China’s EV Pioneers: What China’s EV revolution looks like through the lens of XPeng’s founder (part 2)
He Xiaopeng on China’s EV future, design and autonomy, and his bet on flying cars
Today’s edition features the second half of the 10 highlights I selected from a three-hour podcast conversation between XPeng CEO He Xiaopeng and Luo Yonghao — one of China’s best-known entrepreneurs-turned-internet personalities.
In the previous piece, He looked back on his early entrepreneurial journey and shared why he returned to build an EV company after financial freedom. He also reflected on XPeng’s brand positioning and the staggering scale of R&D investment he believes an AI-driven car company must commit to.
This time, the discussion turns sharper. He lays out why he expects China’s EV market to shift within five years from a knockout stage to an all-star game with only about five serious players left. He shares his perspective on automotive industrial design, offers a comparison of how autonomous driving may evolve differently in China and abroad, and describes his exploration into flying cars — from early trials to plans for mass production. He also delivers candid views on which Chinese companies are genuinely committed to AI and which are merely paying lip service.
Once again, while I’ve distilled the conversation into highlights, the full three-hour dialogue is packed with insights. If you understand Chinese, I strongly encourage you to watch or listen to the entire podcast — it’s well worth your time.
In the past two weeks, Luo has put out two more episodes. I had expected him to keep bringing on guests from China EV industry, but instead his latest conversations featured Chinese comedy sensation He Guangzhi and Renke, frontman of China’s much-loved folk rock band Wutiaoren.
I haven’t yet decided whether to continue this newsletter series or in what form to present it. Personally, I really enjoy listening to Luo’s podcast, but producing this series takes quite a bit of time—his talk with Renke alone ran nearly five hours.
If you’d like me to keep covering this series, please take a moment to vote in the poll.
About He Xiaopeng
Prior to serving as chairman and chief executive officer of XPeng, Mr. He served at Alibaba Group Holding Limited, a public company listed on the NYSE (symbol: BABA) and the Hong Kong Stock Exchange (stock code: 9988), from June 2014 to August 2017, including serving as the president of Alibaba mobile business group, chairman of Alibaba Games and president of Tudou.com. In 2004, Mr. He co-founded UCWeb Inc., a Chinese mobile internet company that provides mobile internet software technology and services, and served as the president of product from January 2005 to June 2014. In June 2014, UCWeb Inc. was acquired by Alibaba Group Holding Limited. Mr. He previously served as an independent director and a member of the audit committee of HUYA Inc., a game live streaming platform company in China listed on the NYSE (symbol: HUYA) from May 2018 to May 2020.
Mr. He received his bachelor’s degree in computer science from South China University of Technology in July 1999. Mr. He obtained the qualification certificate of senior economist (technology entrepreneur) in business administration issued by the Human Resources and Social Security Department of Guangdong Province in January 2020.
6. He Xiaopeng's analysis on the future landscape of China’s EV market.
Luo: You’ve often compared the auto industry to a knockout stage and an all-star game. In your view, when do you think can we finally see the winners and losers of this game?
He: In China, about five years.
Luo: Who's going to survive?
He: Different companies may have different perspectives. Personally, I lean on the aggressive side. I'd say in China, we’ll probably end up with around five major players. And of course, a few overseas automakers will still last too.
Luo: Which one is definitely going to make it?
He: No one can confidently say it has the ticket. For the industry, it’s both a challenge and an opportunity.
Luo: Globally, Tesla is definitely safe, isn't it?
He: If you’d asked me last year, I’d probably agree with you. But this year, even Tesla’s facing new challenges. In China’s auto industry, the top-ranked new players tend to come across turbulence within 12 to 24 months. For the big, legacy giants, the cycle comes around four to five years. So, honestly, it still feels way too early to make a definitive call today.
Luo: So on the one hand, you’re saying the knockout stage of China’s auto market will settle within five years. But on the other hand, you’re saying not a single company has secured a ticket yet. That makes the whole competition sound even more brutal.
He: But also more interesting.
Luo: Indeed. And XPeng has been around for almost 10 years now, right?
He: 11 years.
Luo: Looking at where we are today, how would you rate your domestic competitors? Let's focus on the positive side, after all we’re not the type of podcast to stir the pot. Maybe just tell us which companies you personally respect or admire, and highlight some positives you see in them.
He: I’d say every automaker has its own advantages. For example, Volkswage, our partner and also one of our investors. It's sure that, right now in China, a lot of companies are getting recognized by the media, by the industry, and by users. People think we’re doing great, and that we can be really strong on a global scale. But honestly, I feel we’re still pretty far from that. Ten years ago, we were a little too lacking in confidence, but now it feels like the pendulum has swung the other way. There’s maybe a bit too much confidence.
