A look at Music AI’s realtime stem separation and dynamic volume control (Photo Credit: Music AI).
Generative AI presents serious threats and challenges to the music industry, and it’s something we cover exhaustively at DMN. However, alongside these daunting problems are exciting AI breakthroughs that are transforming the music creation and collaboration process, intensifying fan engagement, and solving consumer electronics problems. Just ask Music AI, the company behind the Moises app and one of the less-controversial players pushing these ethical AI areas ahead.
If you’re a DMN reader or bumping around music industry conferences like NAMM, Music Biz, or SXSW, you’ve probably encountered Music AI. This is one of the more unassuming – yet highly impactful – companies in the music space, although it’s not stirring up lawsuits or battling over government policy.
Instead, Music AI is firmly on the ethical side of AI, which essentially means – to be blunt – they aren’t ripping off content and calling it fair use. On the contrary, Music AI is pursuing significant changes in areas such as stem separation, enhanced music creation, and intelligent consumer electronics applications, among other rapidly evolving fields. Just recently, DMN joined forces with Music AI to further showcase these competencies.
At NAMM in Anaheim earlier this year, I sat down with Matt Henninger, Music AI’s VP of Sales & Business Development, who illuminated just how expansive the intersection of AI and music has become.
It was an eye-opening chat. Henninger was getting pulled in many different directions – he’s an in-demand guy – though in less than thirty minutes, he described a far broader evolution happening in music AI.
This is a discussion that quickly goes beyond the copyright problems presented by upstarts like Suno and Udio. However, even in areas like instruments and consumer electronics, Henninger told me that more sophisticated discussions are just starting to occur.
“Companies are just starting to ask the right questions,” Henninger relayed. “They’re wrapping their heads around the extreme possibilities here.”
So what are these possibilities?
At this stage, they’re only limited by the imagination, though Music AI told DMN that they’re having nuanced discussions that traverse DAWs and collaboration software, sophisticated amps, next-gen instruments, and integrations within devices like TVs.
One of the more fascinating ‘bleeding edges’ is volume control and on-the-fly stem separation. Music AI has teamed up with Qualcomm’s Snapdragon NPU (essentially an AI-specialized, high-processing version of a CPU) to enable on-device, real-time volume control of music, whether audio, video, or both.
It’s called ‘Moises Live’ and utilizes an edge AI model that runs on the Snapdragon X Series processors. This technology, branded ‘AI Smart Volume,’ enables customized audio personalization without relying on the cloud – which means adjustments occur without any lag.
It’s something that takes a moment to process (in your brain’s CPU, that is). Simply toggle a component (such as a guitar or voice) down, and it disappears instantly, enabling isolation or instant EQ customization.
Henninger took things a step further. After all, why limit this amazing stem-separation breakthrough to a music video? Imagine a smart TV that had three volume buttons instead of one. Want to tone down the special effects and bring out the dialogue? Toggle away.
“TV audio EQ can be an annoyance for some people, so now they can optimize the mix,” Henninger relayed. “For people experiencing hearing issues, there’s suddenly a sophisticated, easy solution to reduce excessive ambient noise or bring out dialogue more clearly. Problem solved.”
But that’s just one possibility among the endless AI-powered applications Music AI is cooking up.
The company is busily working across a number of arenas, including continued enhancements to stem quality and a beefed-up ‘AI Studio’ with expanded recording and creativity tools.
Music AI is also continuing to democratize the technology by offering concurrency-based access to models, allowing businesses to leverage state-of-the-art cloud models at fixed prices, thereby removing the concern of unlimited scaling costs.
Shifting into the DAW space, the creative and collaborative possibilities are nearly endless, with audio enhancement and restoration also part of the mix.
It all sounds alluring, no doubt, though Music AI execs were quick to stress that these intelligent, AI-powered capabilities are challenging to implement. “I think companies that are just starting to hire their first [machine learning] engineers are already too far behind,” Henninger relayed. “Bespoke AI companies already have a five-plus year head start.”
The takeaway: start these projects in-house if you dare, though you’re probably going to spend most of your efforts catching up.
If you’re a musician, you’re probably already aware of Music AI’s serious head start in this space.
Music AI, the company behind the Moises app, has been advancing artificial intelligence as it applies to music and audio technology for years. Hop onto Moises, and you’re suddenly jamming on your chosen instrument with perfect accompaniment – on any song you choose.
The best part might be watching others flexing their virtuoso chops – while generating their own fans in the process. Moises’ core technology involves AI-powered audio solutions such as the aforementioned stem separation, which easily isolates individual instrument tracks from a mixed audio file for the ultimate accompanying band. In the process, Moises is taking participatory engagement to an entirely new level and expanding super-fandom in the process.
Currently, Moises has more than 60 million active users and was Apple’s iPad App of the Year in 2024. Now, they’re broadening the scope and application of their significant AI-related breakthroughs. In total, Music AI processes over 2.5 million minutes of audio per day, reflecting substantial operational scale.
Investors are certainly noticing. The company recently secured a Series A round of $40 million led by Connect Ventures—an investment partnership between Creative Artists Agency (CAA) and NEA—and monashees, a prominent Brazilian investor (interestingly, Brazil is one of the fastest-growing music markets according to IFPI figures).
Music AI’s latest investment follows an earlier seed round of $8.35 million, and is being directed towards further development and expansion of Music AI’s technology, including applications beyond Moises into broader enterprise solutions.
And the best part: unlike most heavily-funded tech plays, Music AI’s execs stressed their dedication to ethical AI, which involves training models on licensed content and ensuring fair compensation for creators.
Sadly, that’s the less noisy side of AI music, with a focus on tech instead of acrimonious lawsuits and high-stakes battles in Washington. But it’s just as groundbreaking.
