vladasyn
Saying, "We just basic DAW, if you want other tools- go with 3rd party" is not the way to move toward the future.

Actually...I think it is. The field of machine learning (which I think has far more potential for music than AI) and AI are complex, evolving, and specialized fields. Furthermore, the level of integration with a program is mostly scooping up data - the analysis happens outside the program, and then the conclusions "pour" back into the program the same way (but of course, in a much more involved way) as data from a hardware controller making changes in EQ, automation, etc.
 
If I was a machine learning developer who was interested in music, I would not design a DAW to go along with it. I'd feel that the DAW world is relatively mature, and I would just want to create compatible products rather than re-invent the wheel.
 
As soon as plug-ins were invented, DAWs went down the road of being modular, rather than self-contained. Machine learning is a perfect candidate for a modular approach.
 
I wrote an article for Pro Sound News, Pro Audio and Machine Learning - Ready for Prime Time? that you might find interesting.
 

This is nice article, thanks for sharing. Could you, please, explain this better?
 
"I analyze a mixed file and look for half-cycle peaks that exceed, for example, -3 dB below a reference. There can easily be 30 or 40 such peaks. I can then normalize each half cycle down to -3 dB, which allows raising the overall level by 3 dB without introducing artifacts like pumping, while maintaining the dynamics."
 
So you say that it is a job of AI or ML developer to make compatible product and you dont see any use of DAW company to actually find talented individuals or small companies and collaborate with them to inspire, motivate, encourage them to develop mutual products that would benefit DAW and offer support and umbrella to new developing company? (Also Cakewalk probably not in the position to offer any support right now), but otherwise how do you break ahead of competition? The AI/ML developer may not know what musicians need. They need to be shown and told, "Look, we need you to teach your machine to sort through best takes and build ideal vocal track" and offer them financial incentive. (I am mediocre vocalist and it takes forever to go through 10 takes of each phrase to pick the best one!). If Cakewalk could sort through vocal takes, I would buy that upgrade right away!  

vladasyn
There also apps that use artificial intelligence to fill song with drum tracks and even create melodies and chords.
 

You really want that? 
That's ridiculous. 

I would be happy if it fills song with drums for me. 

vladasyn
This is nice article, thanks for sharing. Could you, please, explain this better?
 
"I analyze a mixed file and look for half-cycle peaks that exceed, for example, -3 dB below a reference. There can easily be 30 or 40 such peaks. I can then normalize each half cycle down to -3 dB, which allows raising the overall level by 3 dB without introducing artifacts like pumping, while maintaining the dynamics."

 
Basically, it opens up headroom by reducing the absolute highest peaks. Then you can raise the file's overall level without exceeding the available headroom.
 

So you say that it is a job of AI or ML developer to make compatible product and you dont see any use of DAW company to actually find talented individuals or small companies and collaborate with them to inspire, motivate, encourage them to develop mutual products that would benefit DAW and offer support and umbrella to new developing company?

 
No, that's not what I said at all.
 
We already have a fine example of what I'm talking about: Melodyne. ARA came about because of PreSonus collaborating with Celemony, and Cakewalk was involved as well. I don't think either Cakewalk or PreSonus wanted to stop development of their flagship products to become experts at pitch correction.
 

But otherwise how do you break ahead of competition?

 
You can point to other programs that do pitch correction, but they won't create a tempo track like Cakewalk can because they're not ARA. If you want pitch correction that's integrated really well with a program, yet has the potential for future expansion and development, at this point your best options are CbB and Studio One. They're ahead of the competition in that respect not because they did the pitch correction, but because they signed on with a company that have incredible expertise in that field, and shared their expertise.

Anderton
 
Actually...I think it is. The field of machine learning (which I think has far more potential for music than AI) and AI are complex, evolving, and specialized fields. Furthermore, the level of integration with a program is mostly scooping up data - the analysis happens outside the program, and then the conclusions "pour" back into the program the same way (but of course, in a much more involved way) as data from a hardware controller making changes in EQ, automation, etc.
 
If I was a machine learning developer who was interested in music, I would not design a DAW to go along with it. I'd feel that the DAW world is relatively mature, and I would just want to create compatible products rather than re-invent the wheel.
 
As soon as plug-ins were invented, DAWs went down the road of being modular, rather than self-contained. Machine learning is a perfect candidate for a modular approach.
 
I wrote an article for Pro Sound News, Pro Audio and Machine Learning - Ready for Prime Time? that you might find interesting.

 
I don't know how many know that VocalSync was over a year of RND and used advanced machine learning techniques to for its detection algorithms. One of the developers who built the DSP is a PhD in machine learning. Detecting patterns in speech and music for alignment is a black art and only a few companies do this. SyncroArts is the market leader in that space.

My first post here was meant to be sarcastic, but I am not sure that is how it was taken by all. There really is a significant philosophical question regarding fully machine generated music which will probably never be addressed in practice. Like most such profound questions about the nature of humanity, it is likely that technology will just make philosophical insight largely irrelevant. From a consumer/publisher point of view, music is a commodity, and when machines can create hit songs more reliably, faster and cheaper than humans by themselves or using machine assistance like DAW's, the average listener and publisher is just going to go for the artificial stuff. There will undoubtedly be boutique human crafted compositions for a long time to come, just as there are artisanal pickled carrots and hand made furniture being sold in today's modern world of the future. But the stuff playing on the radio will mostly be factory built. People will think it is quaint that we once thought that music was a mode of communication person to person. The current method of music created by teams led by "producers" cobbling together beats and loops to mimic the sound of the last big hit is already showing the direction the industry is moving.
 
So my music hobby will be like every other human hobby, just a way of keeping busy exploring my own taste and technique, and the people on forums like his will be like civil war reenactors encouraging each other to live in an imaginary past. From time to time a robot monitoring our self indulgent efforts will find something it does not already do, improve it and start turning out much better stuff in our style for the masses. 

Anyone for tilting at windmills?

@noel, when's the replacement forum coming? thx

I was spear headed in this forum about 5 or 6 years ago for giving my opinion on samples and loop content.
At the time I believed people who used loops (4 and 8 bar loops you can buy in sample packs) were frauds.
Thinking they cant play an instrument and are faking it.
 
In the present I have come to realize these might be inspiring musicians trying to find a way in to making beats.
Times change, when I was younger if you used a drum machine on an album you were a loser.
Rather than be that guy who says " When I was kid we didnt".. or .. "Kids today have no idea".
I would rather embrace them and help them in any way possible and also be open to any new ideas.
 
There is still that shrill factor in me that despises computer based music with no "Real" instruments (such as guitars, bass, drums, exc)
But now a day these are the "real" instruments. 
 
So im all for Artificial intelligences. Not so much to write the songs for us But to assist in the process. 
Im also 100% in on having a duplicate AI husband to do my wife's chore list as well