he always wanted to find a way to be more involved in the audio space. Him creating
00:07:13.900autotune was kind of a fluke. A wife of his colleague was joking about how bad her singing
00:07:19.720voice is, and she was like, oh, if only there was a technology I could sing into that would help me
00:07:24.400sing in tune. And I guess this idea really stayed with him. And he got to thinking, could the tool
00:07:30.100that he used in the oil industry also help correct pitch? The answer was yes. So you might be
00:07:36.880wondering, why is this such a big deal? Well, before auto-tune, there were ways to correct pitch,
00:07:42.300but it was so time-consuming that it was generally not considered to be worth it.
00:07:47.320Meanwhile, auto-tune was incredibly good, easy, and fast to correct pitch. So it was a total
00:07:53.080game changer. This is something that kind of comes up in podcasting a lot too. You know,
00:07:57.660when you're podcasting, the best take is generally your like most natural take. And so when you say
00:08:04.160something and it just sounds so right, but maybe you cough or stumble or, you know, trip over your
00:08:11.700words a little bit. If you redo it, it's never going to sound as good. And so one of the reasons
00:08:16.960why this technology is such a game changer is that it allows you to keep that natural first
00:08:22.380best take, but just polish the parts that don't work. As he described to NPR, the singer's first
00:08:28.880take is often their best. It's full of vitality and emotion. After their take, the producer will
00:08:33.960announce, great, but the second phrase was pitchy, so let's do it again. Well, now the singer's worried
00:08:38.920about pitch and has to focus on intonation, and the vitality and emotion are gone from their
00:08:43.620performance. What Auto-Tune lets the producer do is fix the first take, which makes a lot of sense.
00:08:49.240In 1996, he implemented the algorithm on a custom Macintosh computer and presented the results at the National Association of Music Merchants, a trade show for audio professionals, where it was instantly a massive hit.
00:09:02.240I can understand why because, yeah, the frustration of having to completely redo a take because you're not able to just go in and perfect it is very frustrating.
00:09:12.740So I can understand why when this was released, people were like, oh, my God, huge game changer.
00:09:17.060So auto-tune becomes a thing, but at this point, it's kind of treated like a trade secret.
00:09:23.080Engineers were using auto-tune to discreetly correct pitch without really advertising it,
00:09:28.160until someone sweeps in in 1998 with shimmer and bangles and flowing long black hair.
00:09:36.700Can you guess who I am talking about?0.81
00:09:44.660So Cher's Believe in 1998 was the song of the late 90s, right?
00:09:51.860Not only was it a musical departure for Cher as an artist, but it's a song about transformations.
00:09:58.580So she, as an artist, is going through a transformation in her career, but also singing about the importance of transformation.
00:10:06.200And I think it's an arguable that Believe is Cher's most important song.
00:10:11.320I think it would be really easy for an artist like Cher to kind of become a nostalgia act from her Sonny and Cher days of the 70s, but she didn't.
00:10:20.820She kept evolving and kept innovating.
00:10:23.100At the Grammys that year, the song was nominated for Record of the Year and Best Dance Record, winning the latter.
00:10:28.460And it was a huge commercial hit, too.
00:10:31.120Believe is one of the best-selling singles in music history.
00:10:34.000And this is actually something that is, like, pretty hard to do.
00:10:37.420The music scene of the late 90s was really fragmented, but Believe broke through as this global smash.
00:10:43.700It's also just really endured culturally, like there's a Sex and the City episode where Charlotte thinks her boyfriend might secretly be gay.
00:10:50.260And one of the pieces of evidence is that he puts on Cher's Believe when they're in the kitchen.
00:10:55.700And so I think it's important to really highlight what a big deal this song is.
00:11:02.080like if you've ever been on a dance floor and heard Believe come on, it is like an emotionally
00:11:08.700resonant experience to be dancing to Believe on a dance floor. Like the song just does something
00:11:14.200to you when you hear it. And I also think it's interesting in that it's a song about someone
00:11:19.380who has been dumped, but that act of being dumped is like a badge of honor. It is like usually in a
00:11:27.120love song the person who has been dumped is like sad or you know grieving a lost relationship
00:11:35.440grieving that they weren't enough for this person who dumped them but in believe she is like
00:11:41.240righteous she is not saying like i am sad i'll get over this she is saying i will transform this
00:11:48.700moment is going to be a moment of transformation for me and it's just one of those things like
00:11:52.660the line, do you believe in life after love? Like it just, it's one of those lines that I can't
00:12:00.520believe is not an established phrase or idiom because it just hits so perfectly. So in the
00:12:06.680song Believe, autotune was used to create unnaturally rapid corrections in Cher's vocals.
