A song that uses AI to clone the voices of Drake and The Weeknd has gone viral
bbc.co.uk
bbc.co.uk
While this music is technically impressive, I stopped listening 30 seconds in. The story behind the art is one of technical engineering. There’s no connection between the listener and the original artists.
It’s the same reason an absolutely impressive forgery is worth nothing and the original is priceless: authenticity is important.
nothing wrong with hiring people.
If you dislike autotune and pop music, i think that’s valid. but it’s really a function of artists responding to consumer preference. there are plenty of high quality niche artists who don’t have the mainstream appeal or success
A widely rumored but undisclosed ghostwriter brings the authenticity into question.
AI is going to make the pinnacle reality TV.
People who listen to this genre still care about them (the stories).
In terms of making money or reaching fame anywhere near the level of the original artists, I think it (and all AI generated art) will come up extremely short because it’s essentially a forgery. To buy concert tickets, T-shirts, posters, etc. you need a story.
But when it comes to buying decisions, wether it’s fine art or a concert ticket, the most money is made by selling a story.
For example, if the music was enough, promoting music would look like playing a recording of songs in public places. But instead it looks like artists talking to reporters, posing for photo shoots, back stage access and producing theatrical music videos.
It is. Radio used to be it but now TikTok is the biggest promotion platform there is and those clips don't even show the name of the artist (since it's always some kind of bootleg so the credit usually points to some random username).
Most music events I go to are raves where I've often never heard of the DJs beforehand. I couldn't care less about authenticity, creator-consumer connection or story. I think it's just two different ways to appreciate it, and I'm not sure "authenticity is important" is as common as you make it out to be.
For example, I've always loved 'Cloudbusting' by Kate Bush which (like many of her songs) is a tiny bit silly, and not for everyone. However one day I looked up the lyrics and found that it's about Wilhelm Reich (orgonne energy) and his relationship with his son. As a result, some of the lyrics became much deeper for me, and I liked it more.
I guess I don't care who a musical artist 'really' is for that authentic authenticity in their music. Do I care who Sting is as a person - his relationships, or activism? Do I care if Richard D James owns a Daimler Ferret Mark 3 Armoured Scout Car? Probably not, when it comes to their music.
The only potential disruption I see is in music as pure entertainment. Just open up YouTube trending and you can quickly understand that the everyday listeners taste is... well, not really demanding. Top commercial artists already have an army of writers and producers milking the pop music structure, in this case we will probably see the explosion of even more uninteresting pop jingles.
However, as you said, authenticity and connection between the artist and the audience. There is a billion and one song writers singing and writing about heartbreak, few of them made it, for reasons not always related to the quality of the music itself.
https://www.theartnewspaper.com/2022/10/07/vermeer-at-the-na...
https://www.cnbc.com/2018/08/16/think-you-can-spot-a-fraud-t...
>There are experts who say half the art sold at auction is fake.
https://www.cbsnews.com/news/art-authentication-forgery-jean...
>So, in a sense, every Beltracchi painting is an original. He just lied about who painted it. He says [he] forged a hundred artists and can do just about anyone.
https://www.cbsnews.com/news/art-forger-wolfgang-beltracchis...
>For nearly three decades Ken Perenyi made a small fortune forging works by popular 18th- and 19th-century artists like Martin Johnson Heade, Gilbert Stuart and Charles Bird King.
>Then in 1998, Mr. Perenyi says, two F.B.I. agents showed up on his doorstep, curious about a couple of paintings sold at Christie’s and Sotheby’s, ostensibly by the maritime artist James E. Buttersworth but actually his own meticulous creations.
>Over the next few years, he says, the F.B.I. continued to keep a close watch on him at his bayside bungalow here, tracking his work and where it sold, and talking to his friends and associates. Though the authorities never charged him, the scrutiny pushed Mr. Perenyi to develop what he calls “a new business model”: openly selling his faked oils as the reproductions of the finest masters.
>Now they are bought by Palm Beach decorators, antiques dealers, professionals, business executives and others who want the look of cultured gentility without the price tag.
>Are they sold as authentic by the people who buy them?
>“During the first few years of trying to market and sell my paintings legitimately, I couldn’t say where they went or what people did with them,” Mr. Perenyi, 63, said in an interview at his home.
>His forgeries, he says, financed an extravagant lifestyle that included European trips, exclusive restaurants, Versace couture and “total freedom.” He says they brought him into contact with mob enforcers, the lawyer Roy Cohn and Andy Warhol, who, he says, bought one of his forgeries, a John F. Peto.
>... Mr. Perenyi, who is open in discussing his life as a swindler, is safe in the knowledge that the statute of limitations for his forgeries has passed.
>Mr. Perenyi estimates that hundreds of his fakes remain in circulation. Occasionally he glimpses one (“It’s like bumping into an old friend”) in an auction catalog or in a magazine. “I miss the addictive thrill of fooling the experts,” he said. “It was great sport for me.”
