How to Actually Read the China's AI Ecosystem with Jing Yang

**Jing Yang** (0:00)
This is one of the things that really like surprised me when I was working on this story. DeepSeek had spent a lot of time last year retrofitting their models and software with Huawei chips. I think a lot of people assumed it's because the government ordered DeepSeek to work with Huawei to embrace the domestic ecosystem, but it's actually the opposite. It was DeepSeek that voluntarily started using Huawei, experimenting with Huawei chips. And Huawei actually only found out after, then they started sending people to DeepSeek to help them.

**Bernard Leong** (0:34)
Welcome to Analyse Podcast, the premier podcast dedicated to dissecting the pulse of business, technology and media globally.
I am Bernard Leong, and the last time we spoke, the story was Chinese AI startups going global. Since then, the map has shifted. DeepSeek has raised more than 7.4 billion in its first external funding round, valuing the company at more than 50 billion. Meanwhile, Junyang Lin, the former lead researcher on Alibaba's Qwenn models, has launched a new AI lab that has reportedly completed its first fundraising round at a post-money valuation of around 2 billion USD. These stories raise a bigger question. How is China building their frontier AI under cheap constraints, state capital, open source ambition, and scarce elite talent? With me today is Jing Yang, Asia Bureau Chief at The Information, who has been reporting these stories from Hong Kong, and I think she probably go into China all the time. So, Jing Yang, welcome back.

**Jing Yang** (1:35)
Thank you for having me again. Just for the purpose of legal compliance, no, I don't go to China all the time.

**Bernard Leong** (1:42)
Really? Okay, I'm just joking. All right. Since we last spoken, we were in studio together.
What have you been up to since we last spoke?

**Jing Yang** (1:52)
I was actually reviewing our last conversation, that was in September, I believe, last year, and so much has changed. The one thing that I think, not really a correction, but if you recall, we talked about how we don't know what ByteDance is up to and how they stopped participating in the AI model leaderboard, and then they also haven't had at the time open source any of their big models. And it's a bit of a mystery. But then now we all know that we see dance making waves and controversies. We know what they're doing. So yeah, I guess this is a way of saying, I mean, we all know that we develop AI in today's age. Everything is developing and iterating in such an unprecedented pace. I definitely struggle to follow and make sense of things. And I don't know about you.
If you have any clips, I'll ask you to hear about them.

**Bernard Leong** (2:48)
I feel like every few days, there's a change. And I am not surprised anymore by any changes. In fact, when Seedance came out until today, I think we can't even use API on Seedance. You actually have to basically subscribe to their models as a consumer user. And I think you have another model called Lumia. But if you look at the leaderboard, Seedance is now number one, followed by Google's Yo 3.5 models now, I think, if I'm not wrong.
So I think this whole space is really moving so fast. I used to think that China AI is like every once a time they make an announcement, it will shock everybody. So that's why I got you here today. To help our audience to understand how China's ecosystem is developing on its own terms, not just simply as a reaction to the US, but as a distinct model of frontier AI development, shaped by their talent, open-weight models, chips, capital and developer adoption. And of course, you know, UBTech is going to launch their first humanoid robots. And we know this is where it is all going towards. So I want to first start off having this conversation about DeepSeek because this is the most interesting company ever.
So earlier, it was seen as a capital-efficient research outlier, a lab funded by Liang Wanfeng and HighFlyer, known for pushing frontier performance with unusually efficient model design. And usually when people ask me, where should I learn all the tricks to do open-source AI? I think you should just go to DeepSeek's GitHub repository. Actually, they just let out so many things that people just don't know. So now the report is that DeepSeek has raised more than 7.4 billion. It is its first external financing round, valuing the company at more than US $50 billion. And making it now, I think, China's most valuable AI startup.
If I'm correct, maybe something has changed today. So how should we understand DeepSeek's evolution from a highly efficient research lab into now one of the most closely watched AI companies in China and globally?