00:12:11.660They basically removed that natural slide between pitches and singing, in effect, creating that kind
00:12:16.960of robotic voice that you hear in the song. Believe's producer, Mark Taylor, originally
00:12:22.520did not want to tell people that he had used autotune to get that sound. No, not because
00:12:27.400autotune had a negative connotation yet, but because he wanted to protect the method as
00:12:31.880a trade secret. So the team initially came up with a whole cover story that if they were
00:12:36.260asked, they were going to say that they got that robotic vocal effect by using a vocoder
00:12:41.100pedal. A vocoder, if you don't know what that is, it's similar to autotune, but different.
00:12:45.900Apparently, Cher's label wanted her to remove the autotune effect, and she flat out refused.
00:12:50.880They even started calling autotune the Cher effect.
00:12:54.560So while all of this is going down in 1999, T-Pain is just starting his musical career.
00:13:01.920So when T-Pain was three, he got interested in music because a family friend, jazz singer and producer Ben Tankard,
00:13:08.500allowed him to spend time, quote, twisting knobs at his recording studio.
00:13:13.280At age 10, T-Pain turned his entire bedroom into a little music studio with a beat machine and a keyboard and everything.
00:13:20.460T-Pain actually got his start as a rapper, but after being discovered, he decided that he wanted to sing instead of rap.
00:13:27.060It's the early 2000s, and T-Pain had been looking for a way to make his voice stand out.
00:13:31.180He hears the Dark Child remix of the 1999 song If You Had My Love by Jennifer Lopez, which, by the way, I love that song, which uses a little bit of autotune.
00:13:40.200He's also really inspired by the R&B greats like Teddy Riley, who use things like talk boxes and vocoders, which are kind of similar to Auto-Tune, but a little bit different.
00:13:49.300T-Pain records his debut album, Rapper Turned Singer, in 2006, and it gets to 33 on the Billboard 200 and a certified gold.
00:13:57.760He releases his second album, Epiphany, which is kind of like T-Pain's magnum opus.
00:14:04.020It's like the album that you think of when you think of T-Pain.
00:14:07.180The album includes the song Buy You a Drink, which is probably his biggest hit.
00:14:10.980The song peaked at number one on the U.S. Billboard Hot 100, making it his highest charting single as a lead artist.
00:14:17.640So something to know about this is that when that album first came out, it wasn't like people were criticizing T-Pain as like a gimmicky artist.
00:14:25.680The reviews were admittedly mixed, but people were talking about him as a serious artist to watch, not like some kind of a joke, one-trick pony.
00:14:34.200You know, something that you really need to know about T-Pain is that he legitimately saw autotune as his thing, right?
00:14:41.000He did not see autotune as popping onto a trend or a gimmick or a joke because it was his way of exploring his own voice as a singer.
00:14:50.720But everybody sees how successful T-Pain is at doing autotune, and they all start jumping on the bandwagon.
00:14:57.860Snoop Dogg does it in 2007 on his song, Sensual Seduction.
00:15:01.640Lil Wayne does it in 2008 with his song Lollipop.
00:15:05.060Kanye West releases the album 808s and Heartbreak in 2008,
00:15:08.660a project that T-Pain actually worked on.
00:15:10.980What's interesting about that is that I've heard T-Pain talk about 808s and Heartbreak,
00:15:15.860even though he was, I think, a consultant on the album.
00:15:19.120I think he felt some type of way about how that album was received.
00:15:23.540He talks about how he didn't feel like Kanye West did autotune correctly on that album.
00:15:29.860it sounds like T-Pain thought that the critical praise that 808s and Heartbreak got
00:15:33.720should have gone to him as a pioneer of the technology.
00:15:36.960What's interesting about this is that people associated T-Pain specifically with autotune
00:15:41.460and a kind of cheapening of the music industry in general.
