How do you retain an element of surprise for your own work – are there technologies which are particularly useful in this regard?
Peter: Well, this is why it’s nice sometimes working in a duo – Benjamin and I intentionally sent each other track stems, in audio form so you couldn’t edit, and treated them as ‘found sound’ in Live and let the other mess them up even more. And you’re constantly capturing accidents.
Usually, the stuff you try to do intentionally comes out all wrong and you hate it. Maybe “happy accidents” isn’t even the right word. I think half the time, you’re trying to undermine your own intentions.
Benjamin: There are tons of practices you could use to get out of your habits and mix things up. For me it has always worked best to just use a tool that I haven’t used for a long time whenever I feel stuck and take it from there.
That is a nice aspect of working with Peter: we just surprise each other by deconstructing the others work and putting it back together, sometimes by sending over bits that we just recorded, sometimes by jamming in the studio.
It helps that both of us have the same ignorance towards following a certain music style and always prefer weirdness over perfection.
Production tools can already suggest compositional ideas on their own. How much of your music is based on concepts and ideas you had before entering the studio, how much of it is triggered by equipment, software and apps?
Peter: This is always the reason to play an instrument, back to when I played a piano as a toddler. You make a gesture and then discover this world of sound responding.
The piano has a reputation for being limited, but then you discover the bottom or top notes, where tuning disappears and the sound is complex. I guess on synths I’m often trying to do something similar – push it past where the sound is expected. I played Javanese and Balinese gamelan for some years; there, they believe that there’s a spirit in the instruments you have to appease. (That also means not being rude and stepping over the instrument – just like trying not to give too much of your crotch to fellow passengers in coach on an airplane.) I think of course electronic instruments or software have the same potential.
And speaking of being inhabited by spirits, now I know a lot of the people who designed the thing or wrote the code. So I’ve met the ghosts in the machine – and I do have some respect for their impact on what I do.
Have there been technologies which have profoundly changed or even questioned the way you make music?
Peter: I really loved discovering convolution and granular synthesis for the first time. This was a combination of seeing SoundHack by Tom Erbe and reading about the techniques in Electronic Musician – trying to remember, Larry the O maybe was writing that column.
The ability to smear and stretch time and have the computer produce dream-like sounds was just wonderful for me. I’m still not over it.
Benjamin: Changed definitely (too many to mention), questioned not really - for me making music is not tied to any technology.
To some, the advent of AI and ‘intelligent’ composing tools offers potential for machines to contribute to the creative process. Do you feel as though technology can develop a form of creativity itself? Is there possibly a sense of co-authorship between yourself and your tools?
Peter: I’ve spent a lot of time on machine learning, including co-hosting AI art and music labs and talking a lot to researchers. I think it’s important to differentiate that in very objective terms, the current battery of machine learning techniques is not about generating creativity.
Any indication people give that that’s the case is either anthropomorphizing the technology and mistaking the results, or a matter of marketing it as something it isn’t. Not to be blunt here; I think that position is backed up in terms of how data scientists describe what these techniques do. That being said, there’s still lots of potential in machine learning processes for developing how tools work and producing new sound results.
On one hand, you have the ability to create tools that sort of sample more logically, for instance, or are better at matching EQ. And on the other, you have some really unique possibilities in sound generation. I wound up playing a warehouse in St. Petersburg with grungy, generated sounds for instance, AI-made techno. A lot of other artists (like Herxorcismos) are making real dedicated practices out of it.
There’s always co-authorship between music and tools. We’re always building on tradition and absorbed music. We may look for a blank slate, but there’s never really a blank slate.
Benjamin: I think in its best moments AI can help to give you new ideas while in its worst moments it will make sure that music sounds more similar in the end and that you stop learning things and just pass them over to an algorithm. Making things easier isn’t always what you really want in order to explore your musical ideas.
I do like it a lot though, when an AI is used against its intendd use or just fails miserably in it. This is when I think AI is at its best as a creative tool.
What tools/instruments do you feel could have a deeper impact on creativity but need to still be invented or developed?
Peter: We have a lot more work to do on opening up tuning – my friend Khyam Allami has done great work illuminating this.
But then, as I say that, I don’t really need a new tool (apart from finding a better controller). I just need to go back to some maqam study; my Lebanese ancestors would insist on it, surely.
Meanwhile, even those basic tuning technologies already work reasonably well – they just need to be more widely implemented. Better MIDI implementation for control in software and hardware would also be welcome, even before you get to new more expressive developments like MPE or MIDI 2.0.
Benjamin: I agree there is more to be explored in terms of tuning. Speaking of AI, I’d still love to have something that would analyze and prepare bits of long jams I like to have with my friends - everyone is always way too lazy to edit them. I have a pretty specific set of ideas / features for that, but that would definitely explode this interview. (laughs)
Other than that I feel that we should explore different haptical interfaces when it comes to hardware - there is more than buttons, encoders, keys and what we usually use.
