The clip from the OP was taken down, at least where I am. I didn't even get to hear it!
If anyone still wants to hear this, here is a link to it in my Dropbox. I'll leave this up for a day or so:
https://www.dropbox.com/scl/fo/tkmxo...o2ahufkgm&dl=0
On a more positive note, AI will give us the opportunity to enjoy some new perspectives on Yes. For instance...
At last an album with the Bruford/Moraz combination.
The much anticipated symphonic reggae version of CTTE.
The scary hybrid Open Your Eyes / Heaven and Earth with new artwork in the vein of Jon Anderson's True.
Tormato meets The Quest with special guest Ted Nugent.
The whole catalogue sung by Steve Howe.
The whole catalogue with the drum sounds replaced with Jon Anderson singing Cha Cha Cha.
Having listened to it twice, I will say that it's Yes-ish especially if you like 4/4 time.
Mongrel dog soils actor's feet
I've listened to this a couple of times and I don't believe it's all AI-generated, with the exception of Anderson's voice. I think it's most likely an artist/producer doing a YES pastiche in a post-TALK-era style with real instruments (except for those drums). I think the vocal melodies and lyrics were composed by a human that was then augmented by AI that was fed a ton of Anderson's vocal references to create an admittedly pretty remarkable facsimile of Anderson's phrasing, pronunciation, and tone. But I don't think it was actually composed by AI. Just my .02 cents worth.
This JA is not as convincing but in spots, very close
https://youtu.be/RRcdcNoOsXc?si=43DlnuKoItUgK5gD
Sent from my NE2217 using Tapatalk
Death inspires me like a dog inspires a rabbit
"No AI .. no Yes"
"Normal is just the average of extremes" - Gary Lessor
I'm sure you know far more than I do about the technological limits of AI. So I'll defer to you and knowledgeable others here about feasibility. But... Consider how downright bad this AI Yes thing is as a musical composition. If a human actually wrote this they would have put in some repeated hooks and riffs, a catchy melody, a crescendo, some resolution. This thing goes on and on for absolutely no purpose. That to me suggests pure AI. Just spitting out inoffensive and harmonious drivel for as long as a user requests it.
Listen to it for 20-30 seconds you might say "wow this sounds amazing." I forced myself to listen for about 10 minutes and my brain says "make it stop! now..."
This is 100% AI. The music was not created by a person. The sound is immediately recognisable if familiar with the current output of AI music generators.
I suggest that if anyone doesn't believe this is AI, go to Udio and try generating something. You'll see then.
An astute observation. I struggled to listen to this and actually made it roughly halfway through even though all along, I was keenly aware of an odd sort of discomfort. As you pointed out, it was undoubtedly my involuntary reaction to a mentally destabilizing 20-minute jaunt through the aural uncanny valley.
That said, I wouldn't be at all surprised to learn that various short snippets of 5-10 seconds were in fact composed by a human and then the whole lot thrown into the generative AI blender to create these 'songs.' But perhaps we'll never know, and it frankly doesn't matter... at all.
I wanted to hold out on this, but I was tempted to hear a little and join the conversation, so thank you for posting the link.
The Jon voice is somewhat convincing but I absolutely hate the uniformity of the backing voices, chiming out repetitive echoing lines, it’s not especially Yes-like. The music is ultra-bland too, one paced, lacking any memorable melody. To nail a Yes pastiche they really need better bass lines too, nothing says counterfeit so much as the absence of any trademark Squire playing.
Having said that, I cannot really believe that technology has come to the point where this was all generated from AI, yet it does lack a certain element of human spark and warmth, so in that respect it is believable. Overall my initial scepticism and wariness was well founded. I’ll never need to listen to this again.
I will in no way encourage the use of AI especially in thievery of artist sound and style. I hate my encounters with the old-fashioned robot interfaces and despise the new generation of shit being shoveled at the American consumer.
You're definitely wrong about this. It's totally AI. Lots of tell-tale signs. You could easily churn something like this out yourself if you had the patience to play the AI roulette until you had pieces you were happy with. You'd end up with something just as crappy and pointless as this.
Yeah I don't know how many times I need to say it but if anyone who doubts just goes to Udio and generates something they'll see how this is also totally AI. It's very clear. All this 'there must be a person involved at least writing some of it' might just be copium.
Same reason folks make crappy fan trailers for movies and then put "OFFICIAL TRAILER" all over their video descriptions...for the attention?
Case in point: how many additional listens (and even some degree of likes) did this video get as a result of posting over here?
If you're actually reading this then chances are you already have my last album but if NOT and you're curious:
https://battema.bandcamp.com/
Also, Ephemeral Sun: it's a thing and we like making things that might be your thing: https://ephemeralsun.bandcamp.com
Bookmarks