00:15:45.880But he was legitimately trying to use this technology to create something unique as an artist,
00:20:09.880Well, Nico Case does not use autotune, and in 2006, she told Pitchfork, quote, I'm not
00:20:15.100a perfect note hitter either, but I'm not going to cover it up with autotune. Everybody uses it
00:20:19.440too. I once asked a studio guy in Toronto, how many people don't use autotune? And he said,
00:20:24.300you and Nelly Furtado are the only two people who've never used it in here.
00:20:28.040Even though I'm not into Nelly Furtado, it kind of made me respect her. It's cool that she has
00:20:32.300some integrity. And so I think what we really are seeing here is like autotune being connected to
00:20:40.160like a lack of integrity or a lack of authenticity. This was the 2000s at a time where I think that
00:20:46.340there was a real binary between, quote, real musicians, real artists, and like poppy,
00:20:54.860bubblegummy kind of like gimmicks. And so I think part of the backlash was about this,
00:21:01.320was a response to people being like, oh no, these new fake musicians, they're not real musicians.
00:21:07.860They're using autotune. They're fakers. Real music is like real music, people using their0.85
00:21:13.060real voice. And I can understand that sentiment, but honestly, there's nothing natural about the
00:21:21.840human singing voice. Like your voice is a tool. It's an instrument. And so this idea that any
00:21:28.300manipulation of that means that you're like not a real artist, I find sort of interesting.
00:21:32.800Yeah, that adjective real comes up on the show a lot, right? Just to separate us from them pretty often. Kind of makes you wonder who's the us and who's the them in that scenario.
00:21:44.740So in 2009, the same year that Death Cab for Cutie shows up wearing those blue lapel pins to protest Auto-Tune, Jay-Z released the lead single of his album, The Blueprint 3, DOA, Death of Auto-Tune.
00:21:58.420And he said that he thought that Auto-Tune was like an overused gimmick.
00:22:02.000In that song, he calls T-Pain out by name.
00:22:04.760Also, in 2010, Time Magazine includes Autotune in their list of the 50 worst inventions, on the same list with Agent Orange and subprime mortgages.
00:22:43.460So just like in conversations that we're having around AI and whether or not images that have been manipulated using AI should be labeled, which, by the way, we think they should,
00:22:51.260there was even a call to have it be labeled whenever a live performance used auto-tune.
00:22:56.280Singer-songwriter David Mindell started the Live Means Live campaign, and he wanted there to be a
00:23:00.960logo that said Live Means Live to let the audience know that when that person is performing, no
00:23:05.980auto-tune is being used, no backing track is being used. It is 100% live. And I think that goes back
00:23:12.300to what we were just talking about. Like in the 2000s, there was this, I would argue, false
00:23:18.560dichotomy between authentic musicians and inauthentic musicians. You know, you had a lot
00:23:24.400of rappers calling other rappers quote ringtone rappers which was like a a negative term for a
00:23:32.140rapper that was like too poppy too bubblegum just trying to get money as opposed to like real
00:23:38.880rappers who are doing something else entirely you have a lot of artists being called out for using
00:23:44.340you know backup vocals tracks in their in their live performances in 2013 Beyonce famously was
00:23:51.560lip syncing to a vocal track when she performed at Obama's inauguration. And then she had to do0.88
00:23:57.300a press conference where she belts out the national anthem acapella and is like, any questions? Just
00:24:03.420to prove that she actually can do it. Which, by the way, I've often found conversations around
00:24:08.840having a backtrack going for a live performance. That inauguration, I was at that inauguration and
00:24:16.780it was a very cold day. The weather does stuff to your voice. Like, I don't know. I feel like
00:24:21.720if you're performing at an outdoor important inauguration, there's no shame in using a
00:24:26.900backtrack for a big moment like an inauguration to just need to sound perfect. And if you use a
00:24:31.760backtrack to accomplish that, I don't think it's the end of the world. Like, I think it's sometimes
00:24:35.840the accusations are a little much. And I do think that some of this climate around inauthentic and
00:24:43.080authentic musicians does to me to seem like a way to criticize Black musicians and poppy women
00:24:50.320musicians. Like you can't help but see people who see themselves as like real authentic rockers
00:24:56.660calling out people who use autotune. And it's curious to me who got caught up in the autotune
00:25:05.440criticism. Because like in Radiohead's 2001 album Amnesiac, they use autotune. Tom York
00:25:12.580described the use of auto-tune on that album as, quote, auto-tune desperately tries to search for
00:25:17.860the music in your speech and produces notes at random. If you've assigned it a key, you've got
00:25:22.360music, right? And so nobody was calling out, when all this auto-tune criticism was happening,
00:25:29.060nobody was calling out Radiohead for using it. It was interesting to me who got called out publicly
00:25:33.780and who didn't. In that really good pitchfork piece I mentioned, they put it this way,
00:25:38.580Much of this anti-autotune sentiment presented the idea that the technology is a dehumanizing
00:25:43.440deception foisted upon the public. Attempting to deflect this angle of attack, Hildbrand once
00:25:49.740offered up an analogy with a generally accepted form of everyday artifice, asking,
00:25:54.120my wife wears makeup, does that make her evil? Perhaps because of Cher's involvement in1.00
00:25:59.080Autotune's debut on the world pop stage, critics have often connected pitch correction and
00:26:03.520cosmetic surgery, comparing the effects to Botox, face peels, collagen injections, and the rest.
