(The video embed doesn’t work but I will update it if it works. You can still click the link at top to watch. If you don’t speak bad French like me, you can also use subtitles auto-translated by Google.)
Myself and author and professor Melvin Bukiet, with whom I worked on the Akron Smash Group art project, were interviewed recently by Akron Mirror, which is a podcast put out by the Akron Library. In it we talk about AI and the collapse of believable reality online, and what role libraries should play in the preservation of human knowledge against those who might seek to overwrite it for their own ends.
A tragic end has taken the real human band the Velvet Sundown is based on before their time:
Or that’s what “they” want you to believe. Maybe it’s all a cover up on top of a cover up… The world may never know the truth of this incredible, stupid story…
I noticed that some people, including journalists, seem to have misinterpreted the latest CBC reveal. User Robert1950 on the forum Guitarscanada.com posted a link to the article with the thread title, “Creator of AI Band Velvet Sundown revealed.” No comments so far have accumulated.
A user on Bluesky also seems to have misinterpreted the substance of the actual reveal, referencing the article, “Well the creator just unmasked himself… and he is Canadian.”
A website called TenAsia takes that a step farther and misattributes a quote from the “real” band to me:
“The creator, Tim Boucher, a web platform safety and policy expert from Quebec, stated in an interview with foreign media, “Velvet Sundown exists somewhere between being fully human and fully machine,” and expressed his desire to challenge music copyright and identity through artistic provocation using AI.”
Granted, these are all easy mistakes to make, especially for people who are not professionally following — or inventing — the story. Or, for people who haven’t quite read and absorbed all the details in the article. But no matter! This is the internet – where we’re going, we don’t need to read “articles!”
Speaking of skimming articles, I thought this “song of the summer” piece from CNN was a bit of a dud, since it not only suggests there is none for 2025, but proposes couldn’t we maybe just go back and use one from last summer? Meanwhile trashing what is obviously far and away the *true* “Song of the Summer 2025”:
Or maybe, Errico suggested, the song of the summer is one that best exemplifies the quickening encroachment of AI into creative fields, like AI band Velvet Sundown’s AI song, “Dust on the Wind.” Even the title sounds like a Fleetwood Mac-inspired ChatGPT prompt.
Lastly, in my round-up of mildly interesting miscellaneous Velvet Sundown misinformation bits and bobs is this piece from Charleston, South Carolina’s Post and Courier paper about a local musician (confusingly named after a different geographic locale, Johnny Delaware) who was misattributed as being an AI music act. What’s the word for this, when someone is accused of being AI, but they are not and/or they deny it? “AI-jacketing” maybe?
An article about the phenomenon appearing in The Atlantic that’s raising questions about transparency and legality regarding AI on streaming platforms referred to Johnny Delaware as another potential fake band. The story labeled Delaware a “related artist” to The Velvet Sundown on Spotify.
In fact, he’s a very real artist with quite the following right here in the Holy City — he has a show coming up at the Music Farm on Aug. 8.
In response to the AI claim, Delaware released a video on social media calling out the journalist behind the July 4 story for not properly fact-checking.
Awkward. An archived version of the Atlantic article mentions Delaware in passing, within the context of trying to find other songs that sound similar to The Velvet Sundown:
I pondered for a moment whether any of the other artists on my custom playlist (the South Carolina folk-rock singer-songwriter Johnny Delaware? The Belgian folk-pop quartet Lemon Straw?) might be fake—and how one might try to suss that out.
Delaware in the video suggests the author of the piece didn’t simply Google the musician to see their many videos and other media artifacts. Incidentally, a widely re-posted AP article about spotting AI-generated songs in the wild suggests that listeners “do a background check” before ultimately conceding that there are “no easy answers.”
In the Post and Courier Piece, Delaware is subsequently quoted:
“I mean, I have no control, so all you can do is observe and let it ruin your life or just laugh at it,” he told The Post and Courier.
Delaware said he wrote both The Atlantic reporter and the podcast hosts, who offered to interview him. He declined, at least for now.
