Recent Posts

The big guy is a survivor... and an inspiration.

Had AI been around in the 60s, might Brian have considered using it?

Since it's all hypothetical, I'll say yes - with exemptions.

Consider how Brian in his teens, after getting a reel-to-reel tape deck as a gift, was using that technology which at that time in the late 50's was pretty much a decade old, and that's it. We have home tapes of young Brian stacking his voice through overdubs to replicate the sound of his beloved Four Freshmen. Listen to Brian's trademark high-range vocals: It's damn near the same as Bob Flanigan, in tone, phrasing, and general overall timbre to carry the high voice lines in that vocal stack. It can be uncanny at times to hear Flanigan and compare it to a 60's Brian vocal, circa 61-64. But it isn't actually uncanny when we consider Brian modeled his voice and vocal technique after Flanigan's with the Freshmen.

So breaking down all notions of reality and the time/space continuum, if that young Brian were using new and developing technology at that time to recreate the vocal sounds of his favorite group, why wouldn't it be probable that if a technology existed which would allow him to recreate those vocals as the AI programs are doing now, he'd be at least messing around with it and experimenting? Let's call it "analog Brian", and say if he were trying to have his own "Four Freshmen" in the form of overdubbing himself on a reel-to-reel in his teens singing other material, why wouldn't he do the same thing if another tool existed to do that same thing?

Remember that fragment of Smile which became known as Dixieland? That was a vocal game the Boys used to play in the car and elsewhere where each voice would riff on improvised lines like a Dixieland jazz combo, with the voices as the instruments. Now if tech existed where Brian could actually manipulate what those voices were doing and actually have the voices turn out sounding like a clarinet, trumpet, etc...don't you think he would have at least tried it?

Brian's willingness to work with new technology, new techniques, and new ways of recording music throughout the 60's and into the 70's is sometimes left out of the discussions. One of his favorite albums is and was "Switched On Bach", if that adds anything to the discussion. A totally "new" sound using one of the first commercially available Moogs to exist, and I think it inspired Brian's use of synths well into the 70's where some would consider him like a pioneer of synth-based pop tracks with the Love You album.

If a new technology came out that allowed him to experiment with sound, I'm sure he'd be all in. Would he use it on commercial recordings when he became a professional producer actually cutting records? Who knows...if it served a song he was working on, perhaps so. But if the teenage Brian was trying to recreate the voices and sounds of the Four Freshmen using the high tech recording device of its day, why wouldn't it make sense that he would experiment with similar technology if it existed.

Carnie Wilson Talks Singing Beach Boys Songs With Dad Brian Wilson After Conservatorship Filing

https://www.youtube.com/watch?v=1abygPaCncw

Thanks so much for posting this. Steak, Mac-n-cheese, Dodgers games, and Be My Baby. What an awesome life update. Glad he's doing well!

Oh, that was so heartening to read.

I believe this was the track they were recording after Brian's final tour concluded.  Which would explain Brian's rather fragile vocals.  Apparently Darian is handling all or most of the vocal harmonies.  Overall a pretty good track despite Brian's weakened condition. 

I hear Brian pretty clearly in the harmonies, especially at the beginning and the end. He has what sounds like about two parts in the baritone register. The falsetto sounds like Matt to me, but I'm not sure.

Had AI been around in the 60s, might Brian have considered using it?

I'm going to go with no. There would have been zero reason to use it back then, as they all had fully intact voices.

When we're talking about "AI" here, I think we're mainly talking about fake "soundalike" vocals. (If we're talking about generating *new* music from the ground-up using AI, that's a WHOLE other ball of wax, and to me even more goofy and pointless). And in the realm of vocals, Brian *could have* (and probably DID) sing "Guess I'm Dumb" back in 1965. He didn't need AI.

There was a point in the 80s and even 90s when some musical artists spent MORE time (and probably money) futzing around with programming drums instead of just having someone come in and PLAY them. And I think there's a lot of that happening with AI right now, especially with AI visual art, where people have zero interest putting *any* manual work into it. They type a description and their PC does the rest.

Each time a new thing comes along, people, understandably up to a point, compare it to some previous tech.

