China's Largest AI Hackathon and Shift in Global Talent Flow
Hangzhou’s AdventureX: A Glimpse into the Future of China’s AI Talent
Hi All, I'd like to start this post to discuss the perception of Chinese AI talent.
Thoughts on Chinese AI Talent
A piece published in the widely followed WeChat blog “The Intellectual” (知识分子) prompted me to reflect on the perception of Chinese AI talent. The key points the author made include:
DeepSeek’s research team is relatively young and inexperienced in external views. Still, they have achieved significant academic milestones, with numerous published papers, high citation counts, and a solid h-index value, demonstrating consistent research output across the team. The core 31 authors demonstrate even higher academic metrics, not driven by a few stars, but by a broadly capable group with many having h-indexes above 10.
Over half of DeepSeek’s authors were educated and employed entirely in China without overseas study, showing intense domestic talent cultivation mainly centered on institutions like the Chinese Academy of Sciences and top Chinese universities.
As many have cited that over ~40% of AI researchers globally are Chinese, that definition is still vague. However, this author notes that many Chinese AI researchers now have US education or research experience, but often return quickly to China to work, indicating a reverse talent flow that differs from previous U.S.-dominant patterns.
Resulting in his key point that there is a shift in global AI talent flow. Unlike earlier trends where Chinese (internet and) AI talent moved to and stayed in the U.S., many are now returning to contribute to China’s AI ecosystem, signaling China’s rise in AI innovation and talent retention. And this is a point I’ve made here to the leading Indian tech publication AIM as well.
“(For the brightest) Working for big tech in the U.S. no longer holds the same appeal it once did. Companies like Alibaba or DeepSeek now offer salaries (we’re talking millions and tens of millions of USD) comparable to OpenAI, making it far less attractive to abandon family, friends, and cultural roots,” Shao explained.
Look, even I held ignorant views myself before moving to China and living in this part of the world for over a decade now. However, for the generations of the 80s, 90s, and even the 00s, cheaper labor is not a competitive advantage. (Heck, they’re not cheap.)
Anyway, it’s no news that ByteDance, Tencent, and Alibaba heavily recruited from Silicon Valley and then from each other and other Chinese research labs, shelling out big bucks - not the 100 million USD cheques Mark Zuckerberg is writing, but sizable ones like ~10-50 million USD for the top top top talent.
Unlike these big tech companies, DeepSeek and some other labs have committed to primarily hiring from leading domestic academic institutions directly, wanting them to be green and uncorrupted. Amid all this, the AWS AI lab in Shanghai was dismantled last week, and its leader, Wang Minjie, posted on WeChat that, due to geopolitical reasons, the research lab had been shut down. Noting that despite this setback, he’s optimistic because this is the golden era for AI research and talent. Companies can feel free to reach out to learn more about his team and recruit from the talented roster.
And in other news, Technology Source Share (技术源share) reports indicate that Manus’ official relocation to Singapore resulted in only 40 out of its 120 employees being relocated. And its attempt to “China shed” will mean it’ll have to hire locally in Singapore and even globally.
That’s why I’m covering an event that I think is quite interesting, showcasing the phenomenon in China: the unleashing of raw talent and increasing ambition. There still seems to be some misconception or prejudice from the West that thinks Chinese talent is imported or cheaper somehow in the AI and tech space. But what I want to highlight is that there is actually a lot of energy coming from the youth, and DeepSeek truly ignited that hunger again in the community.
I’m not saying the overall domestic consumption issue in China has been solved, nor is the property price drop no longer a drag on sentiment. BUT, I must say, it is exciting to see so many young people come together, energized and actively looking for ways to build.
With that...
AdventureX
What has been called the “biggest AI hackathon” in China took place this past week. The AdventureX Hackathon, held in Hangzhou (home to DeepSeek, Alibaba, NetEase, Geely, and many other notable companies), was organized by two high school graduates and a group of friends. What makes it special is that it targets empowering young talent, and the team of youngsters that founded the event used nine months to raise one million RMB in funding, as well as secure a free venue, accommodation, and meals for over 100 tech investors and leaders.
For its inaugural event in 2024, over 200 participants came together, with 90% of the hackathon participants being students, ranging from high school to graduate school, and an average age of ~22 years old. This year, the participants were capped at 26 years old, with the youngest participant being 11 years old. Nevertheless, nearly 900 young programmers, developers, and entrepreneurs came together.