00:26:09.600In the video for Believe, Cher actually looks how autotune sounds. The combination of three0.55
00:26:14.640levels of enhancement, surgery, makeup, and that old trick of bright lights that flatten the skin
00:26:20.000surface into a blank dazzle, means that her face and her voice seem to be made out of the same
00:26:24.980immaterial substance. If the Believe promo was produced today, a fourth level of falsification
00:26:30.480would be routinely applied, digital post-production procedures like motion retouching or colorizing
00:26:36.260that operate at the levels of pixels rather than pores, fundamentally altering the integrity
00:26:41.140of the image. The taste for these effects and the revulsion against them are part of the same
00:26:45.960syndrome, reflecting a deeply conflicted confusion in our desires, simultaneously craving the real
00:26:52.380and the true while continuing to be seduced by digital's perfection and the facility and
00:26:57.380flexibility of use that it offers. That's why young hipsters buy overpriced vinyl for the aura
00:27:02.740of authenticity and analog warmth, but in practice, use the download codes to listen to the music on
00:27:08.520an everyday level. So I definitely agree with that take. And I think that our response to
00:27:13.940autotune and the rise of autotune culturally does say something about us, that we want it both ways.
00:27:20.140We want the slickness of something that is clearly digitized while also being able to call it out
00:27:26.420for being inauthentic or being repulsed by it.
00:27:54.880It is commonplace across all genres of music, even though we associate it more heavily with hip-hop, it is in all different kinds of music.
00:28:03.760And autotune is not only used for things like pitch correction or making a singer sound better.
00:28:09.080Artists also use autotune in all kinds of unique ways.
00:28:12.440One of my favorites is the Kate Bush song, A Deeper Understanding, which was originally released back in 1998.
00:28:18.180Kate Bush has said that the song is a critique about our relationship with technology, saying,
00:28:22.820this is about people. Well, about the modern situation where more and more people are having
00:28:27.540less and less contact with human beings. We spend all day with machines, all night with machines.
00:28:32.920You know, all day you're on the phone and all night you're watching the telly. Press a button
00:28:36.360and this happens. And this idea of someone who spends all their time with a computer,
00:28:41.080and like a lot of people, they spend an obsessive amount of time with their computers.
00:28:45.000People really build up heavy relationships with their computers. So the song initially had this
00:28:50.240computer voice on the track. In the original version, it's voiced by her son. But when she
00:28:55.280revisited that song in 2011, Pitchfork reports that she used autotune to make the Siri-like
00:29:00.580voice of the computer sound like a guardian angel, offering surrogate solace and counterfeit company,
00:29:06.740saying, hello, I know that you're unhappy. I bring you love and deeper understanding.
00:29:12.100So she used a heavily criticized technology to critique our relationship with technology.0.74
00:29:35.140So Dan points out that venture capitalist Chris Dixon has this concept about disruptive
00:29:40.100technologies, that disruptive technologies always start out by looking like a toy.
00:29:44.760So when they come out, like the iPhone, I remember someone saying, I don't want to get an iPhone. It looks like a toy. It looks like a phone that a child would use. And I think there's something to this idea that technologies that are going to be disruptive and really take off as the next big thing, they do often start out by looking like toys.
00:30:03.440And so when something looks like a toy, it can be dismissed while also being phased in as something that's going to be disruptive. And so I think that autotune was dismissed as this gimmick, right, this joke. And now it's everywhere, even though so many people initially dismissed its use.
