The Next Jay Chou: Human or AI? | A Conversation with the CEO of *Youci Mountain* and the GM of *Wajijiwa Music* on Suno
Music remains scarce — Suno is just a pretty little toy.
Hi, I'm Koji. "Crossing" is my personal newsletter. The name comes from something Steve Jobs said about standing at the intersection of technology and the humanities.
I'm co-founder of the sleep environment brand Tangdao and the content company The Fair. I previously founded the Jiepang app and was part of the founding teams at Fanfou and Jumei International Holding Limited.
As a classic cross-disciplinary entrepreneur — from mobile internet to e-commerce, from content to consumer goods — I've let curiosity carry me across all kinds of "crossings."
The release of Suno v3 landed like a thunderclap. Songs — the content medium hardest for ordinary people to create — were "democratized" overnight.
No one can resist a spectacle. The internet flooded with hyperbolic headlines: "The music industry is finished," "There will never be another Jay Chou."
A week has passed since Suno v3 dropped. The noise is subsiding, but the heat remains. Rationality is slowly returning. This feels like a good moment to discuss it.
So I interviewed two CEOs from leading music labels to hear their perspectives.
Speaking out on a globally controversial topic in music takes frontline insight, plus some courage and pioneering spirit. I want to thank these two friends in advance:
- Xi Tao, founder/CEO of Youcishan Culture, whose artists include Good Meimei (Qin Hao, Zhang Xiaohou), Chen Li, Jiao Maiqi, Wang Jiayi, Chen Jingfei, and others.
- Xin Zhiyu, general manager of Wajijiwa Music, whose artists under Wajijiwa include Mao Buyi, Ma Boqian, Fool and Idiot, Zhou Zhennan, Zhao Lei, Zhai Xiaowen, Yan Xujia, Xia Zhiguang, Peng Chuyue, Guo Zifan, Air League Band, and others.
❶
What is Suno? — "How can this possibly be real?"
Suno is an AI that generates music. Its v3 release on March 22 marked a leap forward.
You input a request for a song. About 15 seconds later, out comes something surprisingly listenable.
For my first Suno creation, I asked: "Please write a song for my daughters, in Mandopop style. Include their names, 'Rose' and 'Sunflower,' plus the words 'spring' and 'dragon.'"
In 15 seconds, Suno AI generated: lyrics, a female-voiced Chinese song, and a cover image based on the lyrics. Have a listen:
In my heart There is a garden Where roses bloom Where sunflowers shine bright Spring water, clear and transparent Flows through my heart You are my dragon Soaring within me
Overnight, Suno may have generated millions of tracks. Search Xiaohongshu, Douyin, Bilibili, or WeChat — you'll find an endless supply.
One friend fed Bei Dao's poetry into Suno as lyrics and asked it to compose and perform.
Of all Suno outputs, this might be the one I've replayed most. I looped it a dozen times. At first, I was curious about what AI could do. Then I kept looping it because "it seemed to move me" (前提是我知道这首歌词来自北岛，联想到他的故事和态度).
❷
Why is Suno so impressive?
For ordinary people, music creation is the hardest form of creation.
Photos and video are more complex media, but not harder. Today our grandparents can easily create with smartphones, sometimes capturing deeply meaningful footage.
If we look for analogues in AI history, Suno v3 resembles Midjourney v3 and GPT-3.0 — a catalytic version.
If we look for analogues in the history of creative tools, Suno v3's "moment" resembles the "smartphone moment" for photography and videography: from artists-only to everyone-is-a-creator, from public-facing art as primary function to personal consumption as primary function, from pursuing artistic works to realizing personal taste or small-scale shared memories.
After Suno v3, music production and consumption will both change. The entire industry edifice around music will start creaking. Change will emerge from every direction.
❸
Thinking About Suno
First, music splits in two: commodity and work
What's a commodity?
Much music is simply soulless commodity: noisy earworms for short video, ambient sound for long video, clumsy background piano at Western restaurants.
Even when people listen to foreign-language songs, they're often just listening for atmosphere, for flow state.
Post-Suno, music-as-commodity will be infinitely supplied. This portion of the industry will take a hit.
What's a work?
Let me share when music last made me cry.
It was last night, before bed. Huang Qishan's "Mother and Daughter" — I cry every single time. The first three lines are tear-triggering ordnance: "My child, I don't know if I've done well enough, it's my first time being a mother."
The time before that was last Saturday. I was driving with my kids when Wang Zhengliang's "Where Has the Time Gone" came on randomly. Before leaving, they wouldn't put on clothes or shoes. I'd just finished a massive blow-up; my older daughter was in tears. In the car, they instantly turned sunny, competing to share kindergarten stories. Hearing the lyrics "Raising children all my life, my mind full of their crying and laughing," I suddenly teared up at the wheel. Then secretly wiped them dry, not wanting the kids to see.
Then there's Leslie Cheung and Zhao Wei.
"I am what I am, a firework of a different color" — this song can only belong to Leslie Cheung. Anyone else singing it would be cringeworthy, toe-curling. But when he sings it, detached yet deep, fragile yet frenzied — countless images of him flood my mind. Deep empathy.
"There's a girl, a bit willful, a bit arrogant... Hey, that's this girl" — this song can only belong to Zhao Wei, only to the Zhao Wei of the Little Swallow era. Person and song as one. Any other girl singing it would seem deranged. Only Zhao Wei makes it boundlessly innocent, invincibly cute, the girl everyone wants to go spring outing with.
