Mikey Shulman is sitting in front of a 1920s Victrola in his Cambridge office, gesturing toward the vintage record player to make a point about fear.

“Back then, people thought recording music at all was a bad idea,” he says, shellac 78s spinning behind him. “That it would stop you from going to the symphony.”

The Victrola isn’t just decoration. It’s a deliberate reminder that every technological shift in music, from sheet music to recordings and vinyl to streaming, has triggered the same cycle of knee-jerk resistance and eventual acceptance.

Now Shulman, the co-founder and CEO of generative AI music platform Suno, finds himself at the center of music’s latest existential crisis: what happens when anyone can create a song with a text prompt?

The question has already become more than theoretical. By mid-2024, the major labels had taken Suno to court, and the music world was split between panic and fascination. To some, it was the greatest threat since Napster. To others, the next creative frontier. Either way, with a reported unicorn valuation and more than 10 million users, ignoring it is no longer an option.

“The most common complaint against Suno is you make terrible music and you’re replacing artists,” he says. “Those can’t both be true at the same time.”

“The most common complaint against Suno is you make terrible music and you’re replacing artists. Those can’t both be true at the same time.”
MIKEY SHULMAN

At 38, Shulman has the particular confidence of someone who’s already lived through one technology cycle. After earning a PhD in physics from Harvard and teaching at MIT’s Sloan School of Management, Shulman co-founded Kensho Technologies, a Cambridge startup that taught machines how news moves markets. Within four years, Kensho became one of the most respected names in applied AI, selling to S&P Global for over $500 million in 2018.

The exit gave Shulman something rare in tech: time. He tinkered, picked up new instruments, and began experimenting with sound while trying to make computers not just process data, but listen.

By 2022, as generative models like GPT and Stable Diffusion captured the world’s attention, Shulman was already exploring how similar architectures could learn melody and rhythm instead of language. What began as a side project with former Kensho colleagues quickly evolved into something more significant than any of them expected. In many ways, Suno feels like a return to his roots.

“The most fun I’ve ever had is playing music with my friends,” he recalls. He started piano at four, picked up bass at 12, played in bands through high school and college in New York City. “It brings you incredibly close to people.”

From Discord to Disruptor

Fast forward to March 2023. Silicon Valley Bank had just collapsed. Shulman and his co-founders — including his best friend — were standing in a kitchen, watching their audio understanding startup generate modest revenue. They’d built something practical. But they’d also discovered, almost by accident, that they could make music.

Not good music, not yet. But music.

“We thought we would need a hundred times more money and a hundred times more compute,” Shulman recalls. They were wrong. The breakthrough had come sooner than expected, which meant they faced a choice: keep building the sure thing, or pivot to the dream.

“We’re like, what are we doing? The world will probably enjoy what we’re playing with as much as we do.”

They killed the revenue-generating business and launched a Discord bot that made songs. Within weeks, people were subscribing. Not just tech bros looking for the next toy, but regular people who’d never made music before. Parents making songs for their kids. Teenagers turning inside jokes into tracks. Office workers who hadn’t touched an instrument since middle school band suddenly creating.

“Build people the product they actually want to interact with.”
MIKEY SHULMAN

“If I had told you in 2020 there’s going to be a Discord bot where you type in a description and music pops out, you would have said that’s a party trick,” Shulman says. “But it turns out people spend hours a day on it because it’s an incredible mirror into your mind.”

Shulman thought they could stay on Discord forever. He’d watched Midjourney thrive there, building a massive community around AI image generation. His co-founder Martin told him he was an idiot.

It took only five days for 90% of traffic to migrate to their hastily-built web app, which didn’t even have all the Discord bot’s features.

“Build people the product they actually want to interact with,” he says, the lesson still fresh two years later.

The shift from Discord to web revealed something fundamental about Suno’s user base: they weren’t all tech-forward early adopters willing to navigate Discord’s quirks. They were largely people who wanted to make music, period. The simpler the path, the better.

By August 2023, they’d released V0 — so primitive you couldn’t even choose the genre. The model inferred it from your lyrics. The audio quality was, in Shulman’s words, “terrible.” But something extraordinary happened: people loved it anyway.

The Ozempic of Music

Over the next nine months, Suno evolved from a curiosity into the leader of a movement. V1 and V2 improved audio clarity and lyric coherence. V3 added longer songs, better vocals, and distinct instruments. Each update sparked a wave of TikToks and Reddit threads of users turning prompts into mini albums.

By early 2024, Suno songs were quietly circulating in producer group chats. In private, label executives and A-list producers were experimenting with it, much as they once quietly tested Auto-Tune or sample packs.

