So is it inevitable that over the next few years every plugin will have some sort of AI feature?
“There is a bit of a divide in the market in the way plugins are designed,” says Bailey. “There’s analog hardware emulation, then there’s pure software products like iZotope and FabFilter. I know companies in both categories that use deep learning, even companies like Univeral Audio which everything emulates analog hardware, ML gives us techniques to do it more efficiently. My opinion is that we are going to see more and more of them.
As companies like iZotope and UA harness the potential of AI both under the hood and as part of the user experience, other companies are taking a more extreme approach, completely rethinking what AI and the ML could mean for modern DAW and rewrite the rulebook of how music is made using computers.
Never Before Heard Sounds was founded in 2020 by Yotam Mann and Chris Deaner. “I felt like there was a huge opportunity for things that didn’t just show off the power of AI, but that used the power of AI and gave it to real musicians to create music. new and interesting music, ”Mann says. By putting the musician first and the tool second, NBHS aims to rethink the way we interact with AI in the studio and on stage. “As musicians, we are not interested in automatic music creation technology, which is largely in the direction that AI music production seems to be taking – very few people are trying to create tools to increase musicians in the world. instead of replacing them. ”
In addition to collaborating with Holly Herndon on Holly +, they’ve created a website that allows you to resynthesize the audio from any video on YouTube and convert it to a choir or string quartet. Try it yourself here. Remarkably, they also built a prototype hardware that can resynthesize any audio entering a pattern of another sound, in real time – a world first.
“The angle we’re taking is not to try to wrap it in magic and say that some self-magical AI creature is going to solve your musical problems for you,” Mann continues, “but to try to make algorithms as transparent as possible and as useful as possible. This is why we call our tools instruments. They are meant to be held and played with, not to apply some kind of magical layer on top of your production.
NBHS co-founder Chris Deaner, accomplished drummer for US band Plus / Minus, adds, “In some ways what we do is a lot less sexy than a tool that does it all for you, which was and is the promise of IA and ML. For both of us musicians, we’re really interested in the human part of it.
The human part of modeling in particular, as we discussed in Part 1 of this series, brings its own ethical challenges.
“One of the risks of AI is that you end up with a song-generating machine that makes great compositions but never credits any of the things it used to make them,” says Mann. “So for us it was really important to showcase the musicians who participated in it. We think of these generative models, not as AI entities themselves, but as an intermediary between you, the final musician, and the musicians involved in the modeling.