DeusExMaschine

I got a bit of a backlash for using AI tools on my latest Bloghouse EP.

So before everyone turns this into a black-or-white debate, look at this: this is from the manual that came with the MPC3000.

Roger Linn was already saying that sequencers had become instruments in their own right — the piano or violin of their time — and that artists shouldn’t only be credited as “programmed by,” but as “MPC3000 played by.”

I started making music on an Atari ST and an Akai S1000, and I’ve seen the same panic with every generation of tools: sampling, sequencing, plugins, stem separation in Ableton, mastering assistants like Ozone, and now Suno and AI tools.

I’m not ignoring the copyright problem either. My own tracks appear in one of the public AI music training datasets. So yes: consent matters, attribution matters, compensation matters.

But both things can be true at the same time: the way these systems were trained can be a serious problem, and the tools themselves can still become instruments.

They don’t magically make great music. Taste, choices, references, obsession and intent still matter.

So yes, I’m taking a side: AI tools are not the end of music. They’re another evolution in how music gets made.

Who was there?