8 Insights into Crafting Future Sounds

For almost a decade, Google's been diving deep with artists and musicians to figure out how new tech can boost their creativity. Just last year, teams from YouTube, Technology & Society, and Google DeepMind kicked off a new initiative to get artists, songwriters, and producers hands-on with music AI tools and prototypes. They could share their thoughts and help shape these tools. And now, we're excited to share some juicy insights from our work with musicians in the YouTube Music AI Incubator program.

The YouTube Music AI Incubator is a big deal for us, a chance to brainstorm with music pros about AI tools that really meet their needs and amp up their creative game. We had tons of one-on-one chats with global participants, where our research scientists, program managers, and product specialists swapped ideas and got feedback on our ongoing projects. One cool thing we showed off was the Music AI Sandbox – a set of experimental AI tools we announced at this year’s I/O. It's designed to turbocharge the workflow for artists in our incubator.

These meetups have been all about getting honest feedback and having real talks, aiming to shape the future of our genAI music products. We've just released a report that sums up the key insights and observations we've gathered over the past year, working closely with these creative minds.

Here’s a rundown of the top eight insights we've picked up as participants played around with new AI tools to enhance their art:

1. Hearing is believing

A lot of participants were skeptical at first about AI helping with deep music creation. But after giving AI-generated tunes a listen, their doubts melted away. They were amazed and stressed the importance of building these tools with and for the creative community.

2. AI sparks imagination

Folks loved using AI to mess around with new and weird sounds, not worrying about what's musically "right" or "wrong." The unpredictable side of music AI tools makes them different from other music software, sparking creativity and letting artists discover new sounds.

3. More ways to create

While text prompts were okay, participants found them a bit limiting for making music. They wanted more ways to express themselves, which got us thinking about other methods like turning hums and mumbles into digital tunes.

4. AI makes complex sounds easy

Artists found AI super helpful for playing around with complex sounds, like crowd chants or choir singing, to sketch out ideas and see how they might fit into a piece. This helped them decide where to focus their efforts on new musical collaborations.

5. Real-time fun

Playing with AI models in real-time was a blast for the artists. It let them experiment in a playful way, making the process dynamic and collaborative.

6. Tailored tools for different needs

The incubator showed us that artists, songwriters, and producers all have different needs and want different controls from these tools. They see AI as a tool for inspiration, starting songs, finishing songs, or pitching ideas, but they want to stay in control and use their unique skills.

7. Money, credit, and control matter

Almost all the artists in the incubator stressed how important it is to sort out issues like who owns the music, how songwriters get paid, and who's in control. They brainstormed solutions and looked at different ways to tackle these issues. At YouTube, we're not just talking about this—we're working on responsible solutions.

8. Education and access empower

Musicians thought AI tools could be a game-changer for beginners and up-and-coming producers who might not have a lot of tech know-how or resources. They saw AI as a way to make music creation more accessible, offering learning opportunities and empowering people to explore music, no matter their background or experience.

Looking ahead

Our goal is all about co-creation—working hand-in-hand with artists, songwriters, creatives, storytellers, and community members to dream up new possibilities with technology. We think it's crucial to get input from the folks who'll be using these AI tools the most, to make sure they're useful, engaging, and accessible in the long run.

On top of YouTube's ongoing partnerships with all sorts of artists, we're going to keep pouring resources into programs focused on art and tech, like Magenta Studio, Google Arts & Culture, Artists + Machine Intelligence, and Lab Sessions.

Our commitment to sparking conversations about AI and creativity goes way beyond our products. If you want to see how we've been bringing people together to talk about creativity and AI around the world, check out our exhibit on Google Arts & Culture about a discussion we had with artists and The Serpentine Galleries in London. As we keep exploring where technology and society meet, we're keen to connect with and learn from all sorts of communities—artistic and beyond.