China’s AI+ Economy

21 Jan 2026 11:00h - 11:30h

China’s AI+ Economy

Session at a glance

Summary

This discussion focused on China’s AI Plus economy initiative and its approach to artificial intelligence development, featuring perspectives from industry leaders, academics, and global investors. The panel explored how China is transitioning AI from technological frontier to economic driver, with expectations that AI will contribute $15 trillion globally by 2030, with China capturing about 25% of that value. Professor Gong Ke explained that China’s AI Plus action plan, launched in August, emphasizes diffusion and adoption rather than AGI development, aiming for 70% AI agent penetration by 2027 and over 90% by 2030.

The panelists highlighted China’s unique advantages in AI development, including its massive manufacturing and retail industries that provide scalable production environments, cultural openness to new technology, and infrastructure-first approach. Yutong Zhang from Moonshot AI emphasized efficiency in development, noting how Chinese companies achieve comparable performance using only 1% of the resources of frontier labs through fundamental research and engineering optimization. Dowson Tong from Tencent described China’s vibrant AI ecosystem with hundreds of model companies, many open-source, creating competitive pricing and diverse choices for customers.

The discussion addressed workforce implications, with China implementing nationwide AI education programs from primary school through university and training programs for teachers. Regarding energy needs, China is developing green energy infrastructure in western regions to power AI development sustainably. The panel acknowledged both opportunities and challenges, including the need to develop general intelligence capabilities over specialization and ensuring AI enhances rather than replaces human thinking. Overall, the discussion portrayed China’s systematic, economy-wide approach to AI integration as a distinctive model for global AI adoption.

Keypoints

Major Discussion Points:

– China’s AI+ Action Plan and Strategic Approach: The discussion focused extensively on China’s national AI+ initiative launched in August, which emphasizes AI diffusion and adoption across industries rather than focusing on AGI or chips. The plan aims for 70% AI agent adoption by 2027 and 90% by 2030, prioritizing practical applications in manufacturing, healthcare, and finance.

– Economic Value and Investment Perspective: Participants explored whether AI represents a bubble or genuine opportunity, with global investor perspective suggesting substantial value creation potential. The discussion highlighted AI’s expected $15 trillion contribution to global economy by 2030, with China positioned to capture about 25% of that value.

– Efficiency and Resource Optimization: A key theme emerged around China’s approach to AI development through efficiency rather than massive resource scaling. Companies like Moonshot AI demonstrated achieving comparable performance using only 1% of resources compared to frontier labs, emphasizing fundamental research and engineering optimization.

– Infrastructure and Energy Considerations: The conversation addressed China’s energy infrastructure strategy, including the “Dong Shu Xi Sun” (data in the east, computing power in the west) initiative that leverages renewable energy sources in western China to power AI development while maintaining green energy goals.

– Workforce Impact and Education: Participants discussed both positive and negative implications of AI adoption on employment and education, including nationwide teacher training programs, the need for AI literacy from elementary school, and the challenge of helping students use AI for deep thinking rather than instant answers.

Overall Purpose:

The discussion aimed to examine China’s AI+ development strategy from multiple perspectives – academic, industry, investment, and global comparison – to understand how China’s integrated policy approach and market dynamics are driving AI adoption across various economic sectors.

Overall Tone:

The tone was consistently optimistic and collaborative throughout the conversation. Participants demonstrated mutual respect and built upon each other’s insights rather than challenging them. The atmosphere remained professional and forward-looking, with speakers showing enthusiasm about AI’s potential while acknowledging challenges. The tone became slightly more interactive and engaging during the Q&A session, maintaining the positive, solution-oriented approach established during the panel discussion.

Speakers

– Guan Xin: Host/Moderator from China Global Television Network

– Dowson Tong: Senior Executive Vice President of Tencent and CEO of Tencent Cloud and Smart Industries Group

– Hisham Alrayes: Group CEO of GFH Financial Group, Global investor with expertise in wealth management and asset management

– Gong Ke: Professor, Executive Director of the Chinese Institute for New Generation AI Development Strategies at Nankai University

– Yutong Zhang: Founder and President of Moonshot AI

– Audience: Multiple audience members who asked questions during the Q&A session, including:

– Sayaka Tanaka from Japan, founder of Waffle (focuses on gender gap in industry and policy recommendation)

– Saeed Sakri, economist from Oman

Additional speakers:

None – all speakers mentioned in the transcript are included in the provided speakers names list.

