On The Beauty and Dilemma of the New A.I. Randy Travis Song
The trauma of losing the voice of Randy Travis in 2013 after his near-fatal stroke is something the country music community is still attempting to recover from, tempered somewhat by the grace that we still have Randy with us to fete and celebrate in the country music community that he’s been so integral to.
It’s the voice and music of Randy Travis that were the catalyst for the neotraditional resurgence in the ’80s that brought country music music back to its heart. Over the years, we’ve been graced with opportunities to hear unreleased tracks from Randy stored up over the years, while the promise of more unheard Randy Travis songs recorded in his heyday are still out there in the offing.
But in lieu of Randy being able to sing new, original songs due to his continued aphasia, he has partnered with his old label Warner Music Nashville to release a new song called “Where That Came From.” Written by well-known country songwriters Scotty Emerick and John Scott Sherrill, and produced by long-time Travis producer Kyle Lehning with Jerry Douglas, “Where I Came From” sounds like a Randy Travis song written and recorded in his heyday.
The release of the song came with ample teases of Randy appearing in a studio, flashes of microphones in the foreground and background, with Randy sitting there with headphones on, seeming to enjoy the whole experience.
It is important to note that even though Randy is non-verbal, those close to Travis assure that he is still 100% cognitive, aware, understanding, and even communicative both through simple visual affirmations, as well as writing or typing out things he wishes to communicate.
Also, Randy Travis has shown some elementary ability to sing through the years since his stroke, including renditions of “Amazing Grace” and “Forever and Ever, Amen.” However impaired these performances may be, for some it still feels miraculous how far Randy has come. In the case of certain brain injuries and stuttering (Mel Tillis), there are instances where people are not able to easily talk, but are much better at singing since these two functions are handled by different parts of the brain.
But that is not what is going on with “Where That Came From.” Though there has been absolutely no disclosure whatsoever so far, clearly Warner Music Nashville and the Randy Travis camp are using some sort of version of AI to create the performance from Randy Travis. This presents a moral conundrum, and one that up to this point is unprecedented in the music space.
From one perspective, the idea that while living but unable to sing full formed songs, Randy Travis can dictate and sign off on the production of new Randy Travis music feels miraculous. In an idyllic sense, this is the kind of thing that technology and AI can rise to and enhance our lives. How is this any different than a prosthetic or robotic limb being added to a paraplegic? This is technology replacing what has been taken from us.
But in another sense, this is technology creating the facade of an authentic rendering from Randy Travis, when most of us know he didn’t actually sing this performance, while some might be outright duped into thinking he did, especially due to the way the song is being marketed. Despite the remarkable rendering of Randy’s voice in a new, unique song, it goes without saying that it means something different if you know he didn’t actually sing it.
The nuance, the pauses and breaths, the inflections, the imperfections, and the indefinable soul—these are all the things that go into making a song, and a country song especially. It’s inherently human. And though algorithmic enhancements can attempt to emulate such things to fool the senses, you still inherently know what you’re listening to is not real.
Nonetheless, we can still enjoy “Where That Came From” as country fans and as Randy Travis fans as the closest thing to Randy Travis that’s possible. But the biggest dilemma of this release is the lack of disclosure from Warner Music Nashville about what’s going on. As Saving Country Music said when addressing the onset of AI in the music space in early 2023:
“There should be labeling on music so you know where it came from, just like food. It should be marked 100% certified organic, or have a list of the artificial ingredients that went into making it so that way consumers can make informed choices. The insistence by the industry for proper musical labeling should start right now, today. Because AI is here, and powerful, and will corrupt the music landscape in quick order if it is not addressed.”
We all know that Randy Travis did not sing “Where That Came From” in real time, but that it was taken from previous performances cobbled together by technology. Nonetheless, it sets a bad precedent to not disclose this to the public. It’s not the song, but the sort of surreptitious wink-and-nod aspect of it that in some ways has drained the sentimentality out of the experience.
“Where That Came From” actually makes for a great discussion point about AI and music, and so we need to have that discussion as opposed to allowing it to be part of the technological AI creep that eventually could, and potentially will, replace all carbon-based musical performances with technology so the industry can streamline production and cut costs.
Meanwhile, it seems that more of this music from Randy Travis is coming. There was some confusion that “Where That Came From” would be an EP or an album, not just a song due to the wording of some of the marketing. Video teases show Cole Swindell and Cody Johnson in the studio, perhaps alluding to upcoming collaborations. They also seem to be teasing that eventually they will disclose how all of this came about, and they likely will. But doing this in arrears sets a bad precedent.
As Saving Country Music was finishing this article, Billboard disclosed that it will be officially revealed “Where That Came From” was constructed using AI via a segment on CBS Sunday Morning airing on May 5th. Cris Lacy of Warner Music Nashville is quoted as saying in the report to air Sunday, “It’s Randy Travis. Randy’s on the other side of the microphone … It’s still his vocal. There’s no reason he shouldn’t be able to make music … And to deprive him of that, if he still wants to do that, that’s unconscionable to me.”
AI promises to forward potentially promising, but potentially catastrophic developments for the music industry and everything else in life. Asking audiences to suspend disbelief as Warner Music Nashville has done here will only aid an abet the latter outcome if we are not careful.