I just don't think using technology to make alterations to someone's work (e.g. autotune, any number of effects, machine-learning-enhanced audio splitting/separation, etc.) is the same thing as some random person singing into AI software to try to emulate what a *real* person would have sounded like under whatever the circumstances may be (and it seems to often if not usually be a younger version of the same singer).

An autotuned Brian is still *Brian's* voice run through whatever it may be.

With the AI vocals, it's some *other* person singing, and then using a program that has gobbled up many examples of Brian's voice, and then spits out an artificial, computer-generated voice guided by that random person singing.

Everybody has a different "line" I guess. Sort of like some people say the Beach Boys ended with Dennis's death, others say Carl's death, others say they're still *THE* Beach Boys now.

My cutoff on "Fake vs. Real" is the point at which the "voice/singer" in question plays no active role in what I'm hearing.

Make no mistake, it's "interesting" on some level to hear like Al and Mike filling in "Smile" vocals based on Brian's 2004 "Smile". Undeniably. It gives a bit of the flavor for what might have been. I recall hearing, pretty early on in the "AI vocal" fad, over a year ago at least, a 1980 Paul McCartney singing "Beautiful Boy." Did it sound like Paul? Not really, but it was as spot-on as "AI vocals" can get, and it was interesting as a curiosity to hear once, to briefly get a taste of imagining what the real thing would have been like.

But I can't live with the stuff, it doesn't work to actually continually listen to.

Had AI been around in the 60s, might Brian have considered using it?

I think this all comes down to a very basic question: what do you prefer? I remember being on the other side of debates about this for years when it came to pitch correction software. I believed at the time and still believe that fixing flubbed notes via technology is no big deal. Many other folks felt otherwise. The reason those debates went on and on without any real resolution was that neither side was objectively correct or incorrect. It boiled down to what you liked or didn’t like.

At this stage, AI Beach Boys tracks usually feature a familiar backing track with a lead vocal run through a processor that makes someone sound vaguely like someone else. It’s the equivalent of wearing rubber mask. Yeah, if you squint, it looks like the person, but everyone understands there’s no way it can be. If the fantasy of The Beach Boys or Brian Wilson singing certain songs in different eras or with different stylistic approaches means so much to you, I don’t think you’re a bad person for creating AI or enjoying them. I’ve enjoyed some of them myself.

But there’s no way that what’s being produced has much to do with how we’ve traditionally understood the creation of recorded music. It might end up putting musicians out of business, and there’s also nothing that I can do about that. But I can just say at this point, what my preference is.

To add to your points, Guitarfool: This reminds me of when ‘That’s Why God Made The Radio’ was playing on our local rock station, and a friend of mine said “Hmmm that’s a good Beach Boys cover band” not knowing it was the actual Beach Boys (with help on those higher notes from younger talent).

My point being that even The Beach Boys themselves see the value in trying to sound like their younger selves - because that is what people relate/gravitate to. Granted I would so much rather hear Matt Jardine sing harmonies than an Artificial Intelligence Al Jardine, but in a recording setting they’re both somewhat (and I stress “somewhat”) trying to do the same thing - sound like the vintage Beach Boys. I’m not speaking strictly about the official nature of The Beach Boys releases, or the moral implications of AI - I’m speaking mostly about what listeners gravitate toward (and thus what is released) as it relates to the sound of The Beach Boys music.

There are also people on YouTube who try to do exact note-for-note covers of Beach Boys songs (to the point where they dress up in the blue striped shirts for their music videos). If all they are doing is a line-by-line copy, why wouldn’t we just listen to the original? Because some people like seeing the talent that other creatives produce. And those people making the covers enjoy the process/art of it all. Plus it’s neat to hear different inflections/ideas that the band/artist may interject.

TBH I am completely on the fence about this music. On one hand, I completely agree with Guitarfool & others that this music has the capability to enhance our personal enjoyment of the music. On the other hand, I agree with HeyJude about how ‘uncanny valley’ it is. And furthermore, even if I enjoy a song like ‘Thank Him’, I will still get this small feeling in the back of my mind that these creations are somehow ‘wrong’.