Full session report

Comprehensive Report: China’s AI Plus Economy Initiative – A Strategic Discussion on Artificial Intelligence Development and Implementation

Executive Summary

This comprehensive discussion, moderated by Guan Xin from China Global Television Network, brought together leading voices from industry, academia, and global finance to examine China’s distinctive approach to artificial intelligence development through its AI Plus economy initiative. The panel featured Dowson Tong, Senior Executive Vice President of Tencent and CEO of Tencent Cloud and Smart Industries Group; Hisham Alrayes, Group CEO of GFH Financial Group; Professor Gong Ke, Executive Director of the Chinese Institute for New Generation AI Development Strategies at Nankai University; and Yutong Zhang, Founder and President of Moonshot AI.

The discussion revealed China’s strategic pivot from pursuing artificial general intelligence (AGI) to emphasising practical AI diffusion and adoption across economic sectors. With projections that AI will contribute $15 trillion to the global economy by 2030, China is expected to capture about a quarter of this value through its comprehensive AI Plus action plan announced in August of the previous year. This initiative targets ambitious adoption rates of 70% AI agent penetration by 2027 and over 90% by 2030, focusing on tangible economic transformation rather than theoretical advancement.

China’s AI Plus Strategic Framework

Policy Direction and Implementation

Hisham Alrayes provided crucial insight into China’s AI Plus action plan, explaining it as “a national initiative officially announced last year in August” that emphasises practical implementation over theoretical development. The plan targets specific diffusion rates: “70% of AI agent penetration by 2027 and over 90% by 2030,” representing a systematic approach to economy-wide AI integration.

Professor Gong Ke reinforced this practical focus, noting a deliberate absence of AGI and semiconductor discussions in policy frameworks. This strategic emphasis encompasses specific areas including consumption, education, and healthcare, representing a systematic, economy-wide approach to AI integration that moves beyond individual company benefits to create value throughout the entire economic ecosystem.

Economic Value Creation and Investment Perspective

From a global investment standpoint, Alrayes provided a compelling argument against the notion of an AI bubble. “AI is not in a bubble,” he stated definitively. “AI offers substantial value across the full spectrum from power generation to capital markets.” His perspective highlighted China’s unique philosophical approach: “China is looking to create value throughout the economy, very clear, with very specific objectives across the economy, not just as a benefit of those companies. And this is the difference in the philosophy, I think.”

This economic philosophy underpins China’s position to capture approximately a quarter of the projected $15 trillion global AI contribution by 2030, representing a substantial portion of worldwide AI-driven economic value.

China’s Competitive Advantages in AI Development

Market Scale and Cultural Openness

Yutong Zhang from Moonshot AI emphasised China’s distinctive market advantages, particularly its massive manufacturing and retail industries that provide ideal environments for scalable AI systems. “China has huge manufacturing and retail industries providing environment for scalable AI systems with abundant data and use cases,” she explained. This scale advantage is complemented by exceptional cultural openness to new technology, with 85% of Chinese consumers believing autonomous driving is safe and beneficial.

The combination of scale and cultural acceptance creates unique conditions for AI deployment and testing that are difficult to replicate in other markets. This environment enables rapid iteration and improvement of AI systems through real-world application and feedback.

Ecosystem Vibrancy and Open-Source Culture

Dowson Tong highlighted China’s vibrant AI ecosystem, characterised by hundreds of model companies and a strong open-source culture. “China’s vibrant AI ecosystem has hundreds of model companies with strong open-source culture driving down costs,” he noted. This competitive environment has created diverse choices for customers whilst driving down pricing through market competition.

The open-source approach has particular significance for cost optimisation, which Tong identified as “crucial for AI diffusion to wider communities, leaving nobody behind.” This philosophy aligns with China’s broader objective of ensuring inclusive access to AI technology across different segments of society.

Efficiency-Driven Innovation Approach

Resource Optimisation Strategy

One of the most striking revelations came from Yutong Zhang’s discussion of Moonshot AI’s resource efficiency in development. “We have built a comparable state-of-the-art performance model, but by only using 1% of the resource of the comparable frontier labs,” she revealed. This achievement stems from their approach to fundamental research and engineering optimisation rather than simply scaling computational resources.