“It’s a really weird situation,” he said. “I don’t want to be involved with it, but it’s kind of following me around. … Like maybe misinformation about me being a robot isn’t worth me getting into.”
I’ve… obviously taken a really different tack in this saga, and clearly I disagree with the approach Delaware is outlining above. One, I think he does have a lot of control here to shape it in the media and on social platforms. Two, whether or not you want to be involved with them, when things like this come up, they are simply opportunities. Me, I tend to lean into this kind of thing. I guess I tend to, at the end of the day, maybe even encourage misinformation about me being a robot. But that’s just me!
Thought this old archived NYT article by David Byrne about why he hates the label “world music” had some interesting and quotable moments, like this one:
The issue of ”authenticity” is such a weird can of worms. Westerners get obsessed with it. They agonize over which is the ”true” music, the real deal. I question the authenticity of some of the new-age ethnofusion music that’s out there, but I also know that to rule out everything I personally abhor would be to rule out the possibility of a future miracle. Everybody knows the world has two types of music — my kind and everyone else’s. And even my kind ain’t always so great.
What is considered authentic today was probably some kind of bastard fusion a few years ago. An all-Japanese salsa orchestra’s record (Orquestra de la Luz) was No. 1 on the salsa charts in the United States not long ago. Did the New York salseros care? No, most loved the songs and were frankly amazed. African guitar bands were doing their level best to copy Cuban rumbas, and in their twisted failure thay came up with something new. So let’s not make any rules about who can make a specific style of music.
I have been going pretty hog wild on Suno these past few weeks. And one thing that has been firmly solidified in my head is that, because these tools allow basically anybody to output essentially the same type of music (especially if they can copy your prompt), then basically everything depends in the end on the narrative wrapper in which you deliver your finished products to the consumer.
The time is not quite ripe to reveal it here, but over the past few weeks, I’ve put a wrap on 30 full-length albums I made using Suno. During that time, I’ve watched with interest the evolving story around the new AI artist who signed an allegedly $3 million deal with Hallwood media, Xania Monet. One of the wrinkles that has received I think too much attention in this saga has been Xania’s being called out by a human artist named Kehlani, who criticized Xania for being basically not human enough.
“There is an AI R&B artist who just signed a multimillion-dollar deal … and the person is doing none of the work,” a frustrated-sounding Kehlani told followers without directly naming Monet or Jones. “This is so beyond out of our control.” […]
Regardless, Kehlani says, “Nothing and no one on Earth will ever be able to justify AI to me.”
They added, “I don’t respect it.”
I’ve thought a lot about this, and personally, I don’t really like Xania’s songs, one of which has racked up at least 2.3M views on YouTube, and Billboard elsewhere estimates a total of 17M listens across platforms.
Monet’s most popular track, “How Was I Supposed to Know?,” which has ranked in the top 10 on Billboard’s R&B Digital Song Sales for two weeks in a row and hit No. 22 last week on the Digital Song Sales chart overall, has accumulated 22,700 song equivalents in the U.S. and more than 3 million on-demand audio and video streams.
Here’s the track:
Whether or not I actually like it is, of course, entirely irrelevant. Because those streaming numbers don’t lie (I’ve seen no suggestion of inauthentic stream/fraudulent activity anywhere with regards to this). For me the music sounds a little bit on the boring and derivative side. But what I think doesn’t matter, because it seems that potentially millions of people enjoy it.
“… to rule out everything I personally abhor would be to rule out the possibility of a future miracle.”
“Abhor” is a very strong word here, and it’s a long way from how I feel when I listen to this track, which sparks a lot less negative emotion for me. I just don’t particularly like it, rather than hate it or what it stands for personally. But I do think the public reaction, which has largely been unfortunately shaped by Kehlani’s reaction is a little bit on the ridiculous side.
