Who are you at Neutone?

Hi! I’m Naotake and I’m an AI researcher/engineer at Neutone. I’ve been working on music-related AI models since my bachelors at the University of Tokyo. For my Master’s and Doctor’s, I worked on AI models to assist producers to better utilize their synthesizers. I finished my Doctor’s in 2023 and started working for Neutone.

What do you work on?

I’m in charge of most of the AI side of Neutone Morpho, like developing and experimenting with new AI model architectures and training pipeline. I dig through papers everyday to see which ideas can be incorporated into Neutone Morpho or FX.

Why are you doing this?

I’ve been interested in music production, especially sound design, ever since I got into dark D&B and neurofunk music in when I was in high school. I’ve also been fascinated with AI/machine learning since the days of DeepDream around 2015~2016. So naturally, my research was focused on the intersection of two. And I’ve been lucky enough to keep working on it here at Neutone.

What do you hope Neutone Morpho offers artists/musicians?

I think the cool thing about DeepDream was that it was using AI models intended to classify images in an unintended way to produce an otherworldly image. Musical instruments are also often used in ways that their creators didn’t expect. For example, Roland didn’t have acid house in mind when they created the original TB-303. Neutone Morpho is a very new type of plugin with lots of room to create weird sounds, so it’s impossible to tell what everyone will come up with it.

I don’t think that Morpho will help you create the perfect piece of music that you’ve always wanted to make. Rather, I hope you can use it to explore new sounds you never dreamed of, to interact with AI models like no one ever has, and to break norms of existing musical genres.

What’s next in store!?

We’ll keep working on the AI algorithms to improve the quality, reactivity, and efficiency of the Morpho plugin. But we also want to create totally different audio effects and instruments. I think the field of AI audio effects is still nascent and there’s loads of things we can do. There is still a significant gap between a cool AI research paper and something that’s actually usable in music production, and we hope to be the ones closing that gap.

What are you listening to these days? 🎧

I’m always listening to bass music, mostly drum and bass. I’ve been a big fan of Billain for a long time. The sound design and beats are insane and the overall “cyberneuro” aesthetics is cool.

ALEPH’s SEPULCHRE was my favorite album of 2023. I’m not an expert on garage music but the texture and the atmosphere is incredible.