What does the future hold for Reaktor?

Z Gabr

https://community.native-instruments.com/discussion/comment/158048#Comment_158048

I want to believe it…

KoaN

https://community.native-instruments.com/discussion/comment/158048#Comment_158048

This is hard one to predict for sure!

As a guy who spent his whole life obssessing on creating new,unique,personal sounds Ai is scary for sure.

I can still do whatever i want personally but the scary thing is of course how the mass will react,how people will consume,think,react about this change. For me it comes down to discernement,education,learning.

We already see that strongly with electronic music where most people don't know the difference between a DJ mixing songs or a guy spending months crafting his sounds in unique ways "i know some Djs are more creative than others". Combined that with how most softwares want people to make a hit with a few parameters,infinite numbers of prepared samples,elements for making a hit…most people don't know a damn difference between all these things,they think it's all done the same way.

So anything creative,well crafted,personal has a chance to be completely ignored,specially if it is not trendy.I don't blame people,if they don't know they don't know…but would be nice to have a bit of education,culture about this to inform people.

With traditional music it is a bit easier to spot what is good and creative i think,someone who knows nothing about guitars will probably react seeing Jimmy Hendrix play. I guess because you can identify the way the guy moves,the fingers,expression etc. The sound is easier to follow and identify too.

So with Ai i feel this will make this even worse! The distinction between,real not real,human crafted,not human crafted,what is creative,what is not,what is original,what is not will become even more blurry.

So some people will feel they have heard it all already,a guy who does some crazy sound design and electronic music won't impress anyone when you have heard the Ai doing even more crazy stuff. Unless like i said there's some discernement,knowledge,integrity about this all.

But i doubt it…really. it will just blur everything even more and maybe like you said people will get fed up,react and go back towards acoustic stuff because they can't see what is what anymore.

About Autotune…for god's sake i cannot tell you how i despise this sound hehe. First time i heard it,around 2000? I thought it would quickly die and stop….my god was i wrong. And you know about discernement…a lot of people are fed up with autotune also but a lot of people will confuse a Vocoder or any effect on a voice with Autotune. For them it is all the same.

Paul B

If AI does get to the point where 'prompters' can generate music without actually doing the kind of work that requires tools like Reaktor, then all the virtual instruments and DAWs will be dead in the water. 

I don't wish to get too deep into the AI debate, but I've followed the development of what became the thing people call AI today for a long time, and there's a fundamental problem with it that is not solvable with any knowledge we have currently or will have any time soon (and by soon, I'm counting in decades, not years). While development of LLMs has made gigantic strides, actual intelligent machines have made – as far as I can tell – none since the first whispers of computer based AI research many decades ago. We don't have the first idea of how to make an intelligent machine.

To be clear: what I mean by intelligent is – the machine can think.

Because machines can't think and won't be able to any time soon, AI will remain unable to make music that is unique to the AI. That is, while humans will still be able to make electronic music that is infused with their personality, machines will not. Because they have no personality, and personality is not something that can be imitated. It is by nature…personal.

That is not to say there will be no changes in the music world due to AI. But huge numbers of people are still invested in listening to music that has personality. And huge numbers of people still want to make music. Not ask a machine to make it for them. If financial success was a major incentive for people making art, we'd not have most of the art we do. This art was born out of a desire to create.

And I believe people will find ways to distinguish themselves so that it will be possible, perhaps requiring a little more context than solely listening to something on a streaming service, to identify that this particular music you are listening to was not made by an AI.

People talk about an impending AI reactive renaissance of music made by people physically playing instruments as though that ever went away. As someone who goes to a lot of concerts, most of them are by musicians doing exactly this. And the electronic ones are people who have a style that could only come from them. Sure it can be copied now by an AI, but for it to be copied it had to first be made by a human. And the best electronic musicians will continue to evolve their sound. AI can at best be a step behind them at all times.

As for

So some people will feel they have heard it all already, a guy who does some crazy sound design and electronic music won't impress anyone when you have heard the Ai doing even more crazy stuff.

I defy AI to create anything weirder than what humans have already done.

