I was trying to decide whether to write about the newly released CRTC report on the Rogers outage in 2022 or to share some deep insights about late 80s pop hip hop duo Milli Vanilli and why I think their story has resonance for technology ethics and generative AI.
It’s summer. Writing about music felt like it would be more fun!
It’s often hard to explain the myriad ways in which technology, people and social systems intersect and impact each other - never mind also trying to find a compelling story to wrap these abstract issues around. The case of Milli Vanilli meets both of these objectives.
The Scandal
Firstly, for everybody under 45, let’s start with who is Milli Vanilli?
Fab Morvan and Rob Pilatus rose to fame in the late 80s as the pop hip hop duo, Milli Vanilli. In a few short years, they became global music sensations. Three of their five singles went to number one, they sold millions of records, and they even won a Grammy. It was a ‘rags to riches’ success story of two young, Black men making it in the world of pop music.
But, what Milli Vanilli ultimately became known for was an enormous scandal.
It started with this concert - where it became clear they were lip synching when the music hit a glitch. Shortly afterwards, their producer, Frank Farian, revealed that they weren’t just lip-synching live events. Other musicians had sung all the songs on their albums as well. The news shocked the music community. The pair was stripped of their Grammy and they faced a vicious public backlash that ultimately led to Pilatus’ untimely death from a drug overdose in 1998.
Clearly fans, the general public and the Recording Academy of the US ,who preside over the Grammys, all felt deceived by Milli Vanilli. Yet, the story is a bit more nuanced and complex than it might first appear.
Recently a documentary was made that shared additional details about the scandal. The film claims that executives at their label, Arista records, were aware of the deceptive plan, which was conceived by their producer, Frank Farian. Farian was the one who said they would not sing on the album. He had done a similar thing with the disco group Boney M. Morvan and Pilatus had signed a contract, had been given an advance and were indebted to the label. It was not so easy for these young men to walk away. This is not to say there were without any blame, but there was a very big power dynamic at play.
Today, it’s unlikely the Milli Vanilli story would be the type of career ruining scandal it was back in 1991. Lip-synching is no longer a career killing sin. Superstars from Mariah Carey to Beyoncé have reportedly done it at live events with no lasting negative repercussions.
I want my MTV
The 1980s marked a change in the music industry. It was the launch of the MTV (MuchMusic) generation where it was no longer enough for an artist to sound good, they also needed to look good. Pilatus and Morvan were selected for ‘stardom’ because they had a particular look that Frank Farian knew he could build a brand around. They were attractive, they had some dance moves but they were not strong vocalists. Looking the part is pretty much taken for granted these days when it comes to having a career as a musical artist.
However, it was arguably the rise of the music video channels that made it important for artists to be photogenic. The medium shapes the messenger.
Bands are increasingly built for marketability. Think about boy bands - from the Backstreet Boys to NSYNC to One Direction. These are young men picked for their looks, their moves and their personas, over their musical prowess. Think about all of K-pop. In a way, the audio is the least important aspect of the overall performance. A lack of vocal proficiency is no longer addressed with human replacements but with technology.
Auto-Tune, AI and audience expectations
Auto-Tune first launched in 1997 and has been digitally boosting and smoothing out vocals ever since. Virtually everyone uses Auto-Tune now.
Rick Beato is a Youtuber who covers the music industry. In a video entitled “How Auto-Tune destroyed popular music”, he talks about this article in Billboard expressing the anxiety songwriters are feeling in light of generative AI, and then he covers the role Auto-Tune has played in the music industry over close to three decades. There is now a ubiquitous over-reliance on this tool across pop music, according to Beato. Auto-Tune has set audience expectations and reshaped the sound of pop music to become more computerized.
“The general listener doesn’t know the difference and they frankly don’t care. I don’t think they are going to care when musicians and songwriters are replaced by AI. Really the only question is, who gets paid for it?” - Rick Beato
Beato's warning is a reminder of the ways in which we have all been conditioned for the moment we are having right now with generative AI. We could make a similar case about conditioning when it comes to the past 20 years of search engines, or the way texting has primed us for communicating with ChatGPT.
If you want to see a pretty amazing example of digital remixing and vocal correction in action, check out this audio engineer singing a very bad rendition of Toto’s Africa and then fixing it with Auto-Tune and digital editing. Jump to 9:30 to see the final result - it's incredible to hear given where he started.