“We know that we don’t have the luxury to just scale up the compute. So I think our development approach is more about we do a lot of fundamental research and innovations,” Zhang explained. This constraint-driven innovation has become a competitive advantage, demonstrating that efficiency can be as valuable as raw computational power in AI development.

Infrastructure-First Thinking

Zhang also emphasised the importance of infrastructure-first thinking in China’s AI development strategy. “Infrastructure-first thinking with cheap supply through data centres in multiple cities enables innovation,” she noted. This approach ensures that the foundational elements for AI deployment are in place before scaling applications, creating a stable platform for widespread adoption.

Sustainable Energy Infrastructure

Green Energy Integration

Professor Gong Ke detailed China’s approach to sustainable AI infrastructure through what he described as “data in the east and the computer power in the west” initiative. This strategy leverages renewable energy sources in western China, including wind and solar power, to support AI computing needs whilst serving eastern markets.

“China is building energy infrastructure with green energy in western regions using wind and solar power,” Gong Ke explained. This approach addresses the significant energy requirements of AI systems whilst maintaining China’s green development objectives, demonstrating how environmental sustainability can be integrated into AI infrastructure planning.

Workforce Transformation and Education

Employment Impact and Opportunities

The discussion addressed both positive and challenging aspects of AI’s impact on employment. Alrayes acknowledged that while some jobs will be replaced, AI adoption is creating significant opportunities, noting a “deficit of 5 million workers required to have the AI capability” in the Chinese job market, with nationwide upskilling programmes planned to address this gap.

Yutong Zhang provided insight into changing organisational requirements, observing that “organisations need people with general intelligence rather than specialisations, as AI provides on-demand domain expertise.” This shift emphasises the growing importance of learning ability over past experience, as “knowledge expires faster” in an AI-enabled environment.

Educational System Transformation

China’s approach to AI education represents a comprehensive transformation from primary school through university level. Alrayes outlined nationwide teacher training programmes, ensuring educators are equipped to deliver effective AI instruction.

However, he also identified a significant challenge: “The challenge is how to help students to use AI to empower their deep thinking, not satisfied by the instant answer. That’s a very big challenge.” This concern reflects broader questions about maintaining critical thinking skills in an AI-enabled educational environment.

Dowson Tong offered a complementary perspective on AI education, emphasising the role of accessible tools in organic learning. “I think the new generation learns AI not only from schools and the traditional education institutes. By having free tools… encouraging the new generation to be curious and use these freely accessible AI tools would be the best way to develop the habits of using AI as part of the learning tool.”

Industry Integration and Practical Applications

Cross-Sector Implementation

The discussion highlighted AI’s integration across multiple business functions and industries. Dowson Tong described how “AI is being integrated into every function within operations, from programming to accounting, enhancing individual productivity.” This comprehensive integration spans healthcare applications, marketing optimisation, and retail enhancement, demonstrating tangible results across diverse sectors.

The practical focus of China’s AI implementation contrasts with approaches that prioritise theoretical advancement over immediate application. This pragmatic strategy aims to generate measurable economic value whilst building the foundation for more advanced AI capabilities.

Model-Agnostic Approach

Tong also emphasised the importance of flexibility in AI deployment, advocating for a “model-agnostic approach allowing customers to choose different AI models for different purposes rather than forcing single solutions.” This approach recognises that different applications may require different AI capabilities, providing customers with optimal solutions for specific use cases.

Areas of Consensus and Strategic Alignment

Economic Value and Investment Viability

The panel demonstrated remarkable consensus on AI’s substantial economic potential. Both Alrayes and other speakers emphasised that AI represents genuine economic opportunity rather than speculative investment, with concrete projections for global value creation and China’s significant share of that market.

Efficiency and Accessibility Focus

Strong alignment emerged around the importance of efficiency and cost optimisation for widespread AI adoption. Both Zhang and Tong emphasised that making AI more efficient and cost-effective is essential for broader adoption and ensuring inclusive access across different segments of society.

Infrastructure Development Priority

Zhang and Professor Gong Ke agreed on the fundamental importance of infrastructure development for AI advancement. Their shared emphasis on data centres, renewable energy systems, and foundational infrastructure reflects a common understanding that robust infrastructure enables innovation and scalability.

Strategic Differences and Varied Approaches

Development Philosophy Variations

While maintaining overall alignment, speakers demonstrated different emphases in their approaches to AI development. Zhang focused on efficiency through fundamental research due to resource constraints, whilst Tong emphasised ecosystem diversity and open-source collaboration as key differentiators. These approaches are complementary rather than contradictory, reflecting different aspects of China’s multi-faceted AI strategy.

Implementation Priorities

Some variation emerged in discussions of implementation priorities. Alrayes emphasised top-down economic philosophy and systematic diffusion targets, whilst Gong Ke focused on technical infrastructure and sustainable energy integration. These different perspectives reflect the complexity of AI implementation across policy, finance, and technical domains.

International Engagement and Outstanding Questions

Global Perspectives

The Q&A session brought valuable international perspectives, particularly from Sayaka Tanaka from Japan, who sought specific details about AI education curriculum and teacher training methodologies. Her questions highlighted global interest in China’s educational approach and the need for more detailed information about implementation specifics.

Saeed Sakri from Oman raised important questions about negative impacts on traditional jobs and education abilities, seeking empirical assessment of challenges alongside opportunities. These questions reflect broader international concerns about AI adoption’s societal implications.

Unresolved Implementation Challenges

Several significant questions remain from the discussion. The specific methodologies for measuring the ambitious 70% and 90% AI diffusion targets by 2027 and 2030 require further elaboration. Additionally, the detailed implementation of nationwide upskilling programmes and the redesign of educational systems to promote deep thinking rather than instant gratification represent ongoing challenges requiring continued development.

Future Implications and Strategic Outlook

Global AI Development Paradigms

The discussion revealed China’s AI development approach as a distinctive model that prioritises practical implementation over theoretical advancement, widespread diffusion over concentrated development, and economy-wide benefits over individual company gains. This approach represents a coherent strategy supported across diverse stakeholders from academia, industry, and finance.

The emphasis on efficiency, accessibility, and systematic integration suggests that China’s model may offer lessons for other nations seeking to implement AI at scale whilst managing societal impacts and ensuring inclusive benefits.

Technological and Economic Integration

China’s infrastructure-first approach, combined with its focus on green energy integration and sustainable development, demonstrates how AI advancement can align with broader environmental and economic objectives. The “data in the east and computer power in the west” initiative exemplifies how technological infrastructure can support both AI development and renewable energy utilisation.

Conclusion

This discussion illuminated China’s systematic, economy-wide approach to AI integration as a distinctive model for global AI adoption. The AI Plus initiative represents a strategic choice to prioritise practical diffusion and economic value creation over theoretical advancement, supported by unique market advantages including massive scale, cultural openness to technology, and a vibrant competitive ecosystem.

The emphasis on efficiency, sustainability, and inclusive access demonstrates how AI development can align with broader societal objectives whilst generating substantial economic value. China’s expected capture of approximately a quarter of the global AI economy by 2030 reflects not just market size but a comprehensive strategy encompassing policy coordination, infrastructure development, educational transformation, and workforce adaptation.

The discussion’s collaborative tone, combined with honest acknowledgement of challenges, suggests a mature approach to AI development that balances opportunity with responsibility. As China implements its ambitious diffusion targets and other nations observe the results, this model may provide valuable insights for global AI adoption strategies.

The outstanding questions and unresolved challenges identified highlight areas requiring continued research, monitoring, and international dialogue. The success of China’s AI Plus initiative will ultimately be measured not just in economic terms but in its ability to enhance human capabilities, create inclusive opportunities, and contribute to sustainable global development.

Participants encouraged continued engagement using hashtag #AM26 for social media discussions related to this topic.

Session transcript

Guan Xin

Good afternoon, ladies and gentlemen, welcome to this issue briefing of China’s AI plus economy. My name is Guan Xin. I’m from China Global Television Network.

We’re now at a pivotal moment. Artificial intelligence is rapidly transitioning from a technological frontier to a core driver of global growth. By 2030, AI is expected to contribute about 15 trillion U.S.

dollars to the global economy, and China is expected to capture about a quarter of that value. In 2025, 87% of Chinese companies plan to increase AI investment and more than half reported faster deployment. And China’s national AI plus action plan is accelerating integration into vast industries, the manufacturing, healthcare, finance, just to name a few.