But placing all of those discussions aside, it is heartening to hear a new Randy Travis song, graced by the closest thing we can get to the original Randy Travis voice that we all fell in love with, and saved country music in the 1980s.
May 3, 2024 @ 11:23 am
It had me going.
(At least on one listen,) it sounds like a legitimate Randy Travis recording from the ’90s. Given that Travis was always on one label–Warner Bros.–and had a relatively small number of albums–it’s sort of shocking to hear something new from then.
If this is totally artificial, it’s a scary thought.
What next–“New” recordings from Elvis, Sinatra, Jimmie Rodgers, Louis Armstrong, Judy Garland, Hank Williams, the Beatles and Caruso–sounding better than they ever did before?
May 3, 2024 @ 11:42 am
The protection of intellectual property by the estates of the deceased is a very real issue with all of this. The family of Tupac and others have already come out against their voices being used for “new” songs. How to attribute and monetize this is something we need to decide, and unfortunately, we are not having those decisions.
We disclose if tracks have explicit lyrics, are remasters or new versions of recordings. We need, at the absolutely least, to be marking tracks generated with AI so at least we have that very basic element of transparency and disclosure.
May 3, 2024 @ 1:03 pm
I think it could be fine if the artist name was changed from “Randy Travis” to “AI Randy Travis” or Randy Travis via AI” or something like that.
May 3, 2024 @ 1:24 pm
Any explicit track on streaming services is marked with an “E,” and all creators are required to disclose explicit lyrics when uploading tracks. If they don’t, the track could get pulled or recategorized. Similarly, songs that include AI-generated lyrics or other content should be marked with an “A” or “AI.” This seems like the absolute possible least we can do for transparency and to protect human creators.
May 3, 2024 @ 2:43 pm
Reasonable idea, but not feasible. A composer or lyricist can sit alone with a computer in his own room–or at a table at Starbucks–and enter music or lyrics into an AI program and the program can work off of his ideas and come up with its own notes or lyrics to finish the song. How are you going to make the composer or lyricist reveal that he had nonhuman help?
(Unless we want to go even further into Big Brother territory and have the AI program save the session and itself go on the Internet and “rat out” the composer/lyricist if the song gets published and becomes a hit.)
May 6, 2024 @ 2:05 pm
It’s his voice only dubbed in. That’s what it sounds like to me. I do believe they should only do this for music artist that have serious issues like Randy Travis.
May 6, 2024 @ 2:24 am
If new music is done with the participation of the artist themselves, it should be OK for the use of AI in a situation like this. It is when totally unrelated people take a deceased person’s voice and come out with something in their name where the problem lies.
June 24, 2024 @ 8:36 am
Randy’s folks have indeed said how this was produced!! There are videos and interviews discussing it. Get your facts straight geez. Give the guy a break
June 24, 2024 @ 4:20 pm
Read my post, I said “If it is done with the participation of the artist themselves”, which in this case IS Randy because he is still alive, it’s OK. He was there, he helped with the production and OK’d it. I LIKE the new song, I BOUGHT it! I did NOT badmouth Randy at all, so get YOUR facts straight! Sheesh
May 8, 2024 @ 8:14 am
Where this ends, no one knows. RT is involved in this so there are no IP related issues.
However, the AI ethics are non,-existent at this point. What would stop a record company from compiling an AI library of an artist’s voice snippets and then compiling and releasing music from that – since music contracts make artists relinquish rights?
Also what about an artist who decides to not step in front of a mic and do this too? As artists demand redoect ftom fans, artists should in turn respect their fans with disclosure. Such a disgrace that an artist could possibly take that way out. Would that be considered a legitimate creation or a lazy way out?
The arts are one of the ultimate forms of expression. AI has the potential to remove the human element out of it.
(AI is powerful tool easily missused; (deep fakes, students cheating on term papers, useless customer service bots). And it has limitations by being based on incomplete or prejudicial bias built in to who creates it. )
May 4, 2024 @ 6:39 am
Tal vez sean las obras hechas sin intervención de IA las que lleven una advertencia o marca distintiva, lo cual, a quienes nos importe el asunto nos resultará un plus de valor artístico.
Perhaps it is the works made without AI intervention that carry a warning or distinctive mark, which, for those of us who care about the matter, will find it an added artistic value.
May 4, 2024 @ 9:02 am
The song is very good and if you did not know it was AI Randy Travis most people would probably never guess. The merits of AI will be debated for sure but at least in this case a living artist endorsed the project.
To correct an earlier post Randy Travis did not exclusively record for Warner Brothers. In 1998 he moved to DreamWorks Records. Two years later he signed with the Word label where he recorded primarily gospel/Christian music until he returned to Warner Brothers in 2008.
May 5, 2024 @ 7:51 am
Randy is very much alive and involved in the recording of this song through AI. Nobody was concerned when Lisa Marie did a song “with her father”. I think this is amazing. I hope he does more songs like this.
May 8, 2024 @ 4:05 am
It’s so nice of all those big music corporations to develop this technology only to completely shelve it after “giving Randy Travis his voice back.”
