Where Will Digital Audio Workstations Be In 10 Years ?

[telecode]

I was trying to strike a conciliatory tone. I missed the mark. - Scott

My issues with the things the one guy says has nothing at all to do with my own ego. You do not speak for me.
[/quote]
[/quote]

you can't strike a conciliatory note. that poster is a highjacker and flamer. almost every single thread where she shows up i wind up muting. there are a few others like her on this forum.

[vurt]

one g in lego.

[vurt]

as a collector, i find that irksome. sorry.

[toonertik]

is he not a dj rather than a musician?
different thing surely?

although i never understood at such events why everyone faced the dj? personally, the amount of drugs i was on, didn't care where the dj was
but i was everywhere and nowhere!

Upside down.. down side up...
or was it up.. Up... UP..
I loved the M25 wharehouse things...
so many rooms... so many sounds... so many dancers... etc etc..

[toonertik]

PCP all over again

Once in a motel in Midwest City, Oklahoma this AWOL from the Army woman dosed me and my friend so she could f**k him, and tried to when he was in the shower. I was out like a light.

haha... PCP... nah Ket ain't as powerful as PCP..
1976 I met a couple of LA freaks in London and stupidly asked what the story was with "Angle Dust.. PCP"... Muc to my surprise they asked if I wanted to try it... Well being young and adventerous, I said "why not"...
Next day I said NEVER again..
29 years later in a wharehouse club I met a young lady and got lead astray (that's my story and I sticking to it) any way she sorted my out with K...
It was quiet fun, not a strong as PCP...
It was like the movie of one's life at each moment was a st ip of film, with all the frames cut up and shuffled and that it was you saw and was you moment... crazy!

[toonertik]

As to DAWs in ten years...
They haven't really changed that much in the last 20 years or so. There are still linear DAWs, still Reason and Acid.. newcomer with new paradigm was Ableton (but even that paradigm was not new... in 1981 I (and others) was using up to seven special tape players (8 track things for 2 track pro playback) and cutting my own loops and then firing when needed during a recording. Ableton just put that into a DAW... as to AI music, Eno and others were exploring that years ago, sort of.

I don't see those studios that record live musos doing much different.
EDM etc is real just a small part of music, to much surprise to the ... shhhhhh

[telecode]

As to DAWs in ten years...
They haven't really changed that much in the last 20 years or so. There are still linear DAWs, still Reason and Acid.. newcomer with new paradigm was Ableton (but even that paradigm was not new... in 1981 I (and others) was using up to seven special tape players (8 track things for 2 track pro playback) and cutting my own loops and then firing when needed during a recording. Ableton just put that into a DAW... as to AI music, Eno and others were exploring that years ago, sort of.

I don't see those studios that record live musos doing much different.
EDM etc is real just a small part of music, to much surprise to the ... shhhhhh

Yes, I agree. DAW's haven't changed much in their basic functionality. But keep in mind that there has been lots and lots of perfections and improvements that have been put into the products by the respective companies that make them. It will only get better and faster as time progresses.

I still have my old Macbook with Logic 9 on it -- but the reality is, I would never use it because I can achieve so much more so much faster on my newer setup. Even simple things like making transitions or splicing stuff together is so much easier and more well put together in the newer products. And I am not even your regular electronic music geek type that has tried every single synth and DAW on the market. A lot of younger music makers flip around from one vendor to another all the time.

Brian Eno didn't have access to the power and technology required to make AI music. The new AI tech is pacing miles ahead of what was even happening a few years ago, not to mention, many years ago. I think with AI in music, it all comes down to what the advantages are of AI in other fields. A process that is predictable and repetitive can be turned into an algorithm that works endlessly without stopping in a AI system. So music that is predictable and repetitive could be made by an AI. It's sort of like Spotify and they algorithms that runs their platforms to make the services customizable to you. If you are a regular predictable listener that listens to defined music tastes -- those music weekly playlists are going to nail you perfectly and give you all your subscription moneys worth. But if you are an ADHD type listener that listens to every genre under the sun as your interests and moods dictate, those playlists aren't going to work that well for you.

But!! Modern forms of music are not predictable and repetitive. They just sound so to the ears of someone who isn't into the genre. Its sort of like me saying, death metal is predicable and repetative not because I understand the nuances of what is death metal but because I am not a fan of the genre and don't listen to it, so in essence, i can't tell a good from a bad death metal song, album or artist.