“At this point, it’s the Ozempic of the music industry,” someone told Shulman recently. “Everybody’s on it and nobody wants to talk about it.”

The comparison stings and amuses him in equal measure. Behind closed doors, Grammy winners use it daily and songwriters run ideas through it. But publicly? Silence.

“I would much rather that than the opposite,” Shulman says. “Where two people use it and they’re vocal about it.”

“It’s the Ozempic of the music industry. Everybody’s on it and nobody wants to talk about it.”

The secrecy reflects the industry’s deep unease about AI. Unlike other creative fields where AI augmentation has become normalized, such as writers using ChatGPT or designers using Midjourney, music carries unique cultural weight.

The backlash has come from all sides. Established artists called it “soulless.” Music teachers declared it “the death of creativity.” Producer forums filled with rage about “lazy prompters” flooding the market. When musician Nick Cave was asked about AI songwriting, he called it “a grotesque mockery of what it is to be human.”

The most visceral response came in April 2024, when thousands of musicians signed an open letter warning that AI music tools were “trained on our work without permission” and would “devalue human artistry.” The signatories ranged from indie artists to Grammy winners, united in their view that Suno and its competitors represented an existential threat.

And yet, some of the most inventive artists are already exploring what the tools can do. Timbaland discovered you could let the model babble — no words, just flow and rhythm — then write lyrics to match. “The flow is the most important part,” he realized, flipping the traditional songwriting process.

“Music plus technology has a checkered past,” Shulman acknowledges. “There is an inherent tentative or somewhat cautious outlook that most people in music have toward new technologies.”

“Music plus technology has a checkered past. There is an inherent tentative or somewhat cautious outlook that most people in music have toward new technologies.”
MIKEY SHULMAN

The Label Lawsuits

By June 2024, the curiosity had become a threat. Universal, Warner, and Sony filed suit, accusing Suno of training on copyrighted songs without permission and “saturating the market with machine-generated content that will directly compete with, cheapen, and ultimately drown out the genuine sound recordings on which the service is built.”

This September, the labels reportedly amended their complaint with new allegations that Suno had circumvented anti-piracy measures by systematically ripping streams from YouTube. The stakes: up to $150,000 per infringed work, potentially billions in damages.

Shulman declines to directly address the allegations. “It’s not a great idea to comment about specifics in an ongoing litigation, whatever press was out there did not come from us,” he says. “But I think if you zoom out, it would be a shame and worse for consumers if there was a world of AI music and a world of non-AI music.”

“It would be a shame and worse for consumers if there was a world of AI music and a world of non-AI music.”
MIKEY SHULMAN

The tension is existential. The music industry has been here before with the rise of piracy and streaming. Each time, the pattern repeats: initial resistance, legal warfare, then eventual, often reluctant, integration. But AI feels different to many, because it strikes at the soul of music creation itself.

“Every single person in music who makes it gets sued… you just don’t hear about most of them,” he says. “There’s a ton of legal action against people in music. I still maintain there’s so much more to do together. Music isn’t something the tech bros in San Francisco can just disrupt — it’s something we build together.”

Behind the scenes, the same labels suing Suno are reportedly close to striking licensing deals with AI music companies. It’s the hedge-your-bets approach that defined the streaming wars: litigate with one hand, negotiate with the other. At this stage, the genie is not going back inside the bottle.

“Unlike other forms of media, this isn’t a domain where you’d want things to be separate,” Shulman argues. “Tons of hit tracks have AI in them already. Music is music. The way it gets made should be irrelevant to consumers.”

“In that frame, maybe it’s not so crazy that you’d talk to everyone,” he says, smiling slightly.

Studio Evolution

This past month, Suno released V5 alongside Studio, a full multitrack editor that includes stem exports and remix tools. It provides many of the functionalities that producers have long requested, accessible enough for beginners and powerful enough for professionals.

“Nobody thinks about singing into a DAW and saying, ‘make that trumpet,’ except we do,” Shulman says. “It’s an intuitive, easy way to build music track by track.”

The dual release represents Suno’s answer to serving two masters. Casual users who want to “fall down the rabbit hole” need songs that feel like theirs on first listen. Professionals need granular control, weird artifacts, the ability to pull individual stems into Logic and remix them.

“The real goal is to make a song that feels like it’s yours,” Shulman says. “When you use Studio, you can’t say you didn’t make it — you arranged it, edited it, chose everything yourself.”