(sarcasm)
If you want to truly stand up and defend your beliefs and be a gate-keeper it means sometimes being an insensitve asshole to others. I hate seeing Randy Travis in videos. It always looks like grandkids are using their demented grandparent for clout by posting them in videos. Why can’t they let the man live out the rest of his life in peace? This feels exactly like those pictures where they make the grandparent hold up a sign that says “Can I get 1000 likes for my birthday.” It’s sick. Shame on them and shame on the morons eating this crap up.
May 3, 2024 @ 11:24 am
It a wierd thing… I love the song…. Just keep thinking that is really not him and then I started thinking could this be used to create new music from Cash, Merle, Waylon, Hank, Elvis, and other greats…kind of scary…. But man I love hearing anything from Randy
May 3, 2024 @ 11:32 am
Well guess I’m a idiot didn’t think nothing of it and just thought it was a unreleased song
May 3, 2024 @ 11:34 am
“How is this any different than a prosthetic or robotic limb being added to a paraplegic? This is technology replacing what has been taken from us.”
Because it’s not a prosthesis or artificial limb. “AI-generated vocal” is almost the same thing as “synthesized vocal.” There is no original vocal, but a kind of digital model of a person’s vocal EQ curve, to put it crudely. In this case, it’s a digital model of Travis’s vocalizing applied to some input — maybe a male who already sounds like Travis when he sings, such that the digital model has a lighter lift than it would otherwise.
What has been taken from us is not something that can be replaced with a digital dummy, however life-like the dummy is.
“to deprive him of that, if he still wants to do that [make music], that’s unconscionable to me”
Who’s depriving? Travis is free to make digital dummies of himself and we’re free to listen and go “interesting, ok. Next.” But let’s not kid ourselves.
And we all know where this leads, directly to IP protection for vocal models. I’m guessing there’s big money to be made on George Jones Vocal(TM), Dolly Parton Model(TM), and so on.
May 3, 2024 @ 11:52 am
Music producers have been manipulating vocal tracks since the early days of studio recording (doubling the lead vocal, changing speed, etc.). Since then, most music fans have accepted all creative “adjustments” to recordings over the decades. Recently, autotune has been accepted in the mainstream and AI driven tracks will become the standard someday too. It’s just a matter time.
I imagine someday someone will start a recording label creating AI generated performers like Hatsy Heinz, Billie Felson, and Berle Taggert and the younger, music buying public won’t bat an eye. All this technology is taking away authentic musical expression which is one way (and for me the best) for humans to communicate with each other. Sad really.
May 3, 2024 @ 2:00 pm
If the younger music buying public listened to anything that sounded like Patsy Cline, Willie Nelson, or Merle Haggard, I guess I’d have to consider that a major W on some level. But I completely agree.
May 3, 2024 @ 11:53 am
Overall I don’t have a problem with it as long as he’s still alive. They should be more upfront about it though. I didn’t have a big problem with how the new beatles song Now and then was done since at least the remaining members of the Beatles sung on it. Now in the case of dead artist where nobody of substance who was with them is still alive, I have an issue. Sometimes you just got to let it be though I doubt that will be the case going forward.
May 3, 2024 @ 12:02 pm
I am curious how they actually did this. I saw one report where they thought someone else sang the song and they were able to use AI to put Randy’s voice over it, so it sounded like him. This is definitely an odd situation. It sounds great though and makes you want Randy Travis music.
May 3, 2024 @ 12:40 pm
This is all supposedly going to be revealed in detail on the CBS Sunday Morning segment.
May 3, 2024 @ 2:34 pm
CBS Sunday Morning….laugh my …off
They aren’t relevant for anyone and haven’t been in years. It’s these moldy old geezers on camera who live on the coasts, so removed from real America its laughable. Yeah, I’ve been noticing that they have been obsessed with Charley Crockett for a couple years and recently they had Charles Wesley G on. And Sierra Ferrell. None of their anchors even know who these artists are. But they sure do pretend like they are the coolest and most in touch hipsters. It’s a desperate attempt to keep ratings alive.
Wow that was rude of me..sorry about that pointless rant. ????
May 3, 2024 @ 3:29 pm
When you talk to folks in the industry, they concede that appearances on the late night talk shows (The Tonight Show, The Late Show, Jimmy Kimmel) just don’t move the needle anymore like they used to. They’re good for bragging right and impressing your mom, but much else. However, the CBS Sunday Morning segments do tend to result in a bump for artists. It depends on the artist how much of that bump is, but that’s the reason so many artists are doing them.
Yes, only “moldy old geezers” watch it, but moldy old geezers is kind of the Randy Travis demographic.
I will say that Anthony Mason who often hosts does seem a bit out-of-touch. Remember, it was Mason who mispronounced Tyler Childers’ name at the 2018 Americana Music Awards (as “CHILD” no “CHILL”), and Childers let him hear it on the stage. I respect the dude for highlighting this music. But I agree that he often comes across as out of his depth.
May 3, 2024 @ 5:54 pm
Tyler Childers ain’t Bruce Springsteen. Most people have never heard his name. I bought two of his CD’s (after learning about him from this site) and I figured his name was pronounced like its spelled.
Hectoring of people over pronunciations has become a form of intimidation in the new culture. Most people read VP Kamala Harris’s first name with the accent on the second syllable, but if you say it that way (rather than with the accent on the first syllable) you may be calld insensitive or even the “r” word. [Nobody ever got called out or accused of having ill intent if they pronounced the 40th President’s name as “Reegan,” (rather than the correct “Raygan.”)]