Luo: Maybe it's because our domestic new energy vehicle industry have taken off, so people are now swinging to the other extreme and become somehow overly confident.
He: Yes. Take Germany as an example. Can you name a Chinese car brand that's popular out there? Its market is open to us, unlike the U.S., so in theory there should be someone.
So every company’s got its strengths. People love to say things like, “Oh, Volkswagen or Toyota, their software sucks, their user experience isn’t great.” But compared to them, our scale is way smaller. Our ability in areas like management, globalization and building service networks are still in the cradle in their eyes. If you score it out of 100, we might feel like we’ve hit 60 in some areas, but for them, we’re more like a 20 or 30.
That’s why I keep saying no one has the ticket yet. Some companies look really strong, but they’ve still got gaps in key areas. Some others are dominating certain domain, but are too small in scale.
So relatively speaking, Tesla is one of the few companies that can be called "well-rounded", but even they’ve hit ups and downs. Chinese automakers all have their bright spots, but this first knockout round still has about five years to play out. After that, it’s the All-Star Game. Until then, everyone’s got plenty of homework to do.
Take Lei Jun's Xiaomi as anoither example. They've done an amazing job, but they’re also facing challenges. Therefore, the one ultimate principle for us is to move steadily, stay grounded, learn from others, and fix our own weaknesses. That’s the only way to become truly all-around. And honestly, we’re all still on that journey.
7. He Xiaopeng compared the future prospects of autonomous driving in China and the world.
He: I believe in the future, there will only be two kinds of cars. The first kind has a driver, but also has L4 autonomous driving ability, so you could be driving thousands or even ten thousand kilometers ,before you actually need to take control of the steering wheel once.
The second kind won't need a driver’s seat anymore. It drives itself, and any possible override would be operated from distance or on the cloud.
So down the line, cars will split into the man-driven type with L4 and the driverless type also with L4. And I think that shift is going to play out in the next two to five years.
Luo: If fast enough, maybe in two years?
He: Technically, yes. But when it comes to user adoption, I'd say it would take 5 to 20 years. Every country has its own condition and relevant rules and regulations, so as an entrepreneur, the real challege is about figuring out the best development pace for the company.
People in many countries of the world don't even have any clue about autonomous driving. How can you talk them into an L4 self-driving car? So different situations will lead to different development paces.
But from another perspective, I believe that in China and the U.S., we’re probably only 2 to 5 years away from the major turning point, which is going to flip the entire logic of how cars work.
Luo: I’m actually pretty optimistic about China’s tech policy. The government has been pushing the sector strongly, and that makes a big difference.
At first, I used to think it’d be some developed country that would go L4 first, and others would follow after technologies are proved and accident rates drop. But now I feel that the Chinese government might actually be the most supportive one toward L4.
What’s the sense you’re getting from the industry?
He: I have a bit of a voice on this. Over the past couple of years, I’ve been to a number of countries and sat down with regulators to talk about policy and legal changes in transportation. And honestly, what stands out to me is that China’s regulators are both able to keep things under control and, at the same time, among the most aggressive globally in appling new technologies. That’s why I’m confident in achieveing L4 in China.
When I was young, I always thought developed countries had the most open-minded, fast-moving policies to drive technological innovations. But in recent years, I’ve found the reality is the complete opposite. Private companies are not necessarily leading the way in driving technological progress, and governments in advanced economies are also not necessarily be open and eager to push innovation.
That's one of the reasons I think that China’s economy has been able to grow fast. If you go out and compare, you realize our regulators are far ahead of many countries. And that’s something I feel really proud of.
Luo: Especially in promoting technological innovations.
He: I really believe that in China, we're going to achieve L4 in cars both with and without drivers, and be ahead of the whole world in this field.
Luo: That's a confident claim.
He: I'm absolutely confident.
Luo: And it not only owes to our supply chain, but also the whole talent pool of engineers in China. The U.S. has been talking about bringing manufacturing back for a long while, but they can't fill the current engineering talent gap.
In China, making hardware is so convenient, especially in Shenzhen. Any problems with hardware can be solved in a half-day drive in this city. Of course, auto-making is too complecated to be done in such a short period of time. But other hardwards are just a stone's throw.
He: Not just the supply chain. Take Mexico for another example. Suppose we're setting up a factory out there, what do you think is the biggest challenge? Electricity shortage.
