Studio's music school challenges MasterClass's passive-watching model – students leave with finished work, not just inspiration from watching the pros.
ENTRY ANGLES
AI-powered personalized learning platforms with minimal required student effort · AI instructor courses focused on outcomes where AI can substantially complete work for students · Songwriting and creative skill courses using AI to adapt instruction by student ability level
VERTICALS
CAPABILITIES
AI engine development and refinement for personalized instruction, Content licensing and curation, Platform technology for adaptive learning tracks
STUDIO FOUNDER
“the music school of the future,”
MasterClass raised $461.4M teaching people to cook, compose, and film by watching experts they admire. Studio is taking the same underlying market and asking a different question: what if you didn't just watch – but left with something you actually made?
But Studio recently launched something new: what it calls "the music school of the future," with the first cohort opening this February.
This isn't a program for beginners. Applicants need existing experience in songwriting, music composition, or production. The school positions itself for "serious hobbyists," "rising artists," and "established professionals."
The compelling hook: students will create one "great song" every month.
Pricing is flexible – $199 per month, or $1,799 for a full year (a 30% saving). Either subscription can be paused. After enrolling, students complete an online intake interview with the school's AI instructor, which asks about their background and goals.
Based on the student's starting level and objectives, the AI instructor produces an individualized monthly learning plan that includes:
- a curated selection of lessons pulled from Studio's music course catalog, - a set of assignments to complete, - a schedule with deadlines for each lesson and assignment, calibrated to make the month's output goal achievable.
The AI instructor also assigns each student to a group of roughly 20 peers, matched by skill level and learning goals. Groups move through monthly "learning sprints" together, starting and ending on the same day. Group composition can shuffle between months as students drop off or join.
Each sprint runs on a weekly cadence: week begins on Monday, and by Sunday students post their interim work to receive feedback from both the AI instructor and their group peers. Throughout the week, students can exchange messages with the AI and with each other in the group chat.
The goal is one finished, publishable song at the end of the month – though students can experiment with multiple ideas or focus deeply on one. Either way, the school promises a ready track that students can share on social media or upload to a streaming service.
While a subscription is active, students have access to all 110 Studio music courses in the catalog – not just the lessons in their individual plan.
Studio's most recent funding was a $50M round in April 2022, which brought total investment to $60M. A successful school launch would likely support another significant raise.
The old model of online education: create a new course for each distinct goal or target audience. The result: an internet flooded with overlapping courses covering near-identical content, with minor manual adjustments for slightly different skill levels or objectives.
The new model: automatically assemble a personalized course from pre-existing content pieces. The building blocks can be book chapters, articles, podcasts, video clips – recombined in the sequence most appropriate for a student with a specific starting point and a specific end goal.
Platforms built on this model license content directly from creators, assembling large catalogs from which AI engines can build custom "learning tracks" – as opposed to monolithic courses built for a single stated objective.
Several startups are already doing this at scale:
- Go1 – raised $413.7M in investment.
- Odilo – raised $84.9M.
- Obrizium – raised £12.4M.
- Quench – raised $10M.
- Innential – raised €1.05M.
Notably, almost all of these are B2B companies selling employee training to enterprises. Studio is one of the few examples pursuing this model in a consumer context – and on a subject as non-commercial as music, where the fraction of learners who ever earn a living from it is tiny.
Studio doesn't need to license outside content – it already has 110 music courses in its catalog. What it's doing is repackaging that existing content in a trending format (personalized learning tracks) to attract more subscribers. That itself is an interesting insight: the dominant assumption has always been that growth for education platforms comes from adding more courses. Studio is testing whether reframing the same content differently can be just as powerful a growth lever.
But personalized tracks alone would not have been enough to drive meaningful revenue growth. The real engine is the guarantee: every student will finish the month with at least one song they can be proud of.
The school's website says this is possible because students "follow a proven learning process." But that's not the real explanation.
The actual mechanism is that AI is already capable of generating lyrics and music of reasonable quality from a user's prompts. An AI instructor can, in practice, compose a passable song for a student even if the student just describes what they want. If the student also makes their own effort – and the AI presents its contributions as "suggestions" and "feedback" rather than outright authorship – the output aligns even more closely with what the student was hoping for. And the student will genuinely feel they created it themselves.
The most talented students will write good songs using the lessons alone. But more important is the pass rate. Enrollment grows when more students hit their goals – just as parents choose schools with higher university placement rates, and students choose universities with higher graduate employment rates.
Not everyone is naturally talented at the thing they're trying to learn. So how do you maximize the percentage of students who get results?
First, give them an AI assistant available 24 hours a day, capable of extracting the maximum from whatever ability the student has – something no human instructor could sustain, or could only provide at prohibitive cost.
And second, arrange things so that the AI can quietly fill in whatever gap remains between the student's effort and the promised result.
Online education is an enormous market – an extension of the oldest market in human history. It's profitable, which is why nearly every self-described expert in any field is advertising their course online.
The first shift is where the money flows. Revenue in this market no longer accrues only to experts who create content – there's now real money for builders, the technologists and business-minded operators who construct personalized learning track platforms. Technologists build and refine the AI engines; operators handle content licensing. If you're technical, find a business co-founder; if you're commercial, find a technical one.
The second shift is how little catalog you actually need. The insight is to build courses around topics and outcomes where the AI instructor can achieve the promised result with minimal student effort.
The old rule for successful mass-market courses: teach people something any reasonably motivated person can learn. Otherwise, most students fail, word spreads, and the course never scales.
The updated rule: build a mass-market course around something AI can substantially do for the student. That way, the maximum possible percentage of students succeeds. The AI helps different students in different proportions – from 0% for the naturally talented and self-directed to nearly 100% for those who need more support.
That's the model Studio appears to be applying to songwriting. The question worth asking: what other popular learning topics could work the same way right now?
Find one, and you've found the next move.