And history is littered with artists who outsourced and delegated to assistants who actually did the grunt work. Rubens and Van Dyck had huge studios where assistants did the majority of the work. Van Dyck who was renowned for his portraiture often only stepped in at the end to paint the important critical bits like the hands and the face. Thanks, Maestro.

Warhol is another. He promoted the idea that authorship could reside solely in concept and direction and treated himself as director/producer. His name signified the conceptual decision-making even when assistant labourers often did the physical work. Many scholars (and indeed the Warhol Foundation) accept and authenticate Warhol as the author when he conceived, directed, and approved the work perhaps without laying a hand on it.

Or Mark Kostabi...in Kostabi World, a production line of assistants came up with the concepts and naming of the pieces, while others executed the actual paintings. The only input he had was signing the finished works, which enabled him to churn out up to 1000 pieces of art a year. He even knowingly and unashamedly refers to himself as a Con Artist.

Art? We shit it out!

 

That's an oversimlification in my opinion. There are and there is going to be many ways of usign this tech. Of course Suno and other close sources companies and their scaming philosophy can suck my ass, but I there is a lot of ways that you can be using the open source models that don't substract any joy or fun.

I would never ask scaler 3 or any other software, AI or not, for a chord progression or a melody cause what's the fucking point then. The melody and the chords are intrisec to how I feel in the moment. But, for example, spending hours programing strings on any kontakt library, switching articulations, adjusting dynamics, etc etc just to get the "no quite there" repetitive sound those libraries always deliver no matter what vs doing a Audio2Audio, with my melody to obtain a "not quite there sound" anyway but with a richness and subtleties in the articulations that no library is going to be able to deliver is for me a lot more fun. Is like having an sintetic orchesta that you ask takes from your composition, in the box. And this type of artistic direction techniques are going to keep evolving cause the tech have a plenty of room for it.

 

I´ll step in in the op defense here, i know it is a bad rep to defend AI...i´m far for the AI, i actually cant bare watching Youtube videos like car repairs, upgrades because the backgr.....anyway.

What he talks about considering if you´re pretty skilled, you can actually spin the idea very quickly and if you have it in your mind like how Hans says he does it in his head, you can quickly flush the idea without loosing yourself. I know this separates the levels of the skill but he obviously can do stuff and retain the idea while it is coming...

How Arctic says aren't we supposed to have fun... it kinda starts being a line of separation here based on the talent and the experience where the enjoyment or the inspiration last for a fraction and it can get "caught" versus having inspiration and loosing it thru the process of "enjoyment". hobby vs pro

that does mean bad untalented dudes can make crap incredibly quickly too

 

I'm not going to suggest some hardcoded date, but Steinberg's invention of VST (virtual studo technology) was built around the idea of people being able to use virtual versions of stuff they didn't own, didn't want to maintain and keep, couldn't afford, could't learn, couldn't play since it's launch.

This entire subject is about identity protection and ego defense. People here only began to care about AI when it started to show improvement at making music other than Electronic music. The same people saying stuff is "AI slop" are the same people who would complain about House,Techno, Trance, Djs, Rappers, Hiphop. Anything without guitars and a drummer. It was easy to not care when it was about music genres your ego allowed you to dismiss out of hand as "not music"

If the sole measure of "making music" is that you played something with your grubby little mitts, go tell someone like DeadMau5 or Skrillex they aren't musicians for using mostly computers to write. Deadmau5 does not play a single thing when sequencing, even though he has more synths than anyone would ever need. They are further back in the studio and you have to look past the wall-sized modular that cost more than many of our "serious professional musicians" complaining here have made in their entire lifetimes of making music. Don't believe me? Watch his Masterclass. I'll be awaiting links to yours.

A musician makes intentional musical decisions. The tool is irrelevant. Just like you don't get to tell someone they can't pay someone to do their mixing, or what plugins they get to use; trying to assert your authority over what AI others use would be laughable, if it wasn't so self-serving and pathetic. Your real problem with it is that no-one asked you for your permission. You should have realized long before this that they never needed it in the first place.