The core founding team, including Ryan Zhu (also known as Zhu Yuhao), Max Liu, Marvin Cui, Christina Lee, and Kayna Huan, along with their group of friends (something like 20-30 teenagers), organized the entire event.
In an interview, it was said that on the day before the most important day of their lives thus far - the Gaokao, the national college entrance exam - the founders realized there were over 4,500 sign-ups from more than 182 countries and regions. And it was that moment they realized they had found their calling.
The group of young enthusiasts regards this sentence as a guiding principle and has made it the signature slogan of AdventureX, attracting like-minded individuals. It comes from an early interview clip of Steve Jobs, the tech icon who was also a geek in his youth. Ryan told reporters: “Everything around you that you call ‘life’ is made by people who are no smarter than you.”
“Once you realize this fact, life becomes much broader.” He encouraged young people to break the rules, not to accept fate passively, and to leave a mark during their short lives.
These are some candid video shots of the venue and exhibition halls that my contact, who has gone to Hangzhou to scout young talent, shot.
She said it was truly a bustling hub where the latest innovations in cloud services, AI, innovative hardware, and even some Web3 ideas converged. Overall, it was a mixed bag—while the buzz was definitely high, the number of truly standout projects was a bit smaller than expected. Still, a few cool innovations that caught her eye that show where things are heading.
Below are some key takeaways shared by my contact, an AI business leader from Singapore, as well as some consolidated points from various WeChat blogs, including Distill AI, Silicon Person Pro (硅星人Pro), and Nine Thousand Light Years (九千光年). Startups of all kinds showcased their demos, many obviously still a bit gimmicky, and do not exhibit a clear business strategy yet. However, my point for this article today is to highlight the enthusiasm in this space and some emerging trends.
First things first, there was a vast array of products, and the organizers themselves even created several applications to assist with the event's logistics and management. A chatbot helped answer inquiries regarding the event. A tool could help categorize emails automatically. A love match robot helped participants find romance based on their interests, age, and preferences, among other applications.
Brain-Computer Interfaces In Mental Healthcare
Noted by Distill AI, one exciting company that stood out is Emotiv. A company that is working on brain-computer interfaces (BCI). Consider gadgets that can read your brainwaves using EEG technology to track your level of focus and stress. What’s neat is that they’ve moved beyond just lab demos — their products are ready for everyday users. As mental wellness becomes more critical, these kinds of devices could be a game-changer for AI in healthcare.
This is definitely a growing trend in this space. I met with a company that was similar to this one at Beyond Macau as well, where the product uses a camera to read facial expressions and detect various levels of stress, happiness, anger, and other moods. It is reportedly being trialed in Shenzhen clinics for psychological well-being.
AR Glasses with a Practical Twist
So Meta glasses have been the dominant mainstream wearables in Silicon Valley, and now everyone’s waiting on Apple’s Vision Pro. My husband got a pair of the Ray-Ban Meta glasses to snap pics of my running toddler; I don't know how to feel about it, to be honest. The leading Chinese AR glass maker, Rokid, has made headlines across CNN and Bloomberg; others, such as Thunderbird and Xiaomi, are also pushing ahead with wearables.
recently wrote an interesting LinkedIn post, saying that AI glasses sit at the intersection of three drivers: 1) pressure to commercialize large models; 2) search for new endpoints in hardware; 3) quest for a post-smartphone interaction interface. Despite Snap’s failed attempt (perhaps they were too early), wearables like glasses (whether you like it or not) will likely play a much bigger role than they are currently perceived to.EvenRealities, a Chinese startup presenting laser-projection AR glasses that are lightweight and priced more affordably, is not yet fully AI-defied, but it’s a display of what it can do. These glasses show text in real-time, help with translation, and provide navigation right in front of your eyes. The design is lightweight, straightforward, and priced realistically. This practical approach might help AR catch on before the big-name brands fully crack the consumer market.
AI, the Fortune Teller and the Companion
In many ways, we’re seeing companion bots emerge. And they’re now becoming more specific. FateTell showcased an AI-powered fortune-telling app. Utilizing large AI models, it provides personalized destiny readings and interactive divination games. Sure, it’s not about being scientifically accurate—it’s more about giving users an emotional experience and some fun interaction. This shows that AI isn’t just for knowledge workers' efficiency gains, but it could potentially be adopted to meet people’s emotional needs in creative and unexpected ways.