So my next guest we also know reasonably well, and we know this space extraordinarily well. It's very close to home in that it deals with the auto and the vehicle space. We're here in Turin, which is the home of Fiat and the Lingotto, and also the namesake to our fund originally. And so as we think about what's the most physical thing in the physical world, autonomous vehicles certainly meet that. James Peng is the CEO of Pony AI, which is based in China, and has done incredibly impressive things. Let's welcome James to the stage.
Thanks for joining us. Sure. Before we go into anything, I feel it's likely that a lot of people in the audience don't know much about China's development in autonomous vehicles and what Pony AI has created. And what you guys have done is so incredibly impressive in terms of the technological advances and then also the cost base versus the numbers that you hear from Waymo. Maybe you can set the tone for everyone, because it's always so astounding to me a little bit, about your background and what Pony AI has built so far.
Sure. It's great to be here. I'm James Peng. I'm the founder and the CEO of Pony AI. So Pony AI was founded 10 years ago with a goal to build fully driverless applications. Fully driverless means there will be no driver in the vehicle. So it will be much cheaper, safer, and make our mobility systems much more efficient. Over the last 10 years, of course, we have built the autonomous driving systems from hardware to software to the applications, and currently we already have fully driverless operations in the four tier-one cities in China, and we are definitely going abroad as well.
And just for a number of clarity, what is the cost of build of materials for the average Pony AI latest?
The latest version is currently, I think, roughly 35 to 40,000 US dollars, give or take, depending on the slight variation. So it's like a regular car.
35 to 40,000 dollars. That is an incredibly impressive number that everyone should hold on to a little bit. More personally, you've been doing this for 20 years. What led you down this path? What gave you the confidence that this was something worth pursuing, was solvable, that you could pursue it? And then why did you start Pony AI specifically?
Well, I think for autonomous driving itself, it has huge societal and financial benefits, right? Essentially, think about the whole mobility industry is actually inefficient. Each car has a driver itself, and also there's hundreds of thousands of people who die unfortunately because of traffic accidents. I think AI, specifically autonomous driving, will be a solution for this. And also think about the mobility industry itself is a trillion-dollar industry, right? Any improvement, anything happens, it's a huge benefit. And also from a technical point of view, I'm a software engineer by training. I think for autonomous driving itself, it has hardware, software, hardware-software integration, it has user experience, it has vehicle self-mechanical heat dissipation or whatever. It's a combination of so many technologies. So as an engineer by training, it's so fascinating. So it's a great area to work on.
And maybe talk a little bit about the arc of your confidence and the development over the last 20 years. I mean, the DARPA challenge is 2004. As a tech investor, I love to hear that something deep tech has been in the works for over 20 years, because it normally looks like this until all of a sudden it hits sort of this breaking point. And so I'm curious from your perspective what that journey of confidence was like, and when you felt like you kind of hit that breaking point and your confidence started to skyrocket.
No, I think very early on, thanks for mentioning the DARPA challenge and the subsequent development. I think by and large, driving itself in the physical AI world is relatively confined because the roads and the traffic rules are defined by humans. So although there's a lot of unforeseen events happening on the road, by and large it's relatively regulated. So I mean, even 10, 15 years ago, although we didn't know if it's solvable, at least from a high level it looks like it's solvable. I think, of course, seeing is believing. I think three years ago, 2023, early January 2023, is when we actually had fully driverless operation on the open roads in an area that's hundreds of square kilometers. So that's like a real application. Although it's not everywhere yet, but I think that's when I truly believed this is actually solvable from both a technical point of view and also from a regulatory point of view. And afterwards, we started taking real passengers. On average, our vehicles have about 20 to 25 rides per vehicle per day. So that shows it's a real application that really gives the confidence. So I do believe that the whole industry, the inflection point has already happened.
Wow. Okay, that's great. Well, talking about that scale, great question. What is your vision for Pony AI specifically in terms of scale and how big you think, and what this means for the company over the next five or 10 years?
No, as I already mentioned, now we already have real-world applications for real passengers. We're taking tens of thousands of rides per day. So I think the next challenge is really how do we scale to many more cities. In terms of scaling, of course, it's not just a technical challenge, it's also regulatory, it's also user acceptance, branding, also ecosystem partnership. So it's so many moving pieces. The key is to get everything to fit nicely so that we can scale into more cities. Our plan for the next five, 10 years, I think as you mentioned, the biggest challenge or the number one goal is to scale, deploy as many vehicles on the road, and it will have a self-fulfilling effect because as more people see how safe, reliable, enjoyable the ride is, more people will take it, and regulators will then approve the application, and that will be everywhere.
Well, that's another great point that you raised, which is international expansion, primarily to Europe and to the Middle East. Obviously there are the regulatory challenges. Why are they so difficult? How do you overcome them? And is there also a tipping point for regulators where this becomes just sort of the norm in terms of acceptance?
No, I think first and foremost, the regulation is certainly not there. It's always hard to get anything brand new into the marketplace. Also because mobility is a highly regulated industry and is directly related with the safety of the passengers and also the road participants. So the impact of any potential accident is so huge that any regulator will take caution into approving it. So for us, the criteria to decide which market to go into, there are several criteria. One is of course the market itself is relatively big in terms of the size, and also in terms of the average taxi fare, because our pricing will always use taxi as a benchmark. And another thing of course is the supportive regulatory environment and also a good partner in the local market. Europe of course is one of the biggest markets that we are looking at. But of course, Europe is not one single market. We're looking at different cities. Currently we are already operating in Zagreb, Croatia. We're in Luxembourg. We're actually today just announced we'll partner with Uber to enter into London, and of course some other cities as well. So stay tuned, we'll be on the roads of Europe hopefully very soon.