Shulman’s view of authorship runs counter to the music industry’s fixation on control and scarcity. In that world, there’s only so much attention, so much money, so many playlist spots. Every new artist threatens an existing one. Every new technology erodes an old revenue stream.

“It’s a failure of imagination. Spotify already has infinite music. Adding more infinite music isn’t the problem.”
MIKEY SHULMAN

“It’s a failure of imagination,” Shulman argues. “Spotify already has infinite music. Adding more infinite music isn’t the problem.”

He sees a different future: not AI music versus human music, but interactive music versus passive listening. The format of yesterday was physical media. Today’s format is streaming. Tomorrow’s?

“The format of the future is music you can play with, not just play,” says Shulman. “That special social connection that the entire audience feels… while maybe we can’t get all the way to the satisfaction of a live concert, we can make huge strides compared to the lean-back experience that most digital music is today.”

Half of Shulman’s Suno usage involves his four-year-old son. Together, they create personalized songs about fantastical adventures, played on repeat until they’re memorized. “There’s an alternate universe where Suno is just a kids’ product and makes five times as much money,” he jokes.

His own personal creations range from practical to absurd. He’s re-orchestrated piano pieces he struggled with as a child. Created an entire album of gospel songs about Hanukkah. He’s working on a concept album called “Let’s Circle Back to My Place,” featuring cheeky workplace innuendo transformed into music.

But what really drives him is the hunt for new sounds. “Remember the first time you heard Skrillex and you’re like, what the fuck is that?” he asks. “That should happen more.”

He pushes the model’s boundaries, cranks the weirdness sliders, searches for sounds that don’t yet have names. Professional producers join him in this quest, looking for the perfect “artifacty” sounds, the glitches that become features, the mistakes that define new genres.

“I’ll hear something and think it’s mid,” he says. “They pull it into Logic, do five things, and suddenly it’s amazing.”

Music’s Next Movement

From his Cambridge office, deliberately distant from Silicon Valley’s breathless hype cycles (“Remember when Sam Altman got fired and rehired? Everyone in San Francisco was glued to Twitter. We had a normal weekend”), Shulman is building something that looks deceptively simple: a tool that lets people make music.

But the philosophy behind it is radical. The music industry has long treated creation as an exclusive domain for those with talent, training, and access. Suno challenges that idea by leveling the playing field.

“It’s a basic human desire to create,” Shulman insists. “And the validation when other people like what you made. People don’t want to say that part out loud. It seems vain. But you shouldn’t be embarrassed.”

Shulman may prefer to work far from the tech scene’s frenzy, but Suno’s trajectory tells another story. In May 2024, Suno announced a $125 million funding round led by Lightspeed Venture Partners, Nat Friedman and Daniel Gross, Matrix and Founder Collective. It’s now reportedly raising more than $100 million at a $2 billion valuation, evidence that the industry sees not just controversy but opportunity.

Even so, Shulman resists the idea that he’s out to disrupt the music business.

“I wouldn’t even say we’re seeking to disrupt. Like, we’re seeking to grow. And I don’t think we’re actually terribly disruptive to the traditional music industry.”

“Human agency is incredibly underrated. People act like AI is coming and there’s nothing we can do. That’s wrong.”
MIKEY SHULMAN

“Human agency is incredibly underrated,” he continues. “People act like AI is coming and there’s nothing we can do. That’s wrong. We can build a good future or a bad future. We should build the good one.”

The interview ends where it began, with the Victrola. Shulman demonstrates how much effort once went into playing a single song. The ritual of it — selecting the record, cranking the handle, placing the needle — feels almost sacred compared to asking Alexa to play something.

But in Shulman’s eyes, that’s the point. Every generation thinks the previous technology was more authentic. We romanticize vinyl while forgetting that live musicians once protested recordings. We celebrate guitar heroes while forgetting that Bob Dylan’s fans mutinied when he went electric.

“Twenty years ago, people asked if music was made in ProTools or on a console,” Shulman reflects. “Now nobody cares. They asked who would go to a concert to watch someone press play on a laptop. Now those are the biggest shows in the world.”

Suno has generated more than 100 million songs to date. They range from breathtaking to bizarre, from professionally released to privately cherished. Some will find audiences. Most won’t. But together they mark a shift in what it means to make music at all.

“Great artistry is great artistry,” Shulman says. “In five years, we’re not splitting hairs about AI artists versus non-AI artists. There are just artists making great music.”

Editor’s note: This profile first appeared in Luminaries, our weekly deep-dive into the minds moving culture, capital, and code. Subscribe today to receive it straight to your inbox and join private Q&As with our guests.