00:30:20.760Yeah, that's a really interesting point. And it does really connect to the analogy you were making to AI earlier, right? When AI first became a big thing, we were all using Dali to create images that were like sort of jokey.
00:30:38.220And it was like, and like similar with chat GPT, just asking it questions for novelty, for funsies. And so it really does have this quality of feeling a little bit like a toy, but I suppose, you know, quietly on laptops around the world, people are using it for much more serious purposes.
00:30:59.640Yeah, I really think the comparison between autotune and AI as technologies in so much that they will transform these creative industries. And I think the reason why I connect those two conversations in my head is because as we're having these conversations about how AI is going to transform the creative spaces like hip hop, music, the arts, film, television, a lot of those conversations have been really pessimistic.
00:31:27.920And I totally get it, right? Like, the striking screenwriters and actors are not wrong to be really concerned about the way that AI will impact their industry. Those concerns are absolutely fair and grounded.
00:31:42.500However, I think that there's always going to be people who find ways to use technology to create something better. And so I think that we're at this pivotal point in the conversation around AI and how it's going to shape those fields where the question is, do we want this technology to be used to exploit or to innovate?
00:32:05.060I think the screenwriters are saying, let's see how this technology can be used to innovate. If AI is going to help me punch up a script, that's great. The powers that be should not bake my exploitation into that model. I think that that's where we're at.
00:32:20.660And so this might surprise people, but I am kind of positive, kind of cautiously hopeful about how we see this play out.
00:32:28.720Because if there's anybody who can find a way to innovate with new technologies, it's creatives, it's hip-hop artists, it's people like T-Pain and people like Cher, it's innovators.
00:32:38.000And so I really want those people to be the ones who are leading the way in terms of how new technologies do shape creative fields.
00:32:47.600How can they be used to innovate, not exploit?
00:32:49.560And it really comes down to something that tech historian Claire Evans said in the very first ever episode of There Are No Girls on the Internet.
00:33:01.200There have been many instances in the history of music when a new technology has come along that ostensibly is there to displace the musician.
00:33:07.800For example, the drum machine or the synthesizer, you know, these are tools that were designed to replace session musicians with an easier, cheaper version kind of automation of their labor.
00:33:18.340In fact, even in the 80s, like the British Musicians Union tried to ban synthesizers.
00:33:22.960But what artists and musicians did was instead of allowing those tools to replace them, they took control of them.
00:33:29.420And, you know, they took drum machines and they took synthesizers and they invented Detroit techno and they invented new wave and they invented hip hop and they invented, you know, electronic music as it exists today.
00:33:40.520in its many manifestations, they kind of took the thing that was threatening them with displacement
00:33:44.700and incorporated it into what they were doing and made it essential to who they were and used it to
00:33:49.760invent something new that they were integrally as human beings involved with. And I think that
00:33:54.660that act of kind of like, I don't know, like jumping on the grenade or something is like a
00:34:00.080really beautiful thing that artists always do willingly or unwillingly when they are faced
00:34:05.640with new technology. And I think when new technology comes along, you always have that
00:34:09.500choice. Are you going to let it displace you? Are you going to let it intimidate you? Or are you
00:34:13.560going to take it, you know, jump on it, find some new use for it and make it part of who you are
00:34:19.140and give it back to the world in a new form? That choice is always present. And I think that's what
00:34:24.380I try to do in my work across the board. And I think it's the only way that we're going to kind
00:34:29.420of keep on top of all this technology. And I think it's also very human. I think it's what
00:34:34.920people always do. We are always trying to create systems of meaning and beauty out of what is
00:34:43.800coming up ahead. And I think that will never change. Claire really put that so beautifully.
00:34:49.740And that is a sentiment that is echoed by one of my favorite bands, Daft Punk,
00:34:53.260who used autotune in their song One More Time. They said,
00:34:57.480a lot of people complain about the musicians using autotune. It reminds me of the late 70s
00:35:01.440When musicians in France tried to ban the synthesizer, what they didn't see was that you could use those tools in a new way instead of just replacing the instruments that came before.
00:35:09.980And so I think that's where we are right now.
00:35:12.960We are at the crossroads with so many technologies of, are we going to use this to innovate or are we going to use this to exploit?
00:35:20.680And that is the question that we are asking right now about new technologies.
00:35:24.080And that is a question that innovators like Cher and T-Pain answered with a resounding innovate.