We've been doing business in China for so many years, but electricity is something we'd never even thought about.
But in Mexico, when I was there trying to invest in a factory, people there would simply tell me "Sorry, we don't have enough electricity here, and you'll need a lot of money or pull some strings to get your factory powered."
Luo: Yeah, that's something you probably only heard about in China like decades ago.
He: At least in the 90s'. But over the past 30 years we've never heard of anything like this in China.
Yet, many countries in the world are still bothered by this issue. That's why I believe that the high-level infrastructure foundation of China, including things like stable water and electricity supply, broad roads, solid communication network, drainage system and paved lands, makes the fundamental competitiveness of China's supply chain.
China has gone far ahead than other countries in infrastructure, so I firmly beliece that China can get even stronger in the next one and two decades.
Luo: A lot of foreign companies keep talking about moving their supply chains to this or that country, but honestly it’s actually really difficult. It’s not something you can just pull off overnight. China’s well-developed supply chain is the result of decades of accumulation.
He: Yes. And I think a lot of companies that do end up moving out of China will still remain on China's supply chain, which is eventually make our global supply chain even stronger. And I'll say the whole process of transfer may take 10 to 20 years.
8. He Xiaopeng talked about his experience in developing flying cars in China
He: Developing flying cars is a tougher and longer journey, mainly because building an aircraft itself takes a really long cycle, and honestly, its commercial viability hasn’t really been proven yet.
But I’m still confident that in the next era, whether it’s in 3 years, 5 years, or maybe a bit longer, we’re gonna see a major breakthrough in the flying car industry.
Luo: So when can we expect to see your first model of flying car in the market?
He: In the second half of 2026.
Luo: Is our government supporting this sector?
He: Yes it is. China has been strongly encouraging low-altitude economy since 2023. Policymakers have released at least six or seven documents mentioning or emphasizing the development of this sector.
More interestingly, Trump has also said recently that the U.S. needs to get into the low-altitude economy game too, because of China’s influence in this space.
Luo: I've heard that the U.S. also has similar teams.
He: Yes, there are two two publicly traded companies in the U.S. working in this area. But they were originally focused more on electric versions of helicopters to tackle urban traffic issues.
The aviation policies in the U.S. is pretty open, so I believe both China and the U.S. have massive opportunities in the low-altitude economy
Luo: Could you arrange a flight for me of your model if I come to Guangzhou right now?
He: Absolutely.
Luo: How many times have you taken your low-altitude model？
He: Not much, less than 10 times. But we run test flights around 10 thousand times per month.
Luo: Is the test zone still that little test shed wrapped with cloth that you mentioned before.
He: No, in open areas.
Luo: Because of the policy shift?
He: We apply for thousands of test flights in designated areas every day.
Luo: So I can try it out anytime I come to Guangzhou?
He: Definitely OK. I suggest you come over in the first quarter of next year, because starting from this fourth quarter I’ll be personally piloting our new model. So yeah, next year is the perfect time for you to come and experience it.
Luo: You’re flying it yourself? Don’t you need full-on pilot training for this kind of aircraft? Okay, let me throw in a serious question here — throughout your whole low-altitude aircraft development, up to this point, has anyone ever lost their life during the experiments?
He: Not once. But our founder's crashed three times.
Luo: That's inevitable. But not a single casualty during the whole process?
He: Never. We’ve actually been kind of lucky. Our founder, Zhao Deli, has been building low-altitude aircraft for years. And yeah, he’s fallen out of the sky three times and even ended up in the hospital.
Luo: But honestly, without someone like him, this wouldn’t even be possible. This kind of personality is just perfect for the early stage of a startup that’s trying to make innovations.
He: So, if we really want our low-altitude aircraft to become as common as cars one day, we need to make them thousands of times safer than cars. Right now, our target is three thousand times. That means ultra-safe and totally controllable, which is already insanely hard to pull off. Of course, making flying vehicles mainstream is definitely going to take many years.
Luo: But you’ll already be commercializing and selling the very first batch of these aircraft next year.
He: Right.
Luo: I feel like buying one even just to support your idealist spirit. Your pricing isn’t going to be super expensive, right?
He: It should be around 1 million yuan.
9. He Xiaopeng's view on Chinese companies’ real vs. pretended commitment to AI
Luo: When it comes to AI, how exactly is XPeng setting itself apart from the other carmakers?
He: Honestly, I think most companies that say they are "working on AI", they’re not really serious about it.
Luo: It’s more like lip service.