[So there's a point where it comes back around to that Simpsons fireplace gag. Why are we listening to this stuff? Is it worth listening to real stuff instead?I don't think tribute bands and things of that ilk are particularly comparable. It may be comparable in the overarching question of "Should I burn time listening to this when I can listen to the real/original thing?" But tribute bands serve a function of presenting live music. Nobody is trying to make me think that's John Lennon on stage singing in that Beatles tribute band. These AI vocal tracks seem to at least be *striving* to actually sound like the real deal. And what ends up happening is that they mostly sound awful, and the few well-done ones sound 75% towards sounding 75% like the real thing they're trying to sound like.]

As far as where the criticisms of AI came from, I think that's a pretty complicated question.

Certainly, many people take a very staunch zero tolerance attitude towards it because it involves a machine doing some (or most or nearly all) of the work rather than a human with a brain and a soul. There is also the issue of lost jobs in certain sectors.

To me, while those are all pertinent issues as well, one of my main issues is that the the seeming fundamental premise of these AI vocals, the ones that are presented to people, is that they actually sound like the thing they're trying to make it sound like. And I don't think it does. I've probably likened it in the past to those "Magic Eye" posters from the 90s. Some people hear it, some don't. Some people don't have a good enough ear (I guess, and I know this can unavoidably sound kind of snobbish) to hear the nuances that *don't* sound accurate. Or, I guess, don't care?

Now, some might argue that they acknowledge it doesn't sound "just like" the real deal. Okay, so then what are we doing here? The answer is, on one level, harmless tinkering for one's amusement.

So there's a point where it comes back around to that Simpsons fireplace gag. Why are we listening to this stuff? Is it worth listening to real stuff instead?

I don't think tribute bands and things of that ilk are particularly comparable. It may be comparable in the overarching question of "Should I burn time listening to this when I can listen to the real/original thing?" But tribute bands serve a function of presenting live music. Nobody is trying to make me think that's John Lennon on stage singing in that Beatles tribute band.

These AI vocal tracks seem to at least be *striving* to actually sound like the real deal. And what ends up happening is that they mostly sound awful, and the few well-done ones sound 75% towards sounding 75% like the real thing they're trying to sound like.

The lines in bold:

The same things could be said about tribute bands, almost point by point, which is why I raised that comparison. First, you can't judge every example by citing the worst examples of anything. There are restaurants that cook an amazing burger, and others that are pure crap: That doesn't mean all burgers are mostly awful. It's the same with AI creators and live tribute/cover bands, there are total hacks and there are amazing acts out there, they shouldn't all be judged on a bell curve or lowest common denominator by nature of the media or fields in which they choose to make their music.

I'd argue that the tribute bands are doing exactly the same thing you've described in striving to sound like the real deal, and the 75% description is the same description as can be applied to the live tribute bands. I've worked with some through the years, and can say that they are absolutely trying to sound like the bands they're paying tribute to, and in some cases especially with the larger and bigger-budget touring acts, are trying to also look and dress like the original acts, and are even playing the exact instruments the original players used to further put across the fantasy to the audience. They're very dedicated to presenting an authentic look and sound to their audiences.

So they're out there trying to copy the originals and create a fantasy live sound and look, never claiming to be anything but a band paying tribute to another legendary band...and I'd say AI creators are doing the same thing only from the studio versus live.

I guess some people are telling those fans who buy tickets to tribute shows that it's better to stay home and watch or listen to live recordings of that original band, but I doubt that affects the decisions of neither those fans nor the tribute musicians themselves because they're doing something that people enjoy and providing some escapism and fantasy for all involved.

Are all tribute bands proficient or even "good"? Of course not. Are all AI creators and creations proficient or "good"? Of course not. But every burger you get at a restaurant isn't always "good" either. The market and audiences weed them out.

I think the tribute band's version of "John Lennon" or Freddie Mercury is absolutely, 100% trying his best to sound and/or look like John Lennon or Freddie. The difference is they're using analog methods and technology to do so while the AI creators are using modern digital technology to do the same thing. I see a direct connection between the two, the only difference is the medium and technologies being used to accomplish what amounts to the same goal: Create a fantasy experience for fans who want to experience something that is physically and scientifically impossible in the real world.