A VC firm from the Crypto world says AI is too crazy; they are very conservative.

null

Text: elsewhere

Many people say that the AI industry, in some aspects, is becoming more and more like the Crypto industry.

Jingwen from Impa Ventures, someone who once worked in the Crypto industry, says that compared to today’s AI, they even seem quite conservative.

Jingwen describes herself as a 93-year-old girl who aspires to VC but “mistakenly entered” Crypto, and accidentally earned her first big fortune.

During the peak of FBG Capital (a Crypto Fund), she experienced that accelerated crypto world. With just words, she could invest hundreds of thousands of dollars. A story could grow wildly, exploding with coins.

In 2024, she co-founded Impa Ventures with Shiran and James (Liang Jie). Shiran was Jingwen’s colleague at FBG, and James was an old colleague of Shiran at Huachuang Capital.

Jingwen says they are a Problem First fund—looking at the problem first, then the solution.

Impa Ventures currently has a $50 million scale, focusing on early-stage. So far, out of nine investments, eight are to B2B.

Three months ago, Liu Jing shared part of this story in “Chinese Investment Stories.” At that time, he talked with James about the inner world of a investor who missed Pinduoduo. We’ve placed this part at the end of this article to help you understand the team’s temperament of this new fund. Recently, we spoke again with Jingwen and James.

This is the fifth new fund story presented by “elsewhere,” following Nebulon Ventures, Source Rhythm, Creek Stone, and Little Fund.

AI seems to be playing the Crypto game

elsewhere: Before deciding to start an AI investment fund, you spent a long time investing in Crypto. Can you tell us the story back then?

Jingwen: In 2017, I graduated from the Schwarzman College. Coincidentally, my thesis was related to Crypto, so I talked to many people. Someone thought I was looking for a job and introduced me to Zhou Shuoji, the boss of FBG Capital.

At that time, it was really hard to find serious backgrounds in the crypto space. I was from NUS Computer Science and Schwarzman College, so I went for an interview.

The process was quite absurd: the boss was over an hour late, and when he sat down, he asked me what my offer salary was. I said 7,000 SGD, and he directly said, “I’ll double that.” I felt at that moment that this place was just full of people with money and no sense.

elsewhere: But that salary still tempted you.

Jingwen: The original offer was to start in September. I thought, since I was graduating in June and had nothing to do, I’d try here for three months and see what these people are really about.

It turned out to be peak wild growth. At that time, FBG was one of Asia’s top blockchain funds, with LPs like Sequoia and Ribbit Capital. The boss was very wealthy himself, so investments were very casual. My job was to fly globally, attend various events—Berlin, San Francisco, Tokyo, Singapore. Sometimes, I’d meet the boss in the corridor and tell him I liked a project—tens or hundreds of thousands, even millions of dollars—and in just two or three sentences, the money was invested.

elsewhere: How long did this peak last?

Jingwen: Only about half a year. By 2018, the bull market was ending, and 2019 fully entered a bear market. My work also changed 180 degrees: from disbursing money to asking for money. Instead of investing 1 million, I was trying to recover 500,000. That’s it. After that, I basically stopped investing.

elsewhere: Looking back now, what kind of world was Crypto?

Jingwen: Crypto is a world that has been accelerated. Essentially, it’s a process of a new technology emerging and landing. There are many similarities with what’s happening in AI now.

And another point similar to now: stacking teams, hyping narratives, manufacturing metrics, then exiting. In today’s AI game, the backers have shifted to institutions and LPs, but the game logic remains the same.

Talking about Crypto—I don’t think that world is gone. It’s moving toward a more solid direction: real applications, on-chain fundamentals, especially the integration with AI, which I believe has many truly promising directions.

We just don’t want to replicate that game in AI. Impa starts with our own money—we treat our own funds and external capital the same. Only then can we more purely seek that true Alpha.

We are a conservative AI fund

elsewhere: Tell us about your fund.

Jingwen: Starting in 2024, Shiran, James, and I launched this early-stage AI-focused fund. Shiran and James are colleagues from Huachuang, and I and Shiran are colleagues from FBG. We’ve already invested in 9 projects.