He: Exactly. Like when I first got into this industry about ten years ago, everyone was saying, "Software will redefine the car." But honestly, I didn’t really see them doing anything groundbreaking with software.
And then you’d hear people saying, "Data is priceless." But most of them were simply storing the car data. So I asked, "What are you actually going to do with it?" And they’d say, "These data will be a goldmine someday."
To me, that just complete nonsense.
Luo: Are you referring to those poeple from traditional car companies?
He: To be fair, everyone makes similar mistakes, even us. For example, we've always been saying that designing and styling matters in carmaking. But then Xiaomi came out and gave the whole auto industry a pretty harsh lesson: you guys haven’t been taking designing seriously enough.
So what really matters isn't how a company talks, but what they've really done. What's matter is the time, resources, and processes a company puts in, and ultimately the results they deliver.
For example, at an AI-related conference inside the industry, I asked a room full of companies that how many of them are actually doing pre-training. Out of a few hundred entrepreneurs there, only three or four people raised their hands. That tells you everything.
Luo: So you mean, most of them are probably just doing post-training at best, or maybe not even that?
He: Honestly, I doubt they’ve even done post-training. Let me give you another example. This was a while back, when I was at a really important conference. Everyone there was talking about how amazing AI is. But then, a bunch of people came up to me and told me that they've only heard the buzzwords, but have never actually used AI themselves.
Luo: Are you talking about tech companies?
He: No, they're not.
Luo: Then you mean traditional industries. Somehow I can understand that.
He: So, when we talk about really taking AI seriously, it means actually investing in it. For example, in China, there are tons of cloud computing companies, and the top ones are massive. But if you ask around about how many of them are really building strong AI capabilities on their cloud platforms, I’d say not many at all, and you can easily prove that through their cloud services.
If you're not even applying AI in your business, how can you claim you’re taking AI seriously? That’s why, from my perspective, a lot of companies today are just paying lip service when they talk about AI.
Luo: Also in the new energy vehicle industry?
He: In this industry, among the ones working on autonomous driving, the number of companies actually using massive computing power is just a handful.
Luo: Honestly, you can tell right away if a company is really serious about AI just by looking at how many chips they’ve actually bought for AI training.
He: And the truth is, not many companies have actually bought a lot.
10. He on what really matters in EV design
Luo: How big is your industrial design team now?
He: We’re at around 400 people this year. It’s gotten pretty big.
Luo: That's a lot indeed.
He: Most of them are in Shanghai. People in the industry often say we have technology as the left hand and art as the right. The year before last, I realized that, in reality, the tech side gets hundreds of times more investment than the art side. So clearly, the two "fists" aren’t equally strong.
Luo: When you’re making mass-market products, aesthetics still matter a lot, especially with cars.
He: At the end of the year before last, we finally made up our mind to really invest big in aesthetics, and last year we made a ton of adjustments. Out of those 400 people, a lot are from overseas -- some are Chinese, but plenty are foreigners too.
Luo: You have foreign designers too?
He: Quite a lot. Our head of design is from Europe. I think we need a fusion of East and West, especially since we’re selling a global car. In China, if you’re doing hardware plus software, there are two areas that must become your long suits -- industrial design and AI capabilities. Both have to be strong.
But coupling the two is hard: most hardware companies don’t understand software, let alone AI. And AI companies don’t understand hardware, which involves so much, say design, quality, supply chain, manufacturing, and more. So excelling in both areas is really tough. You know what I mean. You’ve done it before.
Luo: Yeah, right. We’ve done a lot of industrial design. I imagine inside your company, the design team drives most of the decisions, right? Cars are huge projects, so for every model you must go through piles of drafts -- dozens, maybe even hundreds -- constantly refining.
When it comes to the final decision, who gets the say? Your head of design, you, or do you guys actually vote on it?
He: At first, I was the one making the final call. But that caused a lot of issues. If I wasn’t professional enough, or didn’t have a good eye, or only noticed a few things when a car actually has thousands of design details, it was easy to make mistakes. Now it’s more balanced, and we’ve added way more user feedback.
Another thing is we make sure there’s a product manager in charge -- someone who keeps an eye on the big picture. And on top of that, designers need real resources and authority to actually do their job. One more big change is how we handle revisions. Before, when a project started, we’d give designers a bunch of constraints and say, "Here are the rules, now design within them." Now we’ve flipped it: we start with how to make the design great first, and then adjust the constraints to fit it. With this new approach, I’m pretty confident that starting with the XPeng P7 and moving forward, our industrial design capabilities are expected to take a big leap in the next few years.
 | A guest post by
|