May 3, 2024 @ 6:56 pm
I don’t know that it’s “hectoring” to expect someone giving out an award to someone to know how to pronounce honoree’s name. It was an honest mistake. I know I’ve made them, and people butcher my last name all the time. I expect it. I was just making a point.
May 3, 2024 @ 12:08 pm
I was hoping for a new album with songs like One In A Row and Fool’s Love Affair. I thought I saw an EP on Apple this morning called Randy Traywick, but it’s not there now.
May 4, 2024 @ 7:13 am
I still have it on my Apple Music. They pushed the the release date back to 10/1/13 I suppose because it was stepping over this new single.
It’s early recordings, sounds like even prior to the Nashville Palace album. That’s a project I’d like to see released in digital for than AI Randy
May 3, 2024 @ 12:10 pm
Mark my words, the next step for Nashville will be to use this technology to make “new music” from dead artists. Come hear a new Elvis song! Look, new Hank Williams! To me, this is a horrendous step.
May 3, 2024 @ 12:35 pm
Already happening… here’s AI Hank Sr covering Alan Jackson:
https://www.youtube.com/watch?v=bmI-1IzDlrI
May 3, 2024 @ 3:02 pm
Doesn’t have to happen anymore. Interesting, tho.
May 3, 2024 @ 8:22 pm
Not only that, but what’s to stop these studios to age the sound, and release it as “unreleased material” instead of AI? And how will we know for sure what’s legit? Has it already happened? Would it classify as fraud? Regardless of one’s opinion on the usage of AI, it 100% needs to be transparent imo
May 3, 2024 @ 12:18 pm
Is this confirmed to be AI? Incredible, weird, and even a bit troubling (can envision a thousand ways in which that could be abused).
That said, I liked the song! Definitely makes me want more Randy Travis!
May 3, 2024 @ 12:41 pm
Yes, it has been confirmed AI in a press release CBS News sent out this afternoon.
May 3, 2024 @ 12:25 pm
This is a complicated issue. At least for me.
May 3, 2024 @ 12:39 pm
If this is my option for new Randy vocals, I’ll pass. It’s not him/no soul. It’s like a bad movie.
May 3, 2024 @ 12:44 pm
If this isn’t Randy involved at all, beyond a nominal production standpoint, then I feel like this isn’t a legitimate addition to his legacy. I could overlook some things if this was a song that he wrote, however. A few years ago, the shred metal guitar maestro Jason Becker, who has been incapacitated by ALS for decades, composed an album’s worth of material, and while it was played by an army of musicians, including a who’s who of heavy-metal neoclassical guitar, Becker still composed every note, produced and oversaw the recording. When news of this hit, initially, and with the short soundbite that was available, I’d figured that Randy had sang the lyrics to the best of his current ability and his vocals were heavily processed. On the recording, you can hear a couple of places where there’s some very obvious manipulation going on, but if this is purely AI, again, I cannot see this as nothing more than marketing, and not a legitimate part of his canon. No more than a hologram of Tupac is a “live performance.”
May 3, 2024 @ 2:05 pm
I agree that if Randy had written or co-written the track, it would feel a little less smarmy. Though I’m not sure if he did it would make a major difference. It just would make it all feel a little more authentic.
May 3, 2024 @ 12:55 pm
Guitarists have been using “impulse response” technology for years. Digital emulations of old amplifiers, sonically indistinguishable. So I imagine there are a lot of shrugs in Nashville over this. Making money is the name of the game.
May 3, 2024 @ 1:08 pm
I’m seeing a lot of folks citing Autotune, reverb and echo, and other enhancements on vocals signals as being similar to this. But this is not about enhancement. It’s about creation. These vocals are being created from something else, not enhanced from an original performance. This is also one of the reasons people think differently about Autotune used to hide pitch imperfections, and echo in a vocal signal, or Autotune as a vocal effect. There is an attempt to deceive that people don’t like.
May 3, 2024 @ 2:10 pm
Well, I think the software here works a bit like IR technology: you pipe in a signal, and the computer applies a kind of (sophisticated) filter to it to change its sound.
In this case, the computing job is much more encompassing than just altering pitch or applying an effect like reverb or echo.
Here the software is (I’m guessing) taking someone’s sung performance and piping it through a software application. The sung performance is there as a skeletal thing, for time and structural dimensions. A set of words is then mapped to this sound-skeleton. The result is probably a very digital-sounding thing.
But then the magic happens when the software applies, in this case, the Randy Travis(TM) filter, which must be like an Impulse Response thing. This would be something like “a digital model of Randy’s vocal timbre, vowel sounds, and vocal timing characteristics.”
This has been a very hard computing problem for decades, but evidently we’ve reached a point where the results are really impressive. Visual CGI has been on a similar path and has been making really rapid and obvious progress. The vocal thing is not far behind.
Producers have talked about certain singers having a good vocal “print” since, well, recording began. Now we’re at a point when vocal prints can be turned into vocal algorithms (if you like) and applied to anything.
The questions loom: what are musicians for? Hell, what are *people* for?
I know what I’m for, and one of those things is people.
