Just the other day I stumbled upon a YouTube channel presenting an AI-assisted tool by the means of which everyone can build their own music VST plugins - without the slightest need for coding/programming skills.
And last week a guy who used AI to reverse-engineer the entire Adobe Creative Suite (Photoshop, Lightroom etc.) stirred quite the resonance in social media.
Even if we are early in the process of becoming able to “make” our own DAWs and Plugins, it is quite obvious that the future of how software, programs, including DAWs and Plugins are being created, curated, adopted, maintained and developed is fundamentally changing.
So, is this the end for professional software like Cubase?
Personally, I don’t think so. I believe that AI will not “blindly” replace good programmers and software, but those companies will survive which can make the best use of AI.
So, what would be your thoughts on how could or should Cubase embrace AI in order to stay ahead of the game and not becoming a “relict of the past”.
I think that most of us agree that Cubase needs to stay a platform that allows the user to take maximum control over every step of the human (!) music poroduction process.
That beind said: where do you see potential to add AI-driven capabilities that do NOT get in the way of your music creation but still represent a step towards being helpful to make your vision come true.
I think that this is an exrtremely imprtant topic for all avid Cubase users as we all want and wish for our software of choice to stay competitive and future-proof.
Not everything labeled as “AI” is Suno or like Suno.
Suno and similar products do NOT support the creative process or the musician’s unique artisitic expression. It just replaces all of that. So, Suno ist sort of an “anti-creativity” and ultimately “anti-Cubase” concept that no serious musician consideres to use as part of their own creative process.
The state of AI is rapidly and exponentionally, evolving.
Discussions on AI that have been had mere months or weeks ago, may already be obsolete, given almost daily new disruptive AI-achievements. For example: Had someone stated half a year ago, that every “hockey mom” could reverse engineer Cubase by pressing a few buttons and create their own, maybe even more capable “Cubase” clone, this idea would have been dismissed as urealistic, laughable or “sci-fi” from a distant future.
Last week, a random guy resverse engineered the entide Adobe Creative suite. Sure, with some flaws, but it is a matter of a couple of months, if not weeks, until anyone may be able to let a llttle army of AI agents build their own “Cubase 30.0” with features Steinberg may not even thought to implement in the upcoming years.
So, the point of this discussion is not to regurgitate old arguments based on old data, but to draw conclusions from the most current state in the field of AI.
As I have previously stated: Not the AI itself will prevail but those who can make best use of it and harness it to their specific use cases.
It’s definetly not the end of professional DAWs! 1. “AI” is ONLY a marketing term. 2. You can’t programme such a complex system following the motto “fire and forget”. Programmers using programming “building blocks” since many years. But humen eyes and brains are still needed. Recently, an IT specialist said that we are a million miles away from AI - real AI. More and more companies stop using “AI” because there is no measurement of if it’s beneficial or if it’s not. They don’t have a monetary befit from “AI”. That’s why the tech companies started sweating because their concept doesn’t work as planned. Recently, an official test from an university said that the most well-knows “AI” systems could solve just around 2 percent correctly of the tasks they gave them. And I don’t want to use an “AI” programmed DAW. It would cause a lot of problems.
Thank you for moving this thread to a subforum that is more suited to this topic.
Personally, I think this discussion is hugely relevant, as just a few weks ago i got invited by Steinberg to participate in a survey on what features I would like to see in Cubase in the future and what role AI tools should play.
In that survey I answered something like: make Cubase as stable and as fast as possible and keep all that AI stuff away.
Why? Because I assumed that AI = generative AI like Suno, which I don’t consider to be a tool for music production but rather an “anti-music production” tool.
It’s been literally the past few weeks and days that I have come to realize that AI is so much more than just generative AI and that it might hugely support and improve tedious workflows, freeing up precious resources and time that are better spent in the creative process itself.
It is not anti-production, it can enhance production processes, which is most likely where ALL DAWs are headed, however by definition Artificial Intelligence “itself”, does not create, media, it generates binary and textual information as a result of learning, hence the term.
AI will never enhance production, it just replaces the human input. You can learn how to read and write music, you can learn how to play an instrument, you can learn how to use a compressor, how to use notation, how cut a sample at the zero crossing, etc. AI is enablement for a lazy generation!
With no electricity I can sit down at a Piano and write music, I can grab one of my guitars and write something, I can sit behind a drumkit and come up with something, I can use staff paper and write something for an orchestral setting…I don’t need AI to do anything. I use a Daw to hit record for audio and played Midi and I move that occasional flubbed Midi note. I write, record, mix…what else is there?
With “C-” computer skill and (“D-”) music theory skill I need an authority like Greg Ondo on the phone continuously to answer my future (Cubase 16 Pro and DGX 680) questions and to make suggestions, alternatively a sophisticated “Cubase AI Helper” is urgently needed bring a flood of new customers into Cubase. For me personally, it’s probably no Cubase AI Helper, no sale.
I would like to see Cubase work with future 8K 43"-55" PC monitors, AR and VR glasses, also all Steinberg/Yamaha software and hardware MIDI 2.0 capable, for future workflow products.
In my opinion, generative AI such as Suno is directly antagonistic to the creative process, because it replaces the human artist in the very creation process.
Or to put it in a nutshell: If AI writes the music for you, where is the fun and where is my unique expression of myself? It’s like asking ChatGPT to write a 1200 word essay on a sppecific topic and in a specific style.
Now, there is a counter-argument to that reasoning, saying that classical arranger-keyboards or arrangement software such as Band in a Box, are essentially doing the same, albeit much less sophisticated than Suno can do. They may not write the “entire” music for you, but arrangements, even soloist and melodic lines.
So yes, if this is the goal, to sit there, press a button and let the machine do the music for you, then so be it - this is music production, and your assessment of that matter was completely correct. However, the more the machines take away from this process, the less this music is “mine”.
In general, I believe that a vast amount of Cubase /Dorico user base is rather interested in having a larger own contribution to the music production than just pressing one button and let the machine do the rest.
So, we are somewhere in a grey field with AI (or algorithms in general) are just “tools” that to a varying degree either support our own creative proocess or take away from it.
Absolutely true. This perspective sees a DAW just as a recording tool without the need for anything “generative” being implemented in the process.
That being said: Even this mere recording process is not trivial.
Here is a case in point: Let’s assume, you are recording guitar, drums, bass and finally vocals - all played and sung by yourself, recorded sequentially one track after the other, and then putting all tracks together in the DAW to make for a wonderful song.
Especially in larger projeects, quantization can become fairly tedious. How about telling the DAW: please “adjust” the timing of all tracks accoording to a reference track and quantize them so the entire song still sounds “natural”.
I’ve said it before, and I’m going to say it again (Oh no! ):
AI is like Marmite.
You either love it, or you hate it.
And arguing the toss about it isn’t going to change anyone’s opinion about it.
I do not disagree, and my goal is to gain insights rather than to “change” anyone’s opinion.
Right now - and I mean just in this very moment as we speak - the music production environment technology is at the verge of a disruptive transformation.
And this is not being “click-baity”, as someone else in this thread assumed, but being genuinly alarmed that the DAW- and plugin- industry may be having their “Kodak”-moment right now: Kodak, a phenomally innovative and hugely successful company in the photography market who used to develop and sell a multitude of various analog cameras alongside manufacturing analog film, missed the technological revolution of digital photography (despite being the first company to ever develop a working digital camera prototype), and failed to adapt to the digital transformation. Now, Kodak is essentially gone, while other manufacturers like Leica or Nikon adapted very well and are thriving up to this day.
So, what happens now and the decisions companies are making right now, will pave the way for them to become the next “Kodak” or the next “Leica”.
And this is where the circle closes: Steinberg actively asking us customers if, what kind if and how much of AI we wish for in the future development of Cubase et al. is actively seeking feedback from their core community on this very matter.
And in order to make my own mind over these question, I need to discuss these things as I hugely benefit from other people’s perspectives.
The board is discussing whether pushing the number of posts about putting AI into Cubase from 1 every week to 2 was a good idea considering the end result is usually negative.