“Chinese founders + global markets” is a big opportunity given by the era, and we’re still in early stages. Day one global is our fund’s biggest feature: Shiran and I are based in Singapore, James is usually based in Shanghai; among our 9 investments, 3 are in Shenzhen, 2 in Singapore, 1 in the US, 1 in Sydney, 1 in Shanghai.

elsewhere: Why is it called Impa Ventures?

Jingwen: Impa is an NPC from “The Legend of Zelda.” When the protagonist starts the main quest, Impa provides maps and guidance. She’s a key part of the hero completing the mission, but she’s definitely not the hero. What we can do is, in their earliest stages, become a part of helping them get things done.

This understanding is actually the starting point of our entire investment methodology—since the founder is the hero, our job isn’t to bet on narratives but to find those who are truly solving problems.

elsewhere: As a new fund, many people probably ask: what’s your differentiation?

Jingwen: Our starting point for projects is different from most funds—we don’t first define a sector, but first identify the problem. So, what we ultimately invest in is naturally different.

We are true AI believers, believing AI is reshaping the fundamental infrastructure of many industries. This isn’t just narrative; it’s happening for real. But we also experienced the Crypto wave—seeing how bubbles grow and burst. So we understand the difference between waves and the ocean.

Believe in AI, but skeptical of many narratives under the AI banner.

elsewhere: How do you distinguish between waves and the ocean?

Jingwen: AI Believer, but Skeptical. Long-term optimistic about the industry, but skeptical about specific projects.

Judgment is based on a few things:

• Problem First, not Narrative First.

Start from the problem, not the sector. We don’t first set an “AI+X” theme and then look for projects; instead, we see a real, underestimated pain point and ask: who is solving this? Is the solution feasible? Founders who can answer these two questions are more valuable than any shiny background.

• Global by default, not Global by design.

When I was in university, I interned in Israel for half a year in VC, which was my first real understanding of how venture capital works. I also saw a group of founders who turned small local markets into global ambitions. Israelis naturally do business with a Day One global mindset. That experience deeply influenced me, and I decided to do VC from that point on.

Later, investing in Crypto, many projects I backed weren’t Chinese teams but truly global developer ecosystems—flying to Berlin, San Francisco, Sweden, and meeting many excellent developers, some of whom later transitioned into AI.

• Young but experienced.

In our team, James is the only one born in the 80s; the others are born in the 90s, 95, or 00s. No historical baggage, but all three partners have gone through complete cycles and have successful exits. Those who’ve seen bear markets judge “real needs” and “bubble narratives” differently.

elsewhere: In AI, what kind of game do you want to play?

Jingwen: Our methodology boils down to four words—“Respect Common Sense,” especially business common sense. Some might think we’re too conservative.

Now, many look at founders through a “colored bubble,” thinking that if they come from big companies or top universities, they must succeed. But what value do you create for customers? Can your productivity improvements truly cover costs?

Impa cares more about value creation.

elsewhere: What are some projects that are not sexy but make sense?

Jingwen: We invested in a project called “Lightyear Reach.” The founder met with over 100 investors, and everyone thought B2B was unsexy. But after talking with him, I found he addressed a real need in China’s supply chain going abroad: many small factories lack overseas marketing capabilities. He uses AI to help them automate customer acquisition and advertising. It’s a typical “distribution intelligence” process.

The model company creates intelligence, but how to sell that intelligence into specific scenarios is the opportunity for startups. These projects may not be fancy, but their business is solid.

In Australia, we also invested in a medical Admin AI project. It doesn’t touch diagnosis or treatment, only solves the most tedious administrative processes like appointments, triage, and summaries. Overseas, these processes were traditionally handled by manpower over the phone, but now AI can fully digitize the entire process.

We prefer projects that first solve existing, confirmed “pain points,” rather than fantasizing about vague needs.

elsewhere: VC is a business of chasing outsized returns. Can such investments be realized?

Jingwen: For example, we invested in a company making digital humans on the client side, betting on 3D real-time digital humans. The topic of digital humans isn’t new in capital markets, but our logic is simple: they offload rendering to the user’s device, only transmitting driving data via cloud, reducing costs by 99% compared to mainstream cloud solutions.

Our calculations show that serving a few hundred thousand overseas active users can generate tens of millions of dollars in ARR. The team has top scientists in 3D digital humans, aiming to achieve the same quality as Cai Haoyu’s Anuttacon LPM video models, but running on phones and browsers. The prospects for multimodal AI and social entertainment scenarios are huge.