May 3, 2024 @ 12:55 pm
I truly don’t know how I feel about this. Randy is still alive and obviously participated in this project. I guess that’s fine – although knowing it isn’t really him really hurts my ability to enjoy the song.
My issue is that dead artists will know have their estates flocking to make new music through AI to make a quick buck. Songs will be chosen that the artists wouldn’t touch with a ten foot poll.
Kinda scary.
May 3, 2024 @ 12:59 pm
It’s a beautiful song, and it’s recognisably Randy Travis. It’s not Josh Turner or Joe Nichols or Scotty McCreary or anyone else who sounds quite a bit like him.
I think it’s life affirming that technology can do this.
May 3, 2024 @ 1:16 pm
Ultimately, the masses won’t give a damn and will lap it up. Same thing is happening in the visual arts space. Instagram is already full of fake AI art. And if you dare to take it to task, you will be driven into submission in the comments. The fakers claim it is just a new tool in the toolbox. Few care. A dining room table analogy gives me some comfort: most of us are perfectly content with cheap mass produced dining room tables made in Asia somewhere with little originality or sense of style. Others with discerning taste (very small percentage) prefer to have a craftsman hand produce a one-of-a-kind table in his shop. That’s the future of music and art. Sad indeed.
May 3, 2024 @ 1:27 pm
Maybe Hatsune Miku will sing at one of the gazillion country awards shows?
May 3, 2024 @ 1:53 pm
I refuse to listen to it.
AI in music (like in everything else)is not a good thing and noone is going to change my mind on that.
Knock on wood but if I ever lose a limb, I’ll lose a limb.
May 3, 2024 @ 1:56 pm
This is functionally no different than having Patrick star singing big iron
It’s dangerous, and once we entertain it there’s no telling what’s next.
Just a few years ago, people were wondering about the morality of re-creating Tarkin for Star Wars
Now the entire media space is abuzz with wondering what AI could have done or if actual artist Scott paid, or if artists are getting laid off because AI is writing the screen plays and doing the animation
May 9, 2024 @ 9:25 pm
Are you saying you didn’t enjoy Patrick singing big iron because I sure did!
May 3, 2024 @ 1:58 pm
There’s an AI version of Hank Williams singing “Straight Outta Compton” set to the music of “Settin’ the Woods on Fire” somewhere on YouTube. Fairly humorous, but the last line of the song is “nothing’ good’ll come from this AI.” That may end up being prophetic.
May 3, 2024 @ 10:11 pm
They shouldn’t have done the whole song but I agree with the last line.
May 3, 2024 @ 2:01 pm
One thing that you kind of mentioned Trigg, is how a bunch of artists are pumping this up. Josh Turner, Cody Johnson, and others have posted mini reels and Instagram stories about their reactions listening to the song. In addition, basically all the comments on Randy Travis’ Instagram post of the song are all artists saying how great it is, etc. It feels very pushed, calculated, non-organic, whatever you want to call it. Just another observation I had. Love the sound of it, but the whole thing just feels strange.
May 3, 2024 @ 2:13 pm
Of course they love it. They can copyright their vocal prints.
Vocal “immortality.”
Vocal Estates.
May 3, 2024 @ 2:01 pm
Bottom Line: The use of AI (especially in this capacity) devalues the music and the legacy of the artist. In my opinion, his involvement is irrelevant.
Randy Travis has a catalog of timeless hits. I am sure there are also unreleased demos, live releases, etc.
This is a disgrace on the country legend.
May 3, 2024 @ 2:24 pm
“Counterfeit” is a pretty good analogy here.
This recording is Counterfeit Randy Travis, just with Randy’s approval.
He doesn’t have anything to lose by releasing it, but the rest of the world certainly does, and it’s like the debauching of a currency. Randy’s recordings have been wonderful. That’s a beautiful thing that is in no way improved, for me at least, by cranking out Counterfeit Randy recordings. It’s like printing money: the more you print, the less it’s worth. Who wins? Just the bankers.
Sad? Worse than sad.
May 3, 2024 @ 7:48 pm
Your commend reminds me—remember when “Johnny Counterfeit” sang on a George Jones song, along with Willie, Waylon & Merle?? That was in 1998. I always assumed Johnny wasn’t available, so they used an impersonator. Wild, and very similar to if they used an AI Cash.
https://www.youtube.com/watch?v=bFK48oz8iTY
May 4, 2024 @ 10:36 am
“Johnny Counterfit” (That’s how he spelled it) was something of a known personality at the time in Nashville, and had his own act, where he’d do impressions of Cash and other singers and famous people like Bill Clinton. It was reported when the George Jones album “It Don’t Get Any Better Than This” was released that “Counterfit” was called to fill in on the title cut and sang together with the real Jones, the real Waylon and the real Hag when Cash was ill and could not participate.
That was obviously a low-tech solotion with nothing like AI involved.
June 24, 2024 @ 8:52 am
Randy’s folks have indeed said how this was produced!! There are videos and interviews discussing it. Get your facts straight geez. Give the guy a break
May 3, 2024 @ 3:23 pm
I was listening to other genres during the Travis era so I have never been more than a casual fan. But this is really weird. It sounds like him, but also a bit robotic. Still, sounds much more authentic than Kasey Musgraves’ vocals on her new album. So there’s that.