Why? Well, because for me, Kehlani’s music is pretty much also on the slightly boring and derivative side. Even if she’s a “human.” Exhibit B:
And here’s my thing: as far as I can tell, Kehlani has no greater right to say she is a “human” than does Xania’s actual creator, one Telisha “Nikki” Jones. Kehlani does not, in my eyes, have some monopoly over what it means to be authentic or human than anyone else does. Kehlani’s statement, referenced above, in part reads:
“Nothing and no one on Earth will ever be able to justify AI to me.”
My question is basically: so what? Why does anyone else need to justify the art that they make to some basically random person who is unhappy about it? Because in my experience of being a person on the internet, there is basically always some random person(s) who will be unhappy about literally anything you do, and will do their level best to cut you down for it.
Regarding Xania’s record deal, Kehlani further stated, “the person is doing none of the work.” But again, Kehlani has no monopoly on what it means to put effort into something, just because her work follows a particular more conventional mode of creation, where Xania/Jones’ follows a new, different, emerging one. Kehlani also has, as far as I can tell, no magical crystal ball that gives her exclusive insight into the very real struggles we all equally face as humans on this planet, trying to survive & thrive against all odds, and against a system which all equally tries to pull us down at every step of the way.
To suggest someone using AI is doing “none of the work” is to fundamentally misunderstand that as artists, the “work” we all do is the work of merely living. And we all do it equally at the end of the day, regardless of what tools or technologies we use to express that business of living creatively. When we accuse someone else of not engaging in the true authentic work of living, of being a creative person in a society which at times seems almost entirely purpose-built for crushing creative people – simply because we don’t like what they created – we essentially pile ever more work on that person, forcing them to deal with more and more of our own accumulated baggage in addition to whatever portion nature or society has already allocated them to bear. None of us can truly see into the soul of anyone else and therefore has the right to sit in some absolute holy judgement over the pain of the effort anyone else has gone through in their lives to get where they are.
I think it’s perfectly fine for Kehlani, or anyone else, to simply not like Xania’s music. As I said, I don’t particularly enjoy listening to it myself. But what I do or don’t like is all but irrelevant in the face of millions of people who do like it, who do find threads that resonate with their own personal experiences of what it means to be doing the work of being human. It feels selfish and narrow to me to try to undercut that very obviously real sentiment – and for what? Because Kehlani and others obviously feel threatened by someone else who has figured out a different solution to the problems put onto artists by capitalism? To me, that’s cheap.
I’ve wondered in this game too, at what point a “grifter” becomes a “hustler” which is more socially & culturally acceptable. Why are we supposed to “respect the hustle” but scorn the grift? It’s the same damn thing. The reality is we’re all stuck in the same sad, bullshit pathetic grind. If people are able to find some way out of that maze – any way at all – and share some light in the tunnel while doing it, well, I personally *do* respect that. Even if I don’t think it’s necessary that anyone else has to justify any of it to me. At the end of the day, every person who follows the artist’s path is only responsible to their own inner light, their own creative voice and urging that keeps them up in the middle of the night, and keeps them going. The rest to me, increasingly, is just so much noise, and to quote Kehlani’s words back on her, I don’t respect it either.
I’ve started talking to the media about my new AI music project, and will reveal more details as those pieces start to get published.
What I can say now is I’ve spent the last 6 or so weeks heavily using Suno AI. And while I’ve found ways to make the models (especially v4.5+) give results that I think can be pretty good sometimes, my experience overall is that the product and user experience side of the service are pretty terrible. Which is why I’m walking away from it and deleting my account. But also because that project is finished, and I like to be able to make clean breaks after things are finished, in order to open up space for new things to come along.
Apart from a pretty rough ride using the actual service, I’ve come to realize something somewhat unfortunate after observing and interacting with other users in the Suno community on Reddit. And this might come off as harsh, but I think might need to be said out loud. I’ve become as a result of those interactions fairly convinced that much of the problems we ascribe to AI are less about AI and more simply problems of the people using the AI.