>A spokesman for Christie’s said that the names of consigners are confidential but noted that a work Mr. Perenyi refers to as his own, a rendering attributed to Heade of two hummingbirds that was sold in 1993, is in the artist’s catalogue raisonné, the definitive compendium of his work.
The difference between Mr. Perenyi’s legal business and his criminal one is that now he makes clear his paintings are reproductions, even though they have the artist’s signature. Fraud applies only when someone actively misrepresents a copy as an original.
>Mr. Perenyi said other clients did not want to speak with a reporter because they present his works as the real thing.
>Authentic paintings attract collectors for a variety of reasons: an artist’s vision, a venerable reputation, investment potential or the ineffable mystique of the genuine. High-end fakes sate a different sort of desire: aesthetic cachet at a fraction of the price.
>Spreading half a dozen of his Guardi replicas across the living room floor, Mr. Perenyi said he developed his artistic technique on his own and learned the forensics by working for a restorer and a frame maker when he was in his 20s. Through extensive research and trial and error, he figured out how to simulate the telltale signs of age: the distinctive spider-web cracking in the paint, the tiny dots of fly droppings, and the slimy green look of old varnish when viewed under ultraviolet light. One of his best, he says, was a Heade-style passion flower that Sothebys sold as a new discovery in 1994 for $717,500. A copy now hangs over his fireplace.
However, many of our narrative characters already are at least part-fiction. I mostly agree with you, but I also suspect that generated art could be attached to narrative characters in a shared fictional cultural anthology to achieve the same end. For example, if someone created a FakeDrake character, and wrote stories about them, and a following around FakeDrake popped up, that character would serve a narrative function and the work of that character could be perfectly authentic. One might expect that the work being secretly produced by a human instead of machine would be viewed as a form of inauthenticity.
1) people know the beat or the artist's voice
2) people know the song because it was promoted to a wide audience
3) it was promoted to a wide audience because the artist had talent (hopefully) plus a compelling image / story that was sold to producers, investors, record labels, DJs, etc.
There are examples of extremely talented musicians playing in public spaces and few stop to listen: https://www.classicfm.com/artists/joshua-bell/violin-busking...
> In total, Bell made $52.17 (£42.18). And this includes a $20 note from someone who recognised him.
Banksy demonstrated this effect too by selling original artwork for $60 in NYC and almost no one bought it: https://www.thedailybeast.com/banksys-biggest-trick-yet-sell...
> At 6 p.m., he closes the table, making only $420 for the day. A limited edition print of Love Is In The Air -- the image of the man throwing a hand grenade of flowers, which was stationed on the center of the table -- sold for $249,000 at Bonham's last June.
> It’s the classic example of art without the frame—and it raises the important question: how much does our experience of art rely on context?
> We linger on The Mona Lisa because it’s framed in the Louvre, behind bullet proof glass and protected by security guards. Outside the Louvre, we walk right past framed copies of the same painting. Context is everything.
https://www.youtube.com/watch?v=VszJPLAtK0U
A couple of other interesting examples also linked in the article.
Up til now, I've been convinced (foolishly, maybe) that there was something special about music that AI just wasn't going to be able to do well. Now, I'm not so sure.
I think modern popular hip hop is probably an easy target - the music itself is relatively simple, and especially for an absolute chart-destroying monster like Drake, there's a huge amount of singing / rapping / voice to use as training data.
I'm somewhat more skeptical on genres that involve complex instrumental playing and more dynamic vocal performances, but I think this track sounding so convincing is evidence that we're making big strides.
On the other hand, there's something about music that I truly don't think AI can touch, and that's this sort of undefined "cool" factor. I'm a rock / metal guy, and there's something just awesome to me about a group of guys up on stage rocking out, but the cool factor is definitely a big part of hip hop too.
Last thought - I'm not sure if this song is as good as I think it is, because I don't really listen to Drake. I recently ran across an AI generated Nirvana song, which was also pretty well done - but it was still just on the edge of the uncanny valley for me. Maybe big Drake fans would feel the same way about this, idk.
Now it looks like it might be one of the first.
The mixing in the song is quite bad and the melody could use a little human touch, but if someone remastered this - it could be a hit!
Like even right now trains could be 100% automated easily. But the risk involved of transporting so many humans that could die if something goes wrong in a freak scenario, and the material damage of derailing a train makes a human supervisor/driver quite viable.
To my ear, it's surprisingly good, but not quite good enough that I'm fooled by it. It might be good enough that if it was slipped into a Nirvana playlist I might not notice - I like Nirvana OK but have never been like, intimately familiar with their catalog.
[1] https://www.thestudentpocketguide.com/2021/05/entertainment/...
1. Is ignorant of the Lex guy's affiliation with MIT, and speculates about how tenuous the particulars could be.
2. Thinks Lex is misusing said MIT connection in unassociated emails.
3. Points to Lex's book-per-week pacing as indigestible given the material, concluding that Lex's persona is fraudulent, and his advocates are all idiots.