I have grown a huge amount of respect for the new younger chillstep artists that are plugging away at the chill step scenes. They put a shit load of work and effort into what they do and you can really tell the good ones from the bad ones because the good ones don't release half made stuff. they put lots of time into what they do to make it what it is and are not on auto-pilot mode. they have to compete with each other and each new artist that jumps onto the scene and the only way to compete is to make the music better than the others. it takes a lot of skill and time to make what is essentially an electronic instrumental song to not sound lame and boring. they use every tool at their disposal and there is true artistry into using those tools and tricks in an artistic and creative way that doesn't overwhelm the listener.

[DuX]

I'm tired of using computers as DAWs since they depend on some OS which is in all cases not made for making music, but for browsing, office stuff, playing games etc. I would like to see a dedicated hardware DAW with an OS that is customised and dedicated for music production only, completely optimised for real-time low latency work with audio, with the lowest possible number of background processes. You press the power button, and there is your DAW ready for making music.

My hope is that somebody will make it possible. Make a dedicated hardware with custom, cut down Linux, or QNC, or BeOS as OS, and equip it with precise multiple MIDI, gate/cv, and audio ins/outs. That thing would sell. I'd be the first to buy it.

I hope that is the future of DAWs, because this way we're tied to corporate OSes who don't care about our work and performance of their OSes for music, and music companies still don't see Linux as something they should support. If they supported Linux, you could run all the music plugins and DAWs on an OS that you can customise as much as you want. This is not possible with Windows or MacOS.

I would also like this piece of hardware to get away from using the mouse for everything as much as possible, cause mousing around is a creativity killer. I've been using the mouse since Atari times, and I still don't find it completely natural for music production. Knobs and faders are more like it.

Cheers!

[telecode]

I'm tired of using computers as DAWs since they depend on some OS which is in all cases not made for making music, but for browsing, office stuff, playing games etc. I would like to see a dedicated hardware DAW with an OS that is customised and dedicated for music production only, completely optimised for real-time low latency work with audio, with the lowest possible number of background processes. You press the power button, and there is your DAW ready for making music.

My hope is that somebody will make it possible. Make a dedicated hardware with custom, cut down Linux, or QNC, or BeOS as OS, and equip it with precise multiple MIDI, gate/cv, and audio ins/outs. That thing would sell. I'd be the first to buy it.

I hope that is the future of DAWs, because this way we're tied to corporate OSes who don't care about our work and performance of their OSes for music, and music companies still don't see Linux as something they should support. If they supported Linux, you could run all the music plugins and DAWs on an OS that you can customise as much as you want. This is not possible with Windows or MacOS.

I would also like this piece of hardware to get away from using the mouse for everything as much as possible, cause mousing around is a creativity killer. I've been using the mouse since Atari times, and I still don't find it completely natural for music production. Knobs and faders are more like it.

Cheers!

They already make something like this. The MPC Live. It retails for about $1500 here.

https://www.akaipro.com/mpc-live

It's good but not quite there. I am miles faster with a mouse + keyboard + keyboard shortcuts + a 8 x 8 controller pad type thingy than I am on that touchscreen.

But, If I were to put on my prediction cap. I would say that the possibility of making and working on music without a mouse and keyboard is realistic within next 3 to 5 years. Right now, the mouseless touch screen interface its getting there, but it's not quite there.

[Timfonie]

We'll be living in the digital age of plenty where we can't be sure if a human or AI created a song.

Expect the reappraisal of live music. Acoustic live music in particular. You can't fool it's created by (the help of) a AI DAW!

This is just laughable. Ludicrous really. I have to wonder if you have the first clue about what goes into real music. No, if you really did you wouldn't have this particular fantasy.

An AI to be comparable AT ALL to a human will have to... let's consider Rachel in the movie Blade Runner.
Note well that the memories are artificial; but for purposes of arguing this, we'll go with the belief in the memories is the same as having real memories. When do you think this is going to come to pass? What has to happen in the evolution of technology? Specifically. This is just an assertion. Where is the support for this? And why do you want this to be true?
Frankly that's quite suspect afaic.

In addition to this, the AI composer vis a vis a human being will have to have interactions with musicians over a lifetime and a base of knowledge equal to a *real* composer; and lacking a sense of touch is going to be lacking. Don't forget the goalpost is "we" can't_be_sure. In general we're convinced. Just as a simple problem, "we're" going to be convinced by the vibrato of a robot? What has to happen for this to be the case. The years of experience and practice, real technique and an actual fully formed personality, to begin with. How many lines of code do you figure this comes out to?

Why would there be a "reappraisal" of acoustic live music? Because AI. You need more than a brain to get your touch together on a stringed instrument; you need all kinds of physical apparatus to be a wind player. Do you actually entertain a fantasy of a robot doing this any time soon? The feedback from touch on an instrument is extremely complex. This is supposedly going to be in the realm of digital binary code? When? How? What of the problems here are you currently studying?
Show your work.