To put it even more bluntly: I think the reason that so much of AI music seems to suck, is that, well, unfortunately, the attitude and aesthetic choices that the people are bringing to the creation of AI music, well… sucks. This applies equally I’m sure across other modalities and service offerings related to AI. It’s not strictly a Suno issue, but for whatever reason, that is where the issue became crystal clear for me: sucky people make sucky things, whether using AI or not. I don’t really like making such blanket statements about groups of people, and there are always endless exceptions to generalities like that. But the amount of weird dumb infighting and pointless belittling and smugness that I witnessed in that “community” when people bring up legitimate concerns around the product offering makes me not too keen on being part of that community.
Maybe this is a variant of “looking for love in all the wrong places” and I’m asking too much from an online group of people dedicated to making mostly to vulgar, joke, or derivative songs. But I think we as artists can do better, and that our discourse needs to be better. That we need to challenge the limits and each other to get to where we’re going, and rise to the next level of wherever these AI tools are leading us. Probably Reddit is the wrong place to do that, obviously. Probably much of the internet is, in fact, the wrong place for that. But what and where is the right place? Does it exist? Maybe it can and should only exist in small close knit friend networks, Signal groups and the like. Those remain near and dear to my heart.
But the sentiment remains for me regardless: people with annoying attitudes and bad aesthetics make things that are annoying and aesthetically bad. And the problem with AI is not merely the technology and its many issues (and I believe there are many real & important ones). But the problem, like everything, is people. And just like with anything, you get out what you put in. Garbage in, garbage out, as the saying goes… So let’s maybe stop putting garbage in, and see what comes out the other side? Just an idea!
I’ve been living under a rock (or several) for I guess some time, because everything in this Wired profile about AI artist (clankercore?) Neural Viz is new to me, except this:
Before long, the filmmaker had built an entire world with its own language, characters, and lore, all of it made with AI.
And also this sentiment:
Many commenters on YouTube told Kerrigan that his videos should be on Adult Swim. But when he met with producers affiliated with Adult Swim, he said, one of them suggested that he might not need them; that the power had shifted to creators. “That sentiment has come up multiple times in meetings with other various studios,” Kerrigan said.
This seems all too familiar, and has got me wondering the existential question of, why does an artist today even need a label? What exactly does that get them? Reach? Engagement? You can get those things on social media, in the press directly, by actually engaging with other humans. Without any kinds of obligations to a third party. All you need is a distributor, a platform, and some means of generating the materiel. And of course getting paid for it (but which ideally happens at the same place you upload).
I was surprised that instead of focusing on a not that good “Studio” experience, that they didn’t just make the much more obvious and in my mind probably more lucrative business step of offering distribution services, direct transloading your content out to Spotify/Apple Music/YouTube, etc. Maybe that’s in the works, but to me that’s much more valuable and a time saver than another AI editing experience that is tedious and somewhat user-friendly with unpredictable mid results. Just let me make my stuff, send it out, get paid, and get the hell out of my way. This is the way.
So what really could labels offer artists that’s the most valuable? Budgets for marketing, access to good editors and tools, budget and eng/ux team to build out custom tools for production. There are probably more, but that’s off the top of my head. Will give this more careful consideration as I digest this Wired piece more!
Interesting piece on Financial Times intended to prepare the music industry for the coming wave of AI slop:
That’s not so bad for streaming platforms — they are incentivised to maximise playing time, and are less concerned about who’s being listened to. But it may well marginalise existing music companies and human artists, especially where content is not particularly original, or where the fan community is less engaged.
The music industry will need to be proactive. Labels may try to create their own AI artists.
Hate to be the bearer of bad news, but that’s already happening…
The other thing I want to highlight here, based on the line about “human artists” being marginalized, is that people using AI as part of their creative process *are also* “human artists.” Just because we choose one tool over another does not make us not still humans expressing creativity through the technologies made available to us in the market. It should not be an us-vs-them mentality. It should instead be a let’s look at where this is going, what we can do with it, and how should it best function in our ideal future kind of conversation…