4. Uses Lex's courteous response to criticism, and subsequent choice to block said critic, as evidence that Lex is a "fake." There's allegedly supporting evidence of this regarding Kanye West/Vladimir Putin? (I'm ignorant of the reference.)
Full disclosure -- I don't use Twitter and gradually lost the mental wherewithal to slog through the "discourse" within that thread. Let me know if I missed/misunderstood something important.
I'm frankly more confused now. A popular podcaster apparently likes revisiting mostly high school reading material, is polite online, and leans on their institutional affiliation. Okay? Did he also like...shoot someone's grandma or something? I feel like a restaurant patron eavesdropping on a dysfunctional couple having a meltdown over ordering an appetizer that isn't reaaaally about ordering an appetizer. What the hell did this guy actually do?
Still, I think music will be the most resistant to these kinds of trends. Anyone who is familiar with classical music or jazz can tell you that even when the same piece is performed by two very talented musicians, there's often an ineffable and unquantifiable "it" factor that separates the merely incredibly talented from the truly great. Sometimes this is incredibly subtle, or isn't immediately apparent to those who aren't familiar with a genre, but especially because music directly involves the human voice and the movements of the human body, I find this je ne sais quoi much more readily apparent in music than any other artistic medium.
Does that mean AI won't be able to capture it? No, definitely not. But I do think as the marginal cost of producing derivative pop music goes to zero, people will gravitate towards more "organic" forms of expression that are more difficult to emulate through brute force. For example, there's a lot of bad noise music and bad free jazz out there. Most of it probably. I'm sure AI can already produce convincing bad noise music. But since that's already the case, the bar is much higher for AI to produce something that's actually good.
A human can look at an art rendering of a humanoid character and immediately spot symmetry errors, missing fingers, strange limbs or joints, and the AI didn't know any better because it was imagining an artistic rendering from a prompt the same as any other art piece. It doesn't "think" of the physical accuracy of what it's drawing. You might be able to ask GPT-4 to rate the physical accuracy of a bad midjourney drawing and get an honest response, in which case AI already has the ability to do better, it's just not trained that way... yet.
In writing music, humans can conjure and enforce large-scale patterns in music. An AI would currently get lost writing a symphony. An AI might construct an interesting harmony, but it wouldn't dwell on and play with the harmony the way a human composer would. Yet.
I think there's enough demand for non-trivial music and non-trivial artistic representations of humans that these problems will be figured out soon. Whether that involves different training, different neural net architecture or a conglomerate of different sub-neural-nets, or multiple independent neural nets with some filtering others' output.
Here are my two favorites:
177,951 views Apr 15, 2023 (with press coverage)
Wow, viral! /s
Impressive project, don't get me wrong!, I'm just annoyed by the press.. I wish they'd raise their standard of quality and honesty.
I just heard a remix of Jay-Z & Kanye's "Ni*as in Paris," with Biggie as Jay and 2pac as Ye. The AI voice-swap sounded 95-98% like Biggie & Pac. But it did highlight how the flows & cadence of Jay & Ye are their own, while Biggie & Pac have their unique deliveries as well.
It didn't necessarily instill in me a sense of fear for artists in the future, or inspire me to want to find the best songs & musicians & rappers to use this tech out on, or anything like that. But it was well-executed, and illuminated other aspects of these artists beyond the very recognizable tone & pronunciation from each respective artist.
I always thought Biggie had the best flow. His delivery almost has that slightly-off-beat drag, even if he can pretty much self-quantize if he wants to (like in Notorious Thugs). Maybe these AI remixers will hone in on such details soon enough, sort of quantize the deliveries in such a way to be purposefully off-beat and unusual (almost like a Dilla beat with his signature style that makes it a point to be off-the-grid so to speak).
In this case, the protections afforded by IP law for music and right of publicity for people are pretty different, and it would be a mistake to assume perfect coherence between the two. That a feature, not a bug.
My question: do you own the likeness of your own voice? How does this differ from a very good impressionist?
If I find an artist to be questionable, it is "ethical" to consume generated music of this nature based on their prior work?
Are we on top of a slippery slope and going downhill towards the obvious next steps?
Can we even trust anything at this point?
No, and I feel badly for anyone who thinks that is a new thing.
See: Milli Vanilli
downvotes incoming from real musicians and audio engineers but for less than 100 bucks you can get your vocals 80% of the way there.
imo it’s amazing for songwriting
you’ll be able to create more music as a result. and get clearer feedback on the end result
imo the mixing engineers are better used to make your ideal vision stronger. also make the drums hit and ensure the guitar doesn’t clash.
if you just use them to make vocals sound pleasant, it’s going to be less valuable
Same goes for what an AI might write that isn't technical information I might make use of.
Same goes for images.
Simply I don't care what "art" might come from AI because there is no human involved. I'm a human - I want human things, not things that undermine my humanity.
Also people way underrate how little there has always been in any form of art that wasn't simply copy-pasting or data-driven and that includes most of the content we consume, but especially modern tv/movies/music.
Peace.
I already do.