Pure fantasy. Again, why would one want this to be true? This strikes me as the babble of someone that believes tech is already replacing real action, the *Producah* mentality of laziness and weak affect and an impoverished agency, to be perfectly honest.

I appreciate your reaction. And I appreciate your concern, at least that’s how your post appears to me. The fierceness of it gives me the impression you feel quite uncomfortable of AI possibly having such an impact. I may be wrong though.

And why do you want this to be true?

I do not.

Again, why would one want this to be true?

Again, I do not. What I consider likely doesn’t equal what I prefer.

I’m fascinated by the advances of AI yet also worried. The worry is not so much about how music is created.
My worries are more about other areas like deepfakes in sound, image and text in general, which undermines trust in media at the same time as it can fuel polarisation. Or AI as a tool to completely dissole privacy.

I compose music and I’ve been reading and thinking a lot about AI for many many years.

I don’t underestimate the many qualities needed to create novel, touching and culturally relevant music. It’s far from using a few simple algorithms.

On the other side I also don’t underestimate the possibilities and rapid pace of advances in the field of AI. It WILL totally change our world.

Note I’m not talking about robots playing physical instrument although that will advance as well. I’m talking about neural network systems which are becoming more and more like black boxes where’some magic happens’without us exactly knowing how. Magic as in ‘very impressive feats’ like beating the best human Go players or Dota players, by just trial and error and learn from and improve upon itself.

The limits of AI are mostly defined by the limits of hardware. Think of the end of Moore’s law for instance as a serious limiting factor. I’m not totally sure if hardware can be advanced enough to support the scale and complexity of neural nets mimicking human brain style power. I guess it will, as long as if humanity survives long enough.

So, without serious hardware limitations there are no limits to AI. It will continue to develop into unimaginable complexities and power. There are many fascinating developments going on, including evolutionary algorithms, generative adversarial networks and AI creating new kinds of AI.

Things we call intuition can partly be compared to the output of very abstract layers of deep neural networks. I say partly because it’s not the whole picture of course. The brain is still much more complex and currently there’s nothing in AI which is anything close to emotions. Yet very abstract concepts are no longer limited to humans.

And creativity? We need humans to define something as creative or not creative, which is a subjective thing. AI-machines cannot judge creativity like humans can. AI can rearrange and combine stuff in all kinds of ways, starting from abstract layers. But they mostly need human awareness and knowledge to find out if some totally novel output is also _meaningfull_ output. Meaningful to us humans that is.

Meaningful creativity in a more concretely defined sense has already happened in many ways. For instance AI which autonomously designs scientific experiments. That’s goal oriented creativity, whereas humans judge creative outcomes more in an emotional way. Our emotions tell us “now this is something creative!” even when it’s not.

When music touches us and we perceive elements which are new to us (because we feel we haven’t heard them before) we may call the music a creative feat. But is creativity limited to emotional intend? Is purely machine-produced and emotion-evoking output not creative because the machine doesn’t have emotions itself? That’s a philosophical question.

My estimate is we will use AI as tools, as ever advancing tools, up to the point where we don’t have to do much more than judging the AI’s output.

But what if AI can actively train itselve to recognize what triggers human emotions? It doesn’t have to feel emotions itself in order to ‘know’ what triggers emotions to us. The internet is a continuous giant source of triggers and reactions from which much can be learned. Also AI is not limited to a single box. AI can learn from and colaborate with (human and non-human) peers in networked supersystems.

We don’t need totally human-like AI to give rise to genuine, emotion-evoking output. Just as we don’t need totally human-like AI to cause us a lot of troubles in the world.

So, to conclude, where will DAW’s be in 10 years time? Well, I don’t expect DAW’s all to evolve into fully automatic generative music systems. Many will incorporate AI, that’s for sure.
I expect more from tech giants like Google, Apple or Baidu to offer very smart apps and ecosystems to explore the limits of what machines and the internet can do, music wise.

[telecode]

Again, I do not. What I consider likely doesn’t equal what I prefer.

I’m fascinated by the advances of AI yet also worried. The worry is not so much about how music is created.
My worries are more about other areas like deepfakes in sound, image and text in general, which undermines trust in media at the same time as it can fuel polarisation. Or AI as a tool to completely dissole privacy.

I compose music and I’ve been reading and thinking a lot about AI for many many years.

I don’t underestimate the many qualities needed to create novel, touching and culturally relevant music. It’s far from using a few simple algorithms.