May 3, 2024 @ 4:01 pm
People have wondered for years if Randy Travis really was a complete asshole out of the public eye. Him doing this and signing off on this BS recording is total proof that he is.
May 6, 2024 @ 5:51 am
Eh. Travis is a complex guy. I wonder how much say he had. I know he is lucid but there might be pressure on him to produce new material.
May 3, 2024 @ 4:34 pm
hello
May 3, 2024 @ 5:35 pm
Leaving all discussion of the AI aspect aside… the real scary part is… that this is better than 95% of the crap played on the radio.
May 3, 2024 @ 5:58 pm
I’ve seen some small channels on YouTube creating AI covers using the voice of Randy, Keith Whitley, etc. in the past few months. They were almost believable, but I didn’t feel good listening to it.
This is an almost flawless production. Sounds like his late 90s vocals. And I don’t feel bad about listening to it or guilty for enjoying it. It’s clearly brought happiness to a man sadly silenced. And perhaps also because it’s a new song, not a cover.
I’m not sure how much this will be used by other artists. My gut feeling tells me Linda Ronstadt would never sign off on a similar project.
May 3, 2024 @ 7:17 pm
My gut feeling tells me this will be used by many greedy artists and estate holders. This is just the tip of the iceberg, and the begging of the end of true creativity. Sad that Travis would allow himself to be talked into something like this.
May 3, 2024 @ 7:57 pm
Saying he was “talked into it” is slightly disingenuous. How do we know he isn’t pleased?
Aside from anything else, this is a good song.
As for the AI, well: bad health stole his voice. He just stole it back.
May 4, 2024 @ 10:12 pm
How do you know he wasn’t talked into it? Anyone with an ounce of integrity wouldn’t do this. So I believe he was talked into. I have the same proof as you do when it comes to your opinion. He didn’t steal anything back. his voice is gone, and this is not him singing.
May 5, 2024 @ 6:28 am
For Randy Travis, at this stage of his life, and with the financial situation he finds himself in (still owing for royalty advances from WB, no insurance payout when he lost his voice – I read his book), I don’t think he has to be talked into much if it means any kind of revenue (no matter how greedy that may seem). But his producer, manager, etc are all pretty respected people (especially Kyle Lehning), so their reputations are kind of at stake here too. I’m not including his wife only because we all know she’s going to push anything probably without thinking long term effects. But the creative and industry people in his life wouldn’t want this to be their legacy if they thought it was going to tarnish it. I also view this totally different than an estate of a deceased singer (regardless when/how they died) creating “new” music because the deceased singer has NO input. Randy could have objected. He didn’t. And he may have objected to other songs or options presented to him by a whole bunch of other people.
When he’s no longer here, then who knows. But the genie is out of the bottle now. I for one will listen to anything Randy’s voice is on, regardless how or who produced it.
May 3, 2024 @ 7:15 pm
This is sad on so many levels. Artists will no longer have to perform or even be involved in their albums, they will just have some asshole on a computer create it all for them. In the end, AI will destroy all creative industries; everything will be owned by one or two companies. New music with be created solely by AI, as will movies, books and art. Live performances, along with live readings of poetry and books, will be the last place where authenticity and true creativity exist. And when I say live, I’m not talking the computer-enhanced performances the majority of acts rely on today. Hold onto to your favorite records, CDs, cassettes, books, magazines and other treasures. This will not end well.
May 4, 2024 @ 6:02 am
Commercially, I think this is a possibility… but creatives gonna create, maybe some with AI, but most will still put pen to paper, brush to canvas, chisel to stone, finger to string, voice to words because they have to, and because analog is where the soul is felt…
May 4, 2024 @ 10:15 pm
People will continue to create, but no one is going to pay for it, not when they can get it for nothing. I have said this here on another thread, but I have a good friend who is a partner in a gaming company. He’s an engineer and tech genius. He said his company will be gone within three years thanks to AI. He paints a very bleak picture of the future for creatives.
May 6, 2024 @ 9:51 am
This is probably the central question of our lifetimes. Do we want a human culture or a machine society to have the upper hand?
Yes, machines can help us. But should they rule us and take our place, or should we subordinate them as tools to ensure that we do?
May 3, 2024 @ 9:28 pm
Brought a tear to my eye. Hearing something new from my second favorite singer of all time is amazing even if it aint legit. If u love music then u love randy and what he did for music. Devastated by what happened to him and what we lost as music people, this touched me in a strong way. Will always love this guy and anything that comes close to sounding like his one of a kind voice. Anyone who loves music should appreciate this. The greats deserve to live on.
May 4, 2024 @ 2:41 am
While I wish them well and feel Randy’s wife means well, things attributed to Randy have really bothered me for some time. Statements he has supposedly made on people’s passing, comments on various topics, and books. His countenence and demeanor are very different than pre-stroke….simple, frankly. I don’t mean to be ugly, but they have used style and vocabulary that I feel are not him at all. And now they say he approved this, and I have doubts he knew what it meant. I believe we should appreciate his body of work and be happy he is here in whatever capacity. He doesn’t need to be the same to be loved.
May 4, 2024 @ 5:30 am
If I heard it on the radio or streamed it, would I know? Probably not. I love Randy Travis and it does sound like him. I am not sure I like this being done but I really do not know. I am confused and not sure about AI in anything at all. A confusing issue for me.