 

Last edited: Feb 5, 2026 at 5:19 PM

@KORG3R & @Demloc i think these are both solid points. skill, experience and wanting to do something definitely play a large part in why people us A.I.

like OP mentioned presets, i dont use them in songs because i spent a lot of time learning synthesis and practicing it. like @Demloc said, wanting the perfect sound to me and not just kinda there, pushed me to learn that instead of searching through presets to just endlessly go back and tweak every time i hear the song

where i agree with @ArticStorm in this context is that there is a fun and sense of accomplishing getting the sound right that i love, i dont use any sample libraries but im willing to bet other feel that way like mentioned with the strings. these points you both mentioned is what i see in people who use A.I. creatively, its not a love there (maybe a love for new tech lol) like i mentioned in an earlier post. seems largely like people not wanting to do something or not having the ability to do it

 

Some people like to do puzzles. Sometimes they look at the box.

 

No need to defend me! It's an open discussion! I wanted all to chip in and share their views. I didn't say I'm pro or against AI. I asked your take to learn about real musician's take on the new direction. It's interesting.

 

Schnabel too, for years now churning out product at his pink factory Palazzo Chupi (in Spanish chupar means "to suck", Chupi is Italian but it is still "sucks" to me, lol). I was totally bummed out when I found out of his of fabrication and staff doing the work. It dimmed my perception of his paintings and I wouldn't go into a Gagosian gallery for years...until I broke my fast a coupla years ago to see a Nan Goldin piece here in Basel...just had to.

 

“You have to know the past to understand the present.”
― Carl Sagan

Synthesizer History Timeline

Spoiler: The Synth List – Formating: Year – Manufacturer – Model – First at what..?

The Synth List – Formating: Year – Manufacturer – Model – First at what..?

• 1837 – C.G. Page (Salem. Mass) – first to produce electronically generated sound (not necessarily associated with a musical instrument). After inventing the Volta in 1800 (an early battery), in 1837 Page was doing experiments with coils and realized when certain coils were attached to a batter they omitted a ringing sound. While he initially thought the ring came from the electrical current was interrupted (battery disconnected), what was actually taking place was the induction through the coils was causing them to vibrate.
• 1885 – Person and Ernst Lorenz – Elektrisches Musikinstrument – the 1st musical instrument designed to produce electrically generated sound. It used electronic vibrations to drive an electromagnet that were connected to resonating boards, which translated these vibrations to sound.
• 1897 – Taddaeus Cahill – Telharmonium – electromechanical instrument.
• X-1905 Helmholtz Resonator (Human Voice „Emulation“)
• 1936 – Oskar Sala – Mixturtrautonium – first electronic instruments using Subharmonic synthesis (not a „synth“)
• 1938-42 Oscs, Filters, Envs, LFO (Mech) – no CV – Hammond Novachord
  
  
• 1939 – Homer Dudley invents the Parallel Bandpass Vocoder (VODER) – A manually key operated speech synthesizer
• 1940 – Homer Dudley invents the The Voder speech synthesizer – A device which used the human voice and an artificial voice to produce a composite. Both were researched as a way to transmit speech over copper wires (id est, telephone lines)
• 1948 – Hugh LeCaine – Electronic Sackbut – First voltage-controlled synthesizer
• 1948 – Dr. Raymond Scott – Wall of Sound – First polyphonic Sequencing Workstation (electromechanical) and the Electronum – first sequencer.
• 1950 – CSIR – Mk 1 – The first known use of a digital computer for the purpose playing music
• 1953 Harald Bode – it had: OSCs, Filter – Ringmod – Reverb – finally named: Melochord
  
  
• 1953 RCA-Synthesizer: Herbart Belar, Harry Olson – RCA Synth Mk1, 2 (rec. auf Vinyl) // /1955 mk1 / 1957 mk2 made mk2 used tape instead of vinyl disc
  
  
• 1956 – Louie and Bebe Barron – Produced the first all-electronic musical score for a major motion picture – MGM’s ‚Forbidden Planet‘
• 1957 – Max V. Mathews at Bell Labs – MUSIC – the first digital synthesizer. Technically, it was a computer program, though it set the stage for every digital synthesizer that proceeded it.
• 1961 – Harald Bode his studio modular lab was- very much like the first synth made by HB except oscillators (tapes were used) – but it had filters etc.
• 1963-65 Prototypes of Moogs Modular Synth (first modules shown to the first 2 clients in 64/65) – the official time should be given as 1964 (shown to people) – but finally it was shown to the public with more modules in 1965
• 1965 (not sure) – Buchla – model 100 modular – 1st „modern“ modular synth
• 1965-67 – Moog – Moog modular synthesizer I, II & III – 1st commercial modular synth. final official version – 1st delivery!
• 1969 – EMS – Synthi VCS-3 – first non-modular mini-synth
• 1970 – Moog – Minimoog – 1st Mono Synth with keys (non-modular)
• 1971 – Tonus/ARP – Soloist – 1st preset mono synth
• 1971 – John Chowning – developed FM synthesis using the MUSIC-IV language (source), a direct descendent of Mathew’s MUSIC program. FM synthesis was later licensed by Yamaha, and used in popular synths such as the DX7.
• 1971 – Buchla – 500 – micro-controlled polyphonic analogue in 1971, it was also programmable as you could save patches to floppy.
• 1972 – Triadex Muse – first of many horrible sounding digital synth/seq workstation thingies
• 1973 – Coupland Digital Music Synthesizer – First Digital (Triadex beat it?) Update via Peter Grenader: „No time to read through all these posts to see if it’s come up yet, but the Coupland was vaporwear…it never existed. I met Mark Vail, who’s now a friend, by writing him a letter informing him that his story about the Coupland in his Vintage Synthesizers book (GREAT book) which mentioned it’s only recorded showing was at the AES show in LA in 1978 was a farce. I was there – at their booth and their suite in the Hilton where the instrument was said to be. I was there on the first day, I was there on the last day. The only thing they had was a small model – about six inches across, sitting on a table. The booth was amazing – this radial orb multiple people could sit in, with a cover that came over each person which played what I remembered was a very impressive demo which swirled around four speakers inside the box. I, and everyone else, were blown away. They kept saying…’it will be here tomorrow, it’ll be here tomorrow’…so I showed up the last day just to see it, figuring by the then it would have arrived…it didn’t. I did see the frst Synclavier at that show however. Their suite was across the hall from the Coupland folk. That completely kicked the crap out of everything else shown that year.“
• 1973 – NED – Synclavier – first digital synth
• 1974 – Roland – SH3a – first commercial additive synth
• 1974 – RMI – Harmonic Synthesizer – first commercial additive synth
• 1976 – Yamaha – CS80 – first synth with poly aftertouch = polypressure
• 1976 – PPG – PPG 1003 sonic carrier – 1st programmable mono/duo synth (this, along with the model 1020, might have been the 1st synths to use DCOs as well)
• 1977 (late) – Oberheim – OB1 – 1st commercial programmable mono synth
• 1978 (late) – PPG – Wavecomputer 360 – 1st wavetable synth
• 1978 – Sequential Circuits – microprocessor control the SCI Prophet10 (briefly) and the P5 — again based on existing E-mu tech stuff
• 1979 – NED – Synclavier – First FM
• 1979 – Fairlight CMI – First Sampler, First Workstation
• 1982 – Sequential Circuits – Prophet600 / First Midi Synthesizer (though some argue the Prophet 5 rev 3.2 is pre-MIDI MIDI)
• 1983 – Yamaha – DX7 – Digital takes over, FM goes mainstream
• 1983 – OSC – OSCar – First real-time additive with analog filters
• 1984 – Sequential Circuits – SixTrak – first multitimbral
• 1985 – Casio – CZ101 – First battery-powered all digital mini-synth
• 1989 – Emu Systems – Proteus – First dedicated ROMpler
• 1994 – Yamaha – VL1 – first physical modelling synth
• 1995 – Clavia – Nord Lead – 1st Virtual Analog (VA)
• 1996 – Rubberduck – still not the first softsynth but came before Seer Systems Reality.
• 1996 – Steinberg – VST – Ok not a synth but enabled a lot to be written as plug-ins and used simultaneously
• 1997 – Seer Systems – Reality – First Modular Soft Synth
• 2912 – KalQuestoTron – the first genetically engineered synth. Each cell is an oscillator, filter, and neural sequencer. Can be delivered via injection to always play ‚hold music‘ in your head.