On the other side I also don’t underestimate the possibilities and rapid pace of advances in the field of AI. It WILL totally change our world.

Note I’m not talking about robots playing physical instrument although that will advance as well. I’m talking about neural network systems which are becoming more and more like black boxes where’some magic happens’without us exactly knowing how. Magic as in ‘very impressive feats’ like beating the best human Go players or Dota players, by just trial and error and learn from and improve upon itself.

The limits of AI are mostly defined by the limits of hardware. Think of the end of Moore’s law for instance as a serious limiting factor. I’m not totally sure if hardware can be advanced enough to support the scale and complexity of neural nets mimicking human brain style power. I guess it will, as long as if humanity survives long enough.

So, without serious hardware limitations there are no limits to AI. It will continue to develop into unimaginable complexities and power. There are many fascinating developments going on, including evolutionary algorithms, generative adversarial networks and AI creating new kinds of AI.

Things we call intuition can partly be compared to the output of very abstract layers of deep neural networks. I say partly because it’s not the whole picture of course. The brain is still much more complex and currently there’s nothing in AI which is anything close to emotions. Yet very abstract concepts are no longer limited to humans.

And creativity? We need humans to define something as creative or not creative, which is a subjective thing. AI-machines cannot judge creativity like humans can. AI can rearrange and combine stuff in all kinds of ways, starting from abstract layers. But they mostly need human awareness and knowledge to find out if some totally novel output is also _meaningfull_ output. Meaningful to us humans that is.

Meaningful creativity in a more concretely defined sense has already happened in many ways. For instance AI which autonomously designs scientific experiments. That’s goal oriented creativity, whereas humans judge creative outcomes more in an emotional way. Our emotions tell us “now this is something creative!” even when it’s not.

When music touches us and we perceive elements which are new to us (because we feel we haven’t heard them before) we may call the music a creative feat. But is creativity limited to emotional intend? Is purely machine-produced and emotion-evoking output not creative because the machine doesn’t have emotions itself? That’s a philosophical question.

My estimate is we will use AI as tools, as ever advancing tools, up to the point where we don’t have to do much more than judging the AI’s output.

But what if AI can actively train itselve to recognize what triggers human emotions? It doesn’t have to feel emotions itself in order to ‘know’ what triggers emotions to us. The internet is a continuous giant source of triggers and reactions from which much can be learned. Also AI is not limited to a single box. AI can learn from and colaborate with (human and non-human) peers in networked supersystems.

We don’t need totally human-like AI to give rise to genuine, emotion-evoking output. Just as we don’t need totally human-like AI to cause us a lot of troubles in the world.

So, to conclude, where will DAW’s be in 10 years time? Well, I don’t expect DAW’s all to evolve into fully automatic generative music systems. Many will incorporate AI, that’s for sure.
I expect more from tech giants like Google, Apple or Baidu to offer very smart apps and ecosystems to explore the limits of what machines and the internet can do, music wise.

You are wrong. The limits of AI are not defined by hardware limits. They are defined by economics. You also don't want to confuse AI with Robotics. They are not the same thing. A very very basic explanation of AI is about finding patterns in a controlled environment quickly and efficiently and doing it better and more cheaply than a human can. Its all mathematics and algorithms and being able to copy and replicate and perfect it easily from place to place.

There are patterns in music. Music is a controlled environment. Mathematics and music are quite close to each other, hence why computing and music creation kind of fit well together. Try making a fancy clay pot with a computer. Good luck with that. But, you sure can make a very convincing orchestral score soundtrack with a computer and no human players involved. People do it all the time. Music listening and computing and AI fit perfectly with each other. Its a controlled environment and its all about counting clicks and pattens in wave files -- hence the astronomical success of Spotify like companies.

However, the argument that creativity is a controlled environment that can be measured is not true. An artist who strives to create a creative work which will break new creative ground in the creative medium that artist operates in is not going to be easy or cheap to create into an algorithm which will be cheap to reproduce in all and any country. How exactly are you going to create an AI that is going to create ground breaking creativity in China and then just copy that code over to the USA or Germany and have it replicate the effect?

Will there be a new genre of music that will take audiences by storm? Of course there will. There always has been. But its going to come from a human who is responding to something that is happening in society around them -- not a computing device analyzing patterns it has access to.

As far as AI and musicians. I think the only ones that need to worry are the ones that are creating predictable repetitive types of music that does not attempt to challenge the listener in any way.

[reggie1979]

so you're saying it will be usurped by an even newer technology and will be viewed with the same reverence that some put on analogue equipment today?
interesting...