— This is what I consider a work: it demands the singer's interpretation, even demands person-song unity.
The next Jay Chou — human or AI?
So when facing this question, my answer is clear: only a human.
Because I absolutely cannot imagine an AI singing "A pretty girl who makes me blush, a bad girl who drives me crazy" and somehow moving me. Only a young, shy, radiant Jay Chou could deliver that blend of rebellious swagger, playful irreverence, and tender charm.
❹
Koji Interviews Xi Tao and Xin Zhiyu
①
👨💻 Koji: Suno-generated songs have spread widely online. Beyond "technically impressive/realistic," have you heard any that truly moved you?
👦🏻 Xi Tao, founder/CEO of Youcishan
No.
🧑🏻🦱 Xin Zhiyu, general manager of Wajijiwa Music
I haven't heard a single Suno output that truly moved me. I don't even want to use the word "work" for these things that can only be called pieces. Suno may be novel, may be democratic — we should maintain sustained attention and respect for new things. But based on everything I've heard so far, the listening experience is more like a pile of beautiful little trash.
Musical works remain scarce. Suno is just beautiful little trash.
②
👨💻 Koji: How do you think Suno will impact and change the music industry?
👦🏻 Xi Tao
Suno won't change the music industry, because whether it's music or painting, art is its core. The underlying logic is: humans are warm, emotional, feeling beings. With this underlying logic, things identical in form become completely different in essence. So Suno can't change the music industry.
What Suno affects are companies and people who were already mass-producing, assembly-lining music. The emergence of AI, or software at different stages represented by Suno, replaces the mechanical things humans do.
🧑🏻🦱 Xin Zhiyu
It's too early to discuss Suno's impact on the music industry. Compared to pure text generation, music involves melody, arrangement, performance, and vocalization. AI generation across these stages is clearly more complex at the model algorithm level. This will take longer.
In the future there may be two product paths. One is adding AI function plugins to existing applications to help UGC platforms with content output — the ecological path of AI to CapCut, CapCut to Douyin. The other is in the professional music track, integrated into existing DAW software to help professional creators improve production efficiency.
More than the above, I'm concerned about music copyright确权 issues. I don't know whether Suno's underlying learning model mechanism involves infringement. Can generated works be legally defined as original? Are AI-generated musical works recognized as copyrighted? Who is the rights holder? These will challenge existing copyright management systems, and different countries take different stances. Clear legal definitions are prerequisites for the content industry's development.
AI never widens the gap between ordinary people and ordinary people. It widens the gap between experts and ordinary people.
③
👨💻 Koji: For yourself (or your artists), do you have any ideas or inspiration for using Suno in future creation?
🧑🏻🦱 Xin Zhiyu
One scenario is immediately viable: Suno can provide creators with melodic motifs in a highly cost-effective way, serving as inspiration triggers, with humans then completing full compositions. This can improve creators' efficiency.
That's concrete and real.
④
👨💻 Koji: Bold prediction — what kind of new "music idol" might Suno produce?
👦🏻 Xi Tao
Whether AI can produce music idols — I find the question itself laughable. If this actually happened, it would prove humans have no emotional warmth left, entirely enslaved by machines, enslaving our own emotions. I completely disagree with all AI directions aiming to replace art. I think it's a waste of time, with the goal of turning humans into machines.
🧑🏻🦱 Xin Zhiyu
AI won't bring us new music idols. Because music-related scenarios — whether performance, singing, or singing-dancing — all require physical and spiritual injection. Things without human flavor cannot move people. Additionally, output consistency and scenario consistency remain major unresolved problems for music AI.
⑤
👨💻 Koji: In the future music industry, what aspects will never be changed by Suno?
👦🏻 Xi Tao
I've always believed AI is an excellent tool, but limited to tool. Human nature is the biggest difference between humans and machines. Everything in the future process will become faster and faster, but human nature will become more and more important. Emotion will never be changed by AI.
🧑🏻🦱 Xin Zhiyu
One thing is beyond doubt: music production barriers are lower, and new song catalogs on streaming platforms will grow by tens of millions annually (currently millions). Under congestion effects, the difficulty and cost of discovering good music will surge sharply, returns will be diluted, and musicians' survival will become harder.
But this also makes good works more precious. We say truly vital works achieve person-song unity. AI may solve the "song" problem, but the "person" level still ultimately depends on people.
However, AI's greatest value is squeezing out those practitioners who coast and cut corners. That work will be replaced by AI. I encourage everyone to reach higher.
❺
Closing: More Music, More Joy
Let me close lightly with something a netizen wrote:
After Suno, all text can easily transform into music.
Beyond telling stories with characters and plots through music, like Bob Dylan, we can convert today's news and podcasts into music, hearing major and minor affairs in song.
Of course, you can also tell your own story with music — try turning your diary into music?
Or like the Shanghai Rainbow Chamber Choir, use music to complain about life's trivial annoyances.
If you're quite humorous, you can create musical stand-up like Wang Mian.
What's unprecedented: You can combine topics completely unrelated to music with music itself, creating an interesting sense of contrast — a product manual, a social media post...
I fed the above passage into Suno too. The result was unexpected.
Haha, more music, more joy: