Skip to main content

AI learning journey China

Shanghai and Shenzhen: an executive immersion into China's AI ecosystem

China has become a global leader in deploying AI at scale, rapidly translating advances into products, industries, and daily life.  On an AI learning journey with a delegation of leaders from Dutch corporates, we set out to move beyond abstract debate. We wanted to experience together how GenAI, agentic AI, and physical AI are being applied in real operating environments.

Across Shanghai and Shenzhen, we engaged with frontier model developers, robotics innovators, logistics operators, research institutes, and platform leaders. Alongside these visits, we discussed what China's approach means for Europe and for our own organisations, and what we can learn from it. We explored what global shifts mean for Europe's future, and how Dutch and European corporates can take a leading position.

Key moments from the AI learning journey

Setting the context at Deloitte China
At the Deloitte office in Shanghai, we were welcomed by Dora Liu (CEO of Deloitte China) and her leadership team for presentations and discussions about the AI technology landscape in China, the 'China edge', and key differences between the US, Europe, and China. Four takeaways stood out.

China as a learning ground
China is not just a market or supplier, but a training and learning ground for understanding how AI can be implemented at scale.

Speed to scale over IP protection
Competition in China focuses on speed to scale rather than strict IP protection.

Strategic independence through local AI models
Chinese public AI models, alongside on-premise open-source models, are being implemented to reduce dependency on US hyperscalers, to lower AI costs, and to improve cost predictability.

Economic urgency as an accelerator
China has a long-standing urgency for change in economic welfare. The question for Europe: what will it take to reach a similar level of urgency?

From foundation models to physical AI at scale
Day 2 explored the variety of AI implementation, with company highlights from our visits to a foundation model developer and a humanoid robotics company.

Minimax: the cost competitiveness race
Minimax delivers LLM, speech AI, music AI, and video AI. Founded in 2022, it is listed on the Hong Kong stock exchange with a market cap of USD 20 billion and around 500 employees.

The company differentiates on cost, claiming to be 10 to even 100 times cheaper than comparable alternatives, and on open-source models rather than API-based models. The performance of LLMs is converging, and the race for cost competitiveness has started.

AgiBot: speed of development in an unconstrained environment
AgiBot, founded in 2023, is a leader in humanoids. A mega factory was established a year after the company was founded, showing the enormous speed at which it is developing. Use cases are emerging in the industrial space, yet the question is where applications at scale will happen. It shows what you can do when you can develop an idea in an unconstrained way, in terms of talent, data, capital, and regulatory environment.

Working with China, not selling into China
Day 3 focused on the business dynamics in China and how public, private, and research institutions work together.

Co-win and NICE: a change of perspective
At Co-win we learned how foreign companies can do business in China. We also got to know the National Innovation Center par Excellence (NICE), which invests in research into emerging technologies such as AI and communications.

It is clear that we need to change our perspective and think about developing in and with China, rather than selling into China. To win, it is critical to find the right niche and innovate at full speed. Most importantly, it is about mindset: work like crazy, take nothing for granted, and be fanatic to win.

One quote captured it well: "money is being spent inefficiently, yet extremely effectively." Big bets are being made in the belief they will pay off over the long term.

Autonomous delivery at scale
In the afternoon, we visited a logistics site to see autonomous vehicles in operation. The provider focuses on middle and last-mile delivery, operating a large and growing fleet. This company has developed its full-stack platform for tracking and managing its vehicles entirely in-house.

Closing the day
After this visit, we joined the Dutch consul and a delegation from the Minister of Foreign Trade, together with corporates that have China-Netherlands connections. We then took a late-night flight from Shanghai to Shenzhen for the two remaining days.

Scale, ecosystems, and rapid prototyping in Shenzhen
Day 4 in Shenzhen focused on the scale of China's technology platforms and the ecosystems that support rapid innovation.

Tencent: scale and integration
Our first stop was Tencent. Founded in 1998, it has become the 8th largest global technology company. It is fully integrated across datacenters, cloud, LLM, apps, and content, with around 1.4 billion users on its platform using a full set of services provided by micro-apps such as payment, chat, and search.

The tour through their experience confirmed the sheer size, scale, and innovation power behind Tencent. Their strategy centers on an "open loop", and they claim to differentiate through a data advantage and better engineering. One quote stayed with us: "Technology is power; doing good is a choice."

Connecting through the Consul General
We had lunch by invitation of the Consul General, organised by NBSO Shenzhen, connecting Chinese technology and AI organisations from different industries with members of our group.

LASER: the low-altitude economy
In the afternoon, we learned about the low-altitude economy (LAE) from the research institute LASER. Through 4D modelling and the development of a multi-layered operating system, they enable the orchestration of drone delivery and transportation in cities.

Huaqiangbei Tech Center: an ecosystem for rapid prototyping
We concluded the day at the Huaqiangbei Tech Center, organised by Henk Werner from Troublemaker. The center is a marketplace of parts suppliers and an ecosystem of manufacturers, facilitating rapid prototyping and the development of market-ready solutions.

A full stack view of AI at Huawei
The final day of our AI journey took place at Huawei HQ, founded in 1998, with 2025 revenue of USD 126 billion.

Huawei: a hardware-first strategy
Huawei's main focus is on ICT infrastructure, smart devices, cloud, digital power, intelligent autonomous solutions, chipsets, and components. A deliberate strategic choice is to be a hardware company rather than a software or application company. With their solutions, they enable smart sectors such as smart retail, smart harbours, smart energy, and smart manufacturing.

With their leadership team, we had an open conversation about the differences and commonalities between China and the Netherlands/Europe, and the products they offer in "digital energy." We concluded the visit by experiencing their EVs and consumer products.

Sparking a deeper curiosity
This entire week has been full of insights. If there is one key takeaway: this journey to China has triggered a deeper sense of curiosity across many dimensions, which we will continue to explore.

Learn more

Would you like to hear more about what we learned in Shanghai and Shenzhen, or do you think a similar journey would be valuable for one of your fellow board members? Please get in touch.

You can also explore our other AI learning journeys here.

Did you find this useful?

Thanks for your feedback