May 4, 2024 @ 9:47 am
Like the old song says “Aint nothing like the real thing baby”.
Its different when artists like Dolly & supposably Willie are savings songs to be released after they die.
This to me is fraud.
Label them “AI generated” like sometimes when older artists rerecord their original.
Nothing good will come from making AI recordings.
May 4, 2024 @ 10:11 am
I was discussing with a friend of mine not long ago about what “acceptable” uses of AI might be and we both agreed that it could be beneficial to people who had difficulty communicating. Randy seems to fit that definition and if he’d actually written the song, I would be ok with him getting it out into the world any way he could. But, in this case, I’m just not sure he had to record it.
He’s already apparently doing concerts with a variety of guest vocalists, so I’m not sure why the same can’t be done in a studio setting. Randy definitely knows how to pick a great song and surely has learned a thing or two about the production process over the past four decades. I can’t think of any mainstream or independent country artist who wouldn’t be willing to lend their vocals to a track on a project like that.
May 5, 2024 @ 5:39 am
Happy 65th birthday (yesterday),Randy !!!!!!!!! “Where That Came From” means it’s nothing less than great that the supremely talented Mr. Travis is back in studio and hopefully will cut more songs soon. Randy was robbed of his health and his myriad fans of his great music over the past 11 years,and continue your recovery and career,Randy !!!!!!
May 5, 2024 @ 7:36 am
Didn’t ever have any thoughts on AI being used in music until I heard this banger.
https://youtube.com/shorts/kl4Yx7qmGx4?si=y9b_CGmxr5gC7-ny
May 5, 2024 @ 9:11 am
A gentleman I was having a discussion with on Facebook (Ronnie Call) seems to refute the AI music theory. Quote “It’s songs from his past that didn’t make the cut for past records yes he is returning to the studio to give his approval for the release finished product it’s real and remastered songs no AI. I know personally I was just back stage with Randy and his wife and several other country music stars at the NACMAI Hall Of Fame Show I’m the CEO of the North Carolina Country Music Association Randy Travis and his wife are such amazing people it was an honor to meet and talk with them just back in March 17th”
Interesting. Either there is misinformation floating around or the possibility exists that the songs are in fact real.
May 5, 2024 @ 9:18 am
This song was generated by AI. This was confirmed Friday afternoon from a press release from CBS News. CBS Sunday Morning aired a segment that explains how it was all done. I personally confirmed the information with Randy’s publicist as well.
May 5, 2024 @ 9:37 am
Me TOO! I was blown away when I first heard this on CBS Sunday MORNING show today and it made instantaneous tears flow…all I could say was OH WOW…it sounds just like him and the song that was played was one of my first and FAV of his songs!
I so look forward to more AI assisted songs from Randy…and now of course the future is SO BRIGHT for other singers afflicted with anything that prevented them before from singing but not composing….it’s gifts sent by Geniuses..
And assisted by God !! Thank the lord those voices we loved now.can b heard again!!! THANK YOU A I ! Thx also to CBS for airing this heartwarming story!! We now have Randy Travis back again plus many others!
May 5, 2024 @ 1:14 pm
If you’re able to hear it, it’s VERY obvious it’s AI. There’s a digital processing to the vocals that just doesn’t sound human. Is it a good song? Yes. Is it a good Randy Travis song? Not even close. AI doesn’t create art, it steals from it and replicates it.
May 5, 2024 @ 3:39 pm
Yes, It certainly resembles Randy Travis at first, but then it sounds autotuned and then completely synthetic. I got slightly nauseous and could not listen through. I found an old real Randy Travis recording as fast as I could, and listen through it to cleanse my ears and brain.
Unique great voices like Randy and for example George Jones, are probably easier to mimic to a degree so that we recognuze who is being mimicked, but at the same time also hardest to actually make sound real.
May 5, 2024 @ 5:23 pm
For anyone interested who hasn’t yet seen the CBS Sunday Morning story, here’s the link: https://www.youtube.com/watch?v=zM2UzZ4TVGw&t=321s
The story explains how the source singer, onto whose vocal Randy’s AI algorithm is applied, is James Dupre, who’s performed a number of Randy tribute shows with Randy’s blessing.
May 6, 2024 @ 5:28 am
Does anyone know, and I haven’t been able to find anything about this, what the legal implications are for using someone else’s “source” voice for AI? Meaning, do they have to sign away all rights, etc? And in this instance, did Dupre have to try to phrase like Travis would? Because when I would listen to the new song and then one of Travis’ originals from the mid 90s, he tends to hang onto words longer than this AI version does. This whole thing fascinates me, I’ve probably listened to this new AI track 1000 times.
May 6, 2024 @ 3:45 pm
MUmarauder, towards the end of the CBS story, they mention that they’re having trouble getting AI vibrato to sound natural. So, because longer notes often have vibrato, that might be a reason why some of these notes aren’t being held as long as they otherwise might be. Alternatively, it might just be that the length of these notes was what felt right for this particular song.
May 5, 2024 @ 7:42 pm
Sorry I’m late to the party; I was waiting with baited breath for this article. All I can say is, after all these years, to hear Randy’s voice in a new recording moved me to tears. I get what you’re saying about the potential pitfalls, and they do need to be dealt with, but let’s appreciate this moment and hope for more from him. This is a special moment for country fans.
