Unique Research · Unique Viewpoint · March 28, 2024
Background: On January 24, 2024, Unique Capital and partners hosted the 2024 AIGC Application Development Summit in Shanghai, with nearly 1,000 AI practitioners. This piece is compiled from the guest dialogue.
Guest
- Guest: Wanfeng Chen (陈万锋) — Founder & CEO, Kuaizi Tech (筷子科技)
- Host: Wei Wu (吴畏) — Partner, Unique Capital
Summit dialogue
Q: How will generative AI empower marketing? What changes will it bring?
Wanfeng Chen: Marketing today is essentially content commerce. In an era when content can drive massive traffic, companies must focus on content commerce — especially video production, management, and distribution. Why emphasize production? Mature text-to-video still cannot scale, yet real businesses need videos that actually push commercial growth.
So Kuaizi solves the middle workflow. Before ChatGPT-class models arrived, we already built a “write–shoot–edit–publish–manage” (编拍剪投管) workflow to connect those roles so content teams can produce — especially video — better and faster to win customers, not only to cut cost.
I have long argued that the last “C” in AIGC should be conversion — AI Generated Conversion: AI should not only generate content but help you convert. Internally we say: producing content for the client is not the goal; the content must help sell.
Q: As an AI founder, does Kuaizi’s business feel hard? Is the pressure heavy?
Wanfeng Chen: Straight to the soul. Entrepreneurship is heavy work — everyone feels “heavy.” It is not only AI founders carrying weight; every sector in China is carrying weight today.
I see three “heavies” in the current environment:
- Hot money is scarce in AI. Versus the U.S., Chinese AI startups struggle to raise large capital — and thus struggle to find paths to revenue.
- Macro is soft. Everyone says business is hard, but you can still dig for upside and hope inside that softness.
- Cash flow. Brand/buyer budgets are constrained; market liquidity is tight. Founders must find solutions.
Q: Clients care about “cut cost, raise efficiency” — which do they care about more?
Wanfeng Chen: Both. There was also the debate whether China needs SaaS. China cannot not need it. When GDP is not rising as fast, you need technology; enterprise digitization will not reverse. Saving money is only part of the calculus. I often say China does not need traditional management SaaS so much as technology that lifts productivity. AI’s main job is productivity — many chances to raise it. Enterprises face pressure to scale without adding headcount.
Q: Will productivity lifts substitute tasks, jobs, even people? When?
Wanfeng Chen: Choosing not to hire more is not necessarily direct substitution. It can mean seeking more efficient ops — lifting existing staff via tech. That may affect new hiring, but it can also be a strategic response to markets and competition.
We see many clients who started small and quickly saw how AI tools raise efficiency. One client adopted our product in early 2022 and saw about a 10× efficiency lift. That seems to confirm a key insight: if AI cannot deliver at least 10× growth or efficiency for the customer, it is not valuable enough.
In today’s market, value must be highly visible. Clients who used our product to lift returns are the ones pushing the market forward. In a complex economy, AI demand is more urgent — especially into late 2023 Q3–Q4 we expected demand to rise sharply. Competition is fierce, but customers invest only when value is obvious. That is how we create revenue, ease information-flow frictions, and keep growing toward larger returns.
Q: Any darkest moment in the past year?
Wanfeng Chen: Through the pandemic our business kept growing — no drop. Looking from late 2022 into early 2023, we expected that to continue. As a young startup, early layout in AIGC gave us a head start — training new teams and getting existing teams deep into the domain.
We created an AIGC core group to catch image-generation advances, especially diffusion models that emerged mid-2022. To win in AI you first need internal cognition and acceptance of AIGC. With AI racing ahead we have reached a new milestone: many companies, through a shift in mindset, have entered the AI era. For each of us that is a turning point.
We urge the team to keep learning and adapting, and to explain why this environment differs from the past. Falling talent costs may bring challenges; we care more about finding PMF — which customers will pay, and whether they get real value. That is a long journey, not a short task.
Successful AI companies often need about two years of iteration before results show; those with clear wins usually have two to three years of history and deep users/data in a niche. Every strong player we see followed that pattern. We expect to keep growing and innovating in this space.
Q: Have investors moved from not understanding AI to understanding it — and become more demanding and pragmatic?
Wanfeng Chen: The investor trend is clear: they want certainty, especially as AI penetrates industries. Enthusiasm is high, but they look for reliable signals — model effectiveness, PMF, revenue, growth — even for early-stage companies.
Investors themselves face challenges; caution answers market uncertainty. Used well, AI can cut cost and grow profit — that is real value — and that capability is especially precious in a fast-changing market.
Q: With many peers, investors and customers compare. What are Kuaizi’s strengths and weaknesses?
Wanfeng Chen: This year’s strategy is balanced development across the team so we meet investors’ demand for “no soft underbelly.” We have gaps and work to close them. We understand commercial ops deeply in our domain.
With my brand background I know brand ops end to end — content distribution and upstream/downstream shifts. Even when the work is not pure AIGC, sales model and service quality remain core. However advanced the tech, value lands through sales, pricing, charging, and service — and we keep lifting team skill.
Our main limit is scarce resources, especially compute spend. Even so we can meet current demand, and customers still rate our value highly. We hope to invest more later; that does not block today’s growth.
Our investment approach is lean at the base layer — it looks thrifty but has already supported self-sufficiency and growth. In a hard environment that discipline has been key to rapid progress over the past two to three years — possibly the fastest growth stretch of my founder career.
Q: A year ago, why go all-in on AI? How did you communicate the AIGC acceleration plan — and what belief underpinned it?
Wanfeng Chen: Scene fit and PMF decide survival. Do them well and you compete; ignore them and you may not live. As enterprise services hit an inflection in 2023, I believed that even if our products still had years of growth left, without AI they would eventually not stand. Many CEOs share that view: without AI fusion, the model risks obsolescence.
The key is finding the right application scene. Missing it brings anxiety; finding it creates unique value. When angel customers use our tools to solve real problems, we know the direction is right.
AI gives rich, complex information; what matters is which specific problem we can solve, and whether we can deliver significant value at lower cost. We can adjust gross margin; customers care about the huge value they receive. That is the commercial truth to remember in the AI era.
3 CN images (event); attach on ingest. Full Q&A preserved.
— END —