From „https://www.sequencer.de/synth/index.php/Synthesizer_History_Timeline“

 

You industry pros really gonna get into a dick swinging contest in a thread about bedroom producers using AI?

Fucking LOL!

 

Now I'm curious about their work!

 

Ok, time to close the thread?

 

Seems it kicked off and I missed the fun.

Back to the topic.
@euxyh103 so what did you decide, are you going to try any of these platforms? I know that Waves, Output, Soundverse DNA and more, all have products they claim are ethically trained on proprietary and licenced datasets, so maybe those are preferable to Suno/Udio.

What I would find interesting is if I could train a model on my own my own catalogue and unfinished projects, which amount to thousands of tracks over decades now, and see what pops out. Colab with a digital version of myself. I suppose the next step in this will be direct brain-to-machine interfaces, then your thoughts are realised in real time. Wild times we live in.

 

I think I'm going to stick with getting some ideas for chords and progression for parts that I'm stuck in via tools like Midiagent, Scaler, chord pads, and Gemini. For vocals I'm using Suno right now just to test tracks before commiting to recording.

 

The thing is, if you didn't go the sister site route: You bought the samples, you bought the mastering sofware with the presets, iztope wanted to sell ozone to you and meant you to use it's presets and so on.
The problem with AI: It steals and sells stolen goods back to artists. In fact it is like selling cracks.

 

Splice just launched beta of Splice Sounds plugin. It uses "ethical" AI to give you variations of their samples. You can upload some audio from your session, it offers recommendations of sounds that will work in your session, then gives you the option to create the AI variations.

 

A couple of years ago i discovered SoundRaw, it's like Suno or Udio, more limited but also easier to access since the beginning. SoundRaw lets you generate songs based on a lot of parameters, but they're all 100% AI, you can change the BPM, the intensity, the key, the duration, etc. It's insanely powerful in theory, but when you go to the results it clearly falls short.

With that i mean, AI benefits from those who don't think outside of the box, if you understand what it does then you understand that it's, at least nowadays, a second opinion, an alternative, not a dependency. Every major AI company tries to sell you the idea that this is a revolution merely by the extension of the word "intelligence", it is a concept so ambiguous and powerful that can make the masses collapse by a sort of "paralysis by analysis" effect, there's so much to do with something like this.

That's always because it has something to steal from, someone who makes it learn.

Why do you go to school? Why did your parents insist so much to the point where they make space on their routine to wake you up and get you to that "horrible" and "boring" place? Because education is important, not to increase an IQ value on a test, but to make you understand what's around you, to make you learn and interpret information in many ways. AI doesn't have fathers, it has instructions, it reads logic and it's insanely powerful in doing so, but the reason that you push forward knowledge as a human beign is not exclusively attached to intelligence, is also emotion. AI can fool you into feel emotions, but you're not listening or reading to someone's internal process, you're basically having a real time simulated essay on the interpretation a machine has about what we do.

In music, it's the exact same. "Emotional" is not a set of chords that sound like the basis for a LinkedIn corpo propaganda, emotional could be confusion, could be lyrics that mismatch the intensity or the tone of the instrumental, it can be context, references, subtle implicit details that culture provides, it has a target to reach, artists use resources not just as a trasmisible medium but also an abstract, imaginative one.

AI understands tasks, not emotional depth or complexity. Your tool can do a bunch of crazy, useful and impeccable sounding craft, at least compared to modern amateur production, but if that is all that we have to surpass as a society, then Tik Tok artists would not exist, people who don't know how to compose music will be segregated. The beautiful side of music nowadays is that everyone can create based on a thought, an emotion, an impulsive idea or a thoughtful and integrally constructed project, AI is borrowing the creativity of thousands to produce a sterile, summary-of-the-everything result, getting the generics and squishing them to the utmost "functional" basic.

So, if you create a tool so powerful that it can replace humans, then maybe don't make it work based on what humans create. Someday, if we're all replaced, who will train AI? Musicians will reject, eventually, sooner or later, to provide something that's used to replace them, AI cannot simply be human when it chooses to, even with AGI. Using it to make a draft, to try to frame or picture ideas, scenarios, sounds, that's excellent, but living and earning money based on the results of a super-intelligent copycat machine that lives of ripping off the others? That's lowlife behavior, peak one i must say