May 6, 2024 @ 5:53 am
Lots of justification for a dangerous volley.
This is how the fall starts.
May 6, 2024 @ 12:00 pm
Skynet
May 6, 2024 @ 3:09 pm
I am truly happy to hear Randy Travis voice again, even though it’s through AI. I do agree that the licensing of the process needs to be registered and regulated, it’s the best way to have all of the great singers of our lives and world be brought back. We have lost way too many such artists in the past years. I for one look forward to seeing and hearing more from Randy and hope that he will be with us for a long time coming.
May 6, 2024 @ 3:26 pm
Where That Came From indeed
May 7, 2024 @ 6:13 am
This is an excellent song, great writing, production and Randy sounds great again. On the other hand, here we go to the unknown. Thank you for starting the conversation.
May 7, 2024 @ 4:25 pm
Curious to see how this plays out, both in the market, the legal sphere and creative ethics. (And I’m looking at this from 40 years in IT, business law and ethics)
The rights of the Tupac Shakur estate are enforceable under several statutes to protect the legacy of Tupac. Those rights were violated by creating and misusing a likeness legally protected owned by the estate. However it appears that there may need to be additional statutes created to protect against misuse of the fair use doctrine – since AI is used to create a “new work” utilizing a protected “template” (Tupacs voice). This also opens up licensing violations as well. All grey areas.
As for Randy Travis, he is the driving force here which avoids alot of AI related issues as far as consent, licensing and copyright. This avoidsvthe Tupac mess. The quality of the AI presentation will dictate how this ends up in practice. True fan’s will certainly be able to discern how well it is presented. AI isn’t there 100 % .
No fan of AI since we haven’t caught up to it ethically. Wheter it be a help desk bot or students cheating. AI has also been used for deep fakes and vulgar purposes as well.
More importantly, what would stop an (unscrupulous)artist to create a library of words, phrases, riffs, breaths etc and create music in the future without stepping in front of a microphone again? Or perhaps a record producer, record company instead. Electronics and computers in music have been around forever. The Moog, Roland, Yamaha and Fairlight were pioneering in their repitore – but most importantly required human creativity to harness those “instruments”.
AI has the potential to remove the human element from one of the most profound forms of expression- the arts. I can see our future, AI new music created from artists alive and dead, packaged into our AI generated playlists all formulated to habits of consumption to feed corporate greed.
An artist obviously should have the ability to manage their legacy as they see fit. Artistsvare also entitled to have their rights respected. I also expect artists who chose to engage in AI endeavors chose to do so with the respect to the fans, with disclosure and transparency.
May 8, 2024 @ 3:44 am
I’ve read lots of comments online of people saying how this is great, wonderful because it “gives Randy Travis his voice back”. That simply isn’t true. It’s scary how many people are swayed into accepting this agregious example of “non-art” through this emotional excuse of RT not being able to sing. This should be rejected whole-heartedly as “art.” How can the cream of the crop of new artists rise when people are increasingly getting duped by their own desires for nostalgia? I am not surprised that this is happening though – this is the logical conclusion of heavy digital assistance in music for the past 20 years. Your average Swiftie doesn’t know (or care) that Taylor uses heavy live pitch correction. And her albums use heavy digital “trickery” to make her sound like that. They are content to think that she is just a good vocalist. People defending this AI track from RT are just as ignorant and stupid. I truly believe that great music comes from some brief connections the artist makes with some other dimension. This connection that artists can make with the cosmos and become a conduit for something new and great will no longer be rewarded if we accept wallowing in ditial filth like the people from Idiocracy.
May 9, 2024 @ 1:37 am
Most people think with their emotions first.
See the last four years.
They will sign away anything if it makes them feel good.
Then five years later, they will wonder “how did we get here?” after full albums are released by dead artists.
Fools and idiots.
May 8, 2024 @ 9:47 am
I have a huge problem when using AI for anything when it comes to any deceased artist that’s like using a voodoo for dummies book to bring a Chucky doll to life.. I honestly have no problem giving a artist there voice back if it’s across the board not the industry picking and choosing who can and can not for an example just look at what axl rose from gun’s n roses vocals sounds like today from what they once were but he is also one of biggest pricks to ever pick up a microphone unlike Randy Travis but to my point i see AI being okay for some artist and not okay others I have more of issue with AI software like ChatGPT used for songwriting which i already have a pretty good idea it’s being used already
May 22, 2024 @ 9:01 am
“How is this any different than a prosthetic or robotic limb being added to a paraplegic?” That’s an easy question to answer. An artificial arm or leg replaces the missing limb and is used by the person as a limb. AI does not allow a person to talk. It records the sound of a person’s voice over words spoken or sung by another person. There’s no comparison between the two forms of technology.
October 4, 2024 @ 7:01 pm
As I see it, if Mr. Randy (Traywick) “Travis,” gets the idea to produce HIS music, regardless as to how it is done, NObody has ANY Right, to downgrade him for doing so. It is Entirely up to his fans, as to how it is perceived, AND accepted! If God Almighty has Given him the Talent, let HIM display and offer it to the fans that Love him! I know I’m in that Category, for him!