Our core goal is to find the extremes in the Power Law—businesses that can grow to hundreds of millions of dollars, and only those, we will consider.

elsewhere: Your dreams are big…

Jingwen: When I was in college, I had a funny dream—I wanted to make the Midas List.

elsewhere: That’s a dream shared by many VC investors.

Jingwen: I knew about the Midas List when I was in college. It’s not that I thought being on it would make me particularly great. My idea was: early-stage investing is what I love, and I want to reach a milestone in what I love doing.

Looking back now, that so-called “dream” seems a bit silly, but often people don’t have goals. Having such a “funny” goal is better than having none.

A conversation with James three months ago

@Jing Liu

When I first started updating the “Chinese Investment Stories” series, I wrote about a person titled: “The Forgotten Person in the Pinduoduo Capital Myth.”

The story is roughly: a Sequoia investment manager met Huang Zheng early on, and later pushed Pinduoduo (then called Pinhui) onto IC multiple times. But due to various circumstances, it never passed. Until he left six months later, Sequoia invested. This was a key round in Pinduoduo’s history, which later became a legendary Sequoia investment.

Whether related to or unrelated to a billion-dollar return, it’s just that six-month gap.

This person is Liang Jie. Between 2012 and 2016, he worked at Sequoia, and it was during this period that he saw Huang Zheng and Pinduoduo.

Honestly, in the investment industry, such stories aren’t rare. As long as you haven’t become a partner—or even a managing partner (GP)—9 out of 10 investors have a pile of stories about long-standing struggles.

But Liang Jie is one of the most memorable people I’ve met. It even became a long-term narrative about him.

A few years ago, he came from Shanghai to Beijing on a business trip, and we arranged to meet at the Westin on Liangmaqiao. It was late at night when we met. I initially wanted to discuss industry topics, but somehow we quickly started talking about this story. He casually pulled out a pitch deck from the Pinhui era. His memory is so precise that he can recall every page’s layout and wording.

After leaving Sequoia, he went to two other funds, and later ran a fund focused on overseas markets. But those were just ordinary years.

Perhaps because I was also new to investing when I met Liang Jie, and among many legends, this reversed story stood out, leaving a deeper impression.

Before 2020, China’s VC industry was fully to B. Liang Jie still insisted on looking at to C and platforms. I remember he said: if in China, only to B opportunities remain, he’d stop. That’s not his conviction.

elsewhere: I didn’t tell you in advance I’d write about you, and this isn’t exactly a “good story.” When you see this, what’s your feeling?

Liang Jie: A bit surprised. But I don’t want to be labeled as “unlucky.”

First, I don’t think luck explains everything. It’s definitely because my own accumulation isn’t enough—like handling relationships and matters, or my conviction isn’t strong enough; secondly, luck is important, but it also requires accumulation. For those at the table, if you don’t play, you’ll never have a chance.

elsewhere: If you could go back ten years, would you be more confident in pushing Pinduoduo (Pinhui) onto IC?

Liang Jie: Definitely. Compared to ten years ago, I’ve made some progress in both business understanding and how to push things forward.

elsewhere: Someone in the comments said: since you believe so much, why didn’t you buy Pinduoduo stock in the secondary market later?

Liang Jie: In 2015, investing in an early-stage company and buying stock after IPO in 2018 are two completely different timeframes and things.

The more appropriate question might be: if I was so optimistic, why didn’t I join Pinduoduo or invest a bit myself back then?

Honestly, I didn’t think about it at the time. The $600 million valuation and the cash I had at hand didn’t seem to match. But that also shows I didn’t really understand—didn’t think I could make a hundred-billion-dollar business. After leaving Sequoia, I also talked with Colin (Huang Zheng) about other possibilities, like introducing LPs or joining his fund.

elsewhere: Do you regret leaving Sequoia? Maybe staying half a year or a year longer would have made it a story about you.

Liang Jie: In mid-2015, a major turning point happened, and that was one of the reasons I left.

That August, at the Sequoia offsite, the consensus was: the focus of venture shifted from to C to to B, and vertical industries like education and healthcare. I was not very willing.