But more than that: I defy it to create something really out there which has the aspects of human weird music that make it more than just weird. AI weirdness will be controlled (or perhaps uncontrolled) chaos. But human weirdness is informed chaos. People making things weird do it for a reason. AI making things weird does it because it was told to.

Ask an AI to explain the process behind the music it produces, and if it even grasps the context of the question (debatable) it will be entirely unable to provide a coherent answer.

With regards to discernment and people thinking electronic music is easy and all people do is push a few buttons, well the first time I heard that was around 25 years ago. And it wasn't new then, though it came mostly from people who should have known better. People who claimed this with confidence because they were self-proclaimed "discerning music listeners" who knew what real music sounded like. It's true that it is easier for anyone to be sure that Hendrix is doing something special by watching him play (though I guarantee you when he was first on the scene many would watch and proclaim it noise made by a man with no talent), while watching Autechre (as much as you can watch a duo who play in the dark) it could be nothing more than a generative algorithm. But at some point you have to take these things on faith and believe that the majority of people who will put themselves through what it takes to be a professional musician wouldn't do that only to fake their music. Musicians using AI to cheat will flood the market with their generic productions and only make it more difficult for themselves. True, it will also be more difficult for people who make music that is personal to them, but they will have an advantage: when asked to explain their creative process they will be able to because they actually have one.

Well crafted personal music has been ignored by many for most of the history of recorded music. There are numerous bands who recorded incredible songs and were know by a tiny fraction of listeners. AI may make this worse, but it could also provide an incentive for real music lovers, of whom there are far more than AI pessimists assume, to make a greater effort to seek out something truly original and personal.

—I'm tired and not sure how coherent that was. But it's pretty well representative of my views. Normally I'd edit the heck out of it but tonight I'm posting it mostly as it came out.

Kubrak

AI has made enormous advance and advance will go on fast. It will undoubtelly make yet another leap with introduction of quantum computers…..

But still, I agree with you, that while it may make nice music soon, it will be just yet another copy of star XY. But the same already do many humans, even pros…. So many bands/singers sound like XY…. They are just better or worse copies….

But yes, to create something real new, that is pretty hard nut even for humans, so it is most probably beyond capability of AI for at least several decades.

https://www.youtube.com/watch?v=_EZQx87DyzM

https://www.youtube.com/watch?v=qgHeCfMa39E

https://www.youtube.com/watch?v=-9EM5_VFlt8

KoaN

https://community.native-instruments.com/discussion/comment/158122#Comment_158122

About personality in music from Ai,it is a subtle thing.I am not sure most people could identify that quickly in Ai vs human like let's say in a blind test. Well for now we can usually spot differences,imperfections with Ai stuff but it is getting near human realism very quickly. I think often we put our own soul into something we like,something that makes us emotional.

As an example my friend is a professional illustrator and is already seeing consequences of Ai stuff. He has a few stories of companies hiring a guy using Ai to generate art.My friend can usually spot what is Ai made but he has many debates with people sharing Ai images without knowing and just thinking it looks cool. Most people won't see the difference and just share what they like. That is why i often come back to discernement,education in that case.

" I defy AI to create anything weirder than what humans have already done."

I used the word weird but i meant more in the sense of new,unseen. Of course the Ai learns from what has already been done but the realignement of this immense data base can create very unique stuff. It is behind soundwise compared to image but it will quickly catch up. And asking the Ai to do a full song is harder than just asking for a sound or clip of course but still,look at how sample packs,presets are popular…won't be long before Ai is able to make any sound design and then you assemble the song yourself.

" True, it will also be more difficult for people who make music that is
personal to them, but they will have an advantage: when asked to explain
their creative process they will be able to because they actually have
one."

I wonder if people will be interested in explanations though…

My friend just talked to me about this Ai. I haven't tested it and the shown clips are mostly genre,things we have heard before "my friend said you can have crazy result with the right prompts" You can hear imperfections a little in the resolution and in the voice at times but damn…from what i hear you can easily edit the song,hire a real singer,musician or whatever and you have tons of hits in their that most people will listen to and not question at all. I am talking about people who just wants to listen to a song and like it,most people.

https://www.udio.com/