I didn't say anything. It's a picture to fill in the blanks, you're assessment is as good as any though.

Oddly, I got more done on a similar machine than I do today

[Timfonie]

You are wrong. The limits of AI are not defined by hardware limits. They are defined by economics.

We are both right. It's not a matter of only physical limits or only financial limits.
Nearing the limits of miniaturisation makes every possible little step forward very difficult.
It's not like when ASML and TSMC suddenly would have 10 times the money that they will jump to 1 nanometer chip structures quickly. Moore's law is mostly about fundamental physical limitations. The financial question is: will it be worth to try and squeeze just a little more out of it, by investing billions and billions of dollars/euro's ? Will that result in economically viable products?

On the other side, one might think of building ever larger data centers when further miniaturisation is no longer an option. Although there are physical limits here as well, the limits are more financial indeed.

You also don't want to confuse AI with Robotics. They are not the same thing.

That must be a reaction to jancivil as he wrote about robotics. I was only talking about the chips and systems in which deep learning can be implemented. So no embodied cognition here.

Try making a fancy clay pot with a computer. Good luck with that.

Try attaching a 3D printer And a camera for feedback.

However, the argument that creativity is a controlled environment that can be measured is not true. An artist who strives to create a creative work which will break new creative ground in the creative medium that artist operates in is not going to be easy or cheap to create into an algorithm which will be cheap to reproduce in all and any country. How exactly are you going to create an AI that is going to create ground breaking creativity in China and then just copy that code over to the USA or Germany and have it replicate the effect?

There are all kinds of creativity, including creative new methods to solve maths problem, or to find solutions for treating a disease. These are relatively culture independend. Creativity in the arts however, is indeed much more culture dependend. That will make it less obvious to easily gain ground in other parts of the world.

[telecode]

There are all kinds of creativity, including creative new methods to solve maths problem, or to find solutions for treating a disease. These are relatively culture independend. Creativity in the arts however, is indeed much more culture dependend. That will make it less obvious to easily gain ground in other parts of the world.

I don't think you get the basic of the economic drivers of AI. The whole thing rests on finding repeatable processes and being able to make an algorithm that can be easily and cheaply replicated and perfected anywhere. So your disease example. Making an algorithm that will go through all the medical inter connected medical research texts and finding symptoms to skin disease X. Once its created, it doesn't matter whether that algorithm runs on a expensive silicon valley datacenter or some cheap Datacenter in the Phillipines. It can be used to displace medical professionals in any country because they are basically doing the same thing -- looking at a thing on your skin and making a determination what the medical assessment to it should be. It doesn't matter if the person is in EU, NA or Asia. That algorithm can be created and copied and perfected and transported easily and displace any person that does that repeatable process.

For you to make an algorithm that will make a cultural impact on say South Korea and K-pop genre. It will cost you a crap load of resources to make that. And you can't move it around and export it easily. It only works in South Korea more or less as a cultural AI type initiative. Your K-pop algorithm is useless in Toronto with the majority R&B/Hi-hop Drake and Beyonce crowds. Your K-pop algorithm is useless in raver clubs in EU dance clubs where everyone is booging away to Euro dance music. Your K-pop algorithm is useless in Finnish Death Metal club circuit. You need to make a different one for every time you want to make something that works culturally and socially in that region. Spotify on the other hand don't have your problem. They just have to deal with tracking clicks, plays, playlists, and managing who gets paid what in which region based on what is played -- so Spotify's huge investments in AI work in any region they operate in. Those are the underlying economic issues around it. Do you get the issues?

[perfumer]

Do you get the issues?

Big data is as important as the AI that sifts through it (and its complexity). The AI is going to work on a statistical basis = choose 75% established rules, and 25% variations (of course that's very roughly said). Feed all the music of Bach in a machine and make it able to get out the fine details, sort them in a structure, then make statistically based choices on what to include in the new fugue, and chances are you'll end up with something very similar to the works of Bach. Same for K-pop, same for any style that's based on relatively strict rules and structure. A friend of mine who is very smart became a data architect, and was previously working on semantic search engines - basically it's the same process, whether it's news articles about football, jihadist networks, or techno music. Find what occurs and how often, then find what's related to what, then make a model and fine tune it.

But that's valid fo cliche-based music, not for experimentation and highly creative but reasoned (not arbitrary) choices. An AI may not be able to write a Zappa song, or an Umberto Eco page of literature, because this requires what machines don't have - real intelligence. (That is, being aware of the playing field, being alive, fit and in control of your skills, then playing and thinking, thinking and playing.)

So TL;DR - it's not just about the economy, there are more serious issues than that.