AI discussions are almost boring nowadays, with the same arguments being thrown around ad infinitum like "it's just a tool, no need to be scared bro", "real artists will always be cooler", "we're all out of a job in 6 months" etc. etc.

I, for one, think we've seen nothing yet, so any confident statement in any way seems irrelevant to me (unless your name is Philip K. Dick).

However, I've seen the latest video by my main man Benn, and... holy jumping cheeseballs.

It's not just about the tech -which, by itself, is pretty fucking impressive, but like I said, I think we've seen nothing yet.

No, what excited me is that this is the first instance I've seen where the developers of the tech are also involved with the ethical and legal ramifications of their creation so the artists will be the true beneficiaries.

Not big conglomerates. Not publishing companies. Not fucking Disney.

This is the first time I'm seeing a potential for AI powered art for the artists, by the artists which, if implemented, will completely bypass The Man.

And unlike Silicon Valley disruptive tech bros whose sole aim is to dethrone The Man just to take his place, we're talking about people who wanna get done with copyright laws as we know it.

Anyway, I've rambled enough. See it for yourself.

Cheers.

 

Youre correct with the assumtion that we have seen nothing yet. To me the AI regolution will be like the invention of electricity. Its like people saw the light bulb and were like ok its cool but we had candles for that before.. but everything that happened after that which relied on electricity wasnt even thinkable yet. Like radio, tv, computers and everything else..
Im pretty sure with ai it will be the same. Things will happen that arent even thinkable today..

 

chinese ai supremacy

quantum and ai are just buzzwords
some aspects work, a lot of it is just bull

 

Benn is goated. Regarding AI, of course we haven't seen nothing yet, the current state of AI tools in music production is laughable, it's crappy and garbage compared to every other field... but it has a bright future. I can already foresee the future where people will create music using devices that can translate what's going on in their mind to the DAW (or whatever future people use). It's obscene how machine learning will change the world a few decades from now.

 

I recently watched the new Indiana Jones movie. I wasn't looking forward to watching Ford, who is almost 80, trying to run around like a 30 year old. But they used AI to make him look younger/put his face on a stunt man/actor and I have to say it's very impressive. It actually brings back the style of the original movie and it's a joy to see a young Ford in a new role.

I envisage a time where we'll be able to input our own movie synopsis, list of favourite actors and AI will create a whole movie, completely original and as action packed or otherwise as we want it. No wonder actors are striking. The only question is how do the studios monetise this?

 

So far (afaik) AI is nothing more than "really good algorithms". Maybe our actual intelligence is really just that too but computer AI has a way to go before I regard it as sentient, which to me is a prerequisite for the kind of thing we're talking about. That, and exhibiting behaviors beyond those originally designed into it.

 

this movie promise was first uttered in the 80´s
the best they could do you can see in heath ledgers last film and now in in the indie.

but they are striking because, some ai tools ma have been used but, it is as always VFX artist who bring the magic, underpaid and exploited.
that's why so many Oscar winning VFX house went bust and some slimy studio exec accepts the awards (in their absence)instead.

 

It'll go the same way it always goes.

People will find out they were fucked over for AI training, a tiny minority will fight tooth and nail in courts for the next decade to reclaim their creative output.

Two or three high profile names will win, because that's how it always goes.
The other 98.21% will lose, because that's how it always goes.

Governments and courts will say 'Taylor Swift won, what more do you want? It's time to move on', because that's how it always goes.

But you can't make an AI truly and genuinely forget anyway. Just like your brain connects many thousands of seemingly unrelated ideas, impressions, experiences to "Taylor Swift", AIs too have no single memory cell under which they could store everything copyrighted to "Taylor Swift". "Taylor Swift" can be perfectly recreated without knowing any of her stuff. An accurate copy of her exists in the billions of TikToks and Shorts of people discussing the way she dresses, behaves, puts on her makeup, humming along to her melodies, etc. So even Taylor Swift, who won a big stack of money, will lose. So really, everyone loses. Except the big corpos.

For a couple more years after that, a small group of people will remain very vocal. "They can't get away with this! How can they keep getting away with this??" - but they will fall silent eventually.

There already is and will be a grassroots movement where people install their own stacks of single-purpose, free and independent AIs from open repositories.

But big corpos won't allow it forever. It'll start with takedown letters to GitHub, then court cases, then they'll threaten to leave for a country where """the protection of innovation is guaranteed by the law, unlike here"". Or maybe they'll threaten us with paying negative taxes. The courts and governments will inevitably side with the corporations, because that's just how it is.