Baby, I forgot your number
Back to Milli Vanilli. Baby, Don’t Forget My Number was one of their hits and one of a large number of songs throughout the history of pop music that highlights the centrality of the phone to dating culture. Blondie's Call Me is an earlier example of this while Carly Rae Jepsen’s Call Me Maybe, is a more recent example. Before the days of swiping right, this is how people connected.
Who remembers a phone number these days? How many phone numbers could you recall unaided? Probably not many.
We have pretty much fully offloaded this task to technology. It’s no longer important. We don’t even think about trying to memorize a phone number. The purpose of having the number has also changed. It’s now mostly a mechanism for texting vs talking.
What other tasks do we feel are essential right now that in the next 20 years will look as passé as remembering a phone number? Once again, we see the ways in which technology and culture work to inform each other in this song about phones, which, like the Milli Vanilli scandal itself, feels very outdated.
Gotta blame it on something
“Three decades later, it’s surprising that such a media firestorm took place over lip-synching, which has been commonplace, to varying degrees, for many years.” - Variety (2020)
The case of Milli Vanilli highlights the inequities of who is judged most harshly when an ethical incident occurs. Arista records and Frank Farian, for their parts in the scandal, went relatively unscathed. Farian, who passed away last year, had a long and lucrative career in the industry. Arista is now part of Sony Music. Morvan went on to do a stint in radio and TV, attempted a much less successful solo-career as a musical artist in the early 2000s (where he did actually sing) and also worked in various film and TV projects. The artists who sang the tracks for Milli Vanilli were paid for their work. Yet, they did not get the level of compensation or fortune that comes with fame. In other cases happening around the same time, digital tools made it possible to 'sample' vocals from artists who were not compensated.
Auto-Tune has also reshaped artists real-time relationship with their own voice. Artists came to 'hear' themselves in relation to the technology, blurring the lines of the 'raw' vs the 'tuned'.
"The crucial shift with Auto-Tune came when artists started to use it as a real-time process, rather than as a fix-it-it-in-the-mix application after the event. Singing or rapping in the booth, listening to their own Auto-Tuned voice through headphones, they learned how to push the effect. Some engineers will record the vocal so that there is a “raw” version to be fixed up later, but—increasingly in rap—there is no uncooked original to work from. The true voice, the definitive performance, is Auto-Tuned right from the start." - PitchFork
Have we redefined the rules of what it means to be authentic? Or the threshold for what is considered deceptive? When it comes to expectations in the music industry, the answer appears to be - yes. Digital technologies, adopted over time - including but not limited to Auto-Tune - have reshaped the social contract and ethical norms of what is deemed permissible by fans. The importance of a live show and the overall spectacle of performance have made lip-synching less of an issue. Audio is just on piece of the music experience.
If we think about generative AI, which promises to write the song, play all the instruments, arrange and mix the tracks, make the requisite music video - in essence, to do all things digital that now entail 'making music' - what role do humans have in this process? Perhaps Frank Farian got it right; the human's role in this particular socio-economic and cultural arrangement is to be the embodied persona, the 'front person', LARPing a musical artist, instead of making actual music.*
In her book, Cloud Ethics, Louise Amoore asks:
How are algorithmic arrangements generating ideas of goodness, transgression and what society ought to be?
The story of Milli Vanilli, their lip-synching scandal and the subsequent rise of Auto-Tune provides a glimpse into how past socio-technical arrangements set a “new normal” in the music industry. The once scandalous has been made banal. It’s a case we can learn from as we consider the widespread use of generative AI and whatever becomes our next normal.
Blame it on the tech. Yeah, yeah.
Fun Fact - Cher’s Believe which debuted in 1998 was one of the first songs to blatantly use Auto-Tune as a effect.
This PitchFork article is a fascinating history of Auto-Tune and its inventor's pivot to music from a background in algorithmically mapping oil fields.
*On a more hopeful note, I believe live music performed by artists who do actually play the music will be highly valued.
By Katrina Ingram, CEO, Ethically Aligned AI
Ethically Aligned AI is a social enterprise aimed at helping organizations make better choices about designing and deploying technology. Find out more at ethicallyalignedai.com
© 2024 Ethically Aligned AI Inc. All right reserved.