To be fair, during my push, Neil Shen’s response was still positive. But honestly, did I have conviction that Pinduoduo could succeed? Not really. I just thought the story was big, the growth was fast, and the founder was very strong.

elsewhere: How long did it take you to calm down?

Liang Jie: After the subsequent funds, I was actually fine. Every stage has more important themes.

Luo Xiang once said: after he became famous, he received a lot of criticism, which made him uncomfortable. A friend asked: when you became famous and received many undeserved awards, did you feel ashamed? He said no. Then why accept undeserved praise happily, but not undeserved criticism?

I think in this industry—or any industry—top performers are a small group who are both smart, diligent, and lucky; if we think we’re not stupid and can still do things, and compare ourselves to those extremely lucky few, that’s too greedy.

elsewhere: The story of missing the IC stage is common in VC. Why is your story so memorable (besides the fact I wrote about it!)?

Liang Jie: Maybe because later, the scale of Pinduoduo made it unavoidable. Like dating a girl who didn’t work out, but later became a superstar…

elsewhere: I remember you studied engineering. How did you get into VC?

Liang Jie: I did my bachelor’s in materials, then a master’s in microelectronics. During my master’s, I read “The Big Game,” which opened a window for me as a science and engineering student: capital markets are so fascinating and important. A seed of VC started to sprout.

In 2006, after graduating with my master’s, I desperately wanted to enter VC. But I had no chance. In 2008, during the financial crisis, a Swiss fund called Adveq interviewed me five rounds and almost offered me a position. Their founder visited Shanghai once, and I still vividly remember our chat at Jinmao Hyatt.

Later, I spent $499 on a financial modeling course, because many doubted I lacked a finance background and didn’t know how to build models. I also sent over 100 resumes to Chinese VC firms. Finally, in 2011, I got my only offer—Huadong International.

elsewhere: Then came Sequoia.

Liang Jie: Yes. Of course, after leaving, I realized even more that Sequoia was already the center of the universe.

elsewhere: How have these years been?

Liang Jie: I started my own fund, “Skyline Ventures,” but it didn’t scale up (no institutional funding)—disappointed, but had to accept it—AI arrived, but I felt it was hard to participate—met like-minded partners and returned to the game—discovered many opportunities and few participants—grateful and hopeful.

elsewhere: Do you sometimes feel that luck was just a bit short?

Liang Jie: Many smarter and more capable people have left this industry. I’m still involved in this wave of technology, and I feel very lucky.

Early-stage investing has a long feedback cycle, and a single outstanding success can overshadow many flaws. Everyone hopes for that small probability of good luck, which is irrational and unhealthy.

Last year, I went to Hangzhou’s Dali Flower to watch the China vs. Australia game, lost 0:2. I met a TV station interviewer afterward. The reporter asked: China lost, why are you still smiling? I said: the Chinese players on the field today didn’t play particularly badly; they all performed well, especially Wang Yudong. Australia was obviously stronger. Losing 0:2, I have nothing unacceptable.

It’s like a kid’s exam—if they usually score 70, and this time 75, what’s there to be unhappy about?

elsewhere: Once he scores 75, you still want him to try for 90, right?

Liang Jie: Of course, I’d be happier if he scored 90. But that expectation shouldn’t be there.

elsewhere: As an investor, can you try to evaluate yourself?

Liang Jie: You can compare it to playing football. Playing football is something I can get into a flow state very easily. Even when I don’t play well, I never stop. I started playing street football in a small town in middle school, and I’ve kept at it ever since.

In my team after graduation, I was a main player (as long as I showed up regularly, I was basically a starter), but I was never the one deciding the game. But over time, I realized that in recent years, I’ve gradually become someone who can influence the game.

Why? First, I have a competitive spirit; second, even as others decline sharply, my body can maintain or even improve; then, my understanding of football deepened.

Analogous to investing, I think I can keep doing this. I missed the chance to become an overnight success, but I can persist until the end and influence the game.

elsewhere: Finally, I want to ask: are you aiming for 75 or 90 this time?

Liang Jie: Most things in life are beyond our control: our birth, IQ, opportunities. If we achieve something, we should thank factors outside ourselves. Talent is given by heaven, opportunities by the era—nothing to force. Only gratitude and being true to oneself.