And so will most people actually, because they've long since forgotten that it was their own data and tax funded research that enabled the big corpos to develop the AIs in the first place.

"That's our patents, technology and data! Those AI-hippies are enabling the dictatorship in Arstotzka! We're the good guys here! Won't anyone think of poor, upright Google?!".

And that'll be the end of it.

 

But for how much longer?

 

I liked the video. Funny thing is, people here on the forum are always downplaying the problem. Could it have something to do with the type of service provided (music for multimedia products) and, in this case, the lack of unions for audio workers? Ask the extras, scriptwriters and VFX artists if they are calm (and consider that some of these workers have managers and agents, even regulations and unions that, at this moment, are working hard for them). The video shows that a good result with AI still depends on a lot of "cleanliness". And I think it's great how much of humanity has difficulty in projecting events over time. But what about a year from now, or two, will it continue like this? AI is currently limited. In the visual image/video processing fields, it is already a reality! In the sound field, it may take a little longer, but it's just a matter of time and processing limits: All it takes is for these limits to be made more flexible for art professionals to be shitting bricks!

 

Last edited: Sep 4, 2023

I don't mind AI in movie FX because sometimes artist are not able to fulfill the directors vision.
But with music it's different, you take away the human element and you take away the soul of the music.
It's just a matter of time before the term 'Musician' or 'Artist' will be changed to 'AI Musician' and 'Ai Artist'

 

AI was developed for internet monitoring and censorship, which doesn't require any creativity. Likewise ML is very good at copying direct sound, e.g. NAM amp modelling, but music comes from the heart and represents human feelings and emotions. I think AI should stick to what it does best and let humans be human.

 

But if AI becomes intelligent (the "I") it will theoretically be human, at least intellectually. Or better.

(Is saying "intellectually intelligent" a tautology??)

 

T1
\

You got that right. I just finished an album this week and it is currently on the way. All of the songs in the album are written by ME, produced by ME, sung by ME, but the voice was converted into AI and pitched up to get a kawaii hyperpop feel. The link below is one of them.

(My main genres are trap, hyperpop and industrial pop.) I basically turned myself into a vocaloid, all for only $9 a month. TAKE THAT YAMAHA! Looks like I didn't need your sorry ass after all. WOHOOO. I don't have spend another two years learning how to use a $299 software on top of my 3 years worth of music experience. FUCK CORPORATE AMERICA, BITCH. I'm GONNA BREAK THE INTERNET!!!!!!!!

 

So generative music isn't music?

Humans being human = humans using technology to achieve things that would be impossible if they solely relied on their natural abilities.

How is it ok with movie fx but not with music? I mean it's still an fx.

The discussion here isn't whether AI is sentient lol.
It's about a really good implementation of the technology, but more than that, it's about a new way to think of copyright laws.

 

AI will never know what it's like to suffer from a broken heart

 

It's all semantics. To me, if AI isn't sentient, it's just a stupid buzzword, like virtual reality. That term means something, but it's used to just mean "I have a display in a set of goggles".

 

Well, not literally I suppose. Although there's always cyborgs.

 

That's my personal opinion but here's why. There are thing's AI can do in Graphics that if done manually would blow the movie budget out of proportion. There's a reason why the director of Avatar had to wait 13 years to create his movie.

As with music.
I wouldn't call myself an artist If I used AI to create music.

Ennio Morricone's La Califfa is probably a great example of a composition that AI level hasn't reached yet or probably never will.

 

Last edited: Sep 5, 2023

Benn Jordan's intent is noble, but they have little chance of making a difference for most. There are literally tens of thousands of really good singers to provide nice, clean data sets to base future voices on, who won't know or maybe care about future royalties. And as the OP stated "I think we've seen nothing yet", in the not so distant future this data will be able to be played, fashioned, tweaked, and "performed" to sound like anything, or anyone, and sound really, really good. Voices will even be built not based on any particular artist's performance data set at all, just the mathematics of the human voice. The "artists" will be the ones who use the software to create the performance. Andy Blaney's incredible compositions and mock-ups using Spitfire products are an example. He didn't play the instruments, but he's a virtuoso with the MIDI programming. Benn proved in the video that his own lame voice could be made to sound much better with today's tools. Tomorrow's tools will be so much more powerful.