And China’s scale, integrated policy approach, and a vibrant market offer a unique model for the global AI adoption. So to delve further into China’s AI plus development, I’m privileged to be joined here by an outstanding panel of speakers, and they come from different backgrounds, from the academia, industry, frontier innovation, and global finance.

Please allow me to introduce them. Professor – oh, let me start with Mr. Dowson Tom, Senior Executive Vice President of Tencent and CEO of Tencent Cloud and Smart Industries Group.

Very warm welcome to you both, Dowson. And next to him is Ms. Yu Tong Zhang, founder and president of Moonshot AI.

Thank you so much for joining us. And next to him is Professor Gong Ke, Executive Director of the Chinese Institute for New Generation AI Development Strategies at the Nankai University. Welcome Professor Gong.

And last but not least, Mr. Hesham Alrayes, Group CEO of GFH Financial Group. It’s such a pleasure to have you with us.

So this session is about 30 minutes long, including our discussions and a very brief Q&A session. And just a reminder, if you would like to share your thoughts about this session on social media, please use the official hashtag AM26. Now let’s dive right in.

Let me start with Hesham, because I’d like to have a global investor perspective. We’re now in this unprecedented fervor of AI investment, but discussions about the AI bubble, you know, the intensity of capital investments meets uncertain paths of return also grow quickly. So I know you’re an investor who navigates cycles.

And in what areas are AI generating tangible, measurable economic value today?

Hisham Alrayes

Thank you for having me today. I think we had a discussion this morning in the governance session, and there was a poll, and 60% of the people think it’s not a bubble, and the opportunity to capture the value is still there. And I think that there is substantial value and returns to be extracted out of this new market-changing invention and change.

And not only in AI-specific, but also through the full spectrum from power generation to data centers, to chips, to technology, software, and capital market. So the advancements of technology require heavy investments, and we come from the wealth management and asset management, and you see the developments throughout the world, whether through in China or in the States or also from our region, the requirements of government investment solo cannot drive and take it to the next level.

So you’ll see a lot of the founders and companies they raise globally, and when their capital markets are stronger, you’ll see more advancements, stronger research, and higher delivery, and a shorter cycle for delivering return investments.

So I think the opportunity is massive, and we are just at the beginning, and the ability to transform this throughout the economy to create a true value in a short period of time, whomever does that will be the winner.

And I think China AI plus is a true serious desire to deliver on that to win the race.

Guan Xin

First of all, China’s AI plus action plan is what we’re going to zoom in on now. Professor Guan, give us an academic perspective on China’s AI plus action plan and how it shows about the potential of AI in creating growth drivers across China’s economy.

Hisham Alrayes

Yeah, actually the AI plus action is a national initiative officially announced last year in August. So in this document, if you read the document, you cannot find any words talking about AGI. You cannot find any words talking about the chips.

It does not mean the Chinese government don’t understand the importance of AGI, don’t understand the importance of chips. But that means the Chinese government lay more emphasis on the diffusion, on the adoption, on the penetration of AI, and to make AI really make values in production and in the daily life. So the main direction is to move AI from chat to product, to services.

So the goal was set in two steps. First, by 2027, next year, the diffusion rate of the AI agent and the intelligent terminal will exceed 70 percent, and by 2030, over 90 percent. So that’s the goal of the diffusion in China.

That’s the goal of AI plus.

Guan Xin

Right. And let me turn to Yutong Zhang. Leading an AI company at the frontier, how do you think about China’s approach, like focusing on the diffusion and does not mention AGI?

And if we talk about the diffusion of AI, what are the biggest blockers or enablers of AI adoption in the economy?

Yutong Zhang

I think China is a very unique market. I think China’s market is huge in many perspectives. It has a huge manufacturing industry, a huge retail industry.

I think many industries give us the environment to build a real scalable system in production. There will be many data, many use cases that people can try to use AI to couple with. So I think this is one of the advantages in China’s market.

And I think, secondly, is the openness to new technology, because we have seen it continuously The pattern is several of the technology wave, you know, in EV, in solar, in smartphone, in autonomous driving.

I’m personally surprised that 85% of the Chinese people think autonomous driving is safe and also can improve their lifestyle and, you know, productivity. So that’s why I think China has like thousands of robo-taxi already, you know, run in like tens of the cities. So I think this like openness to technology, really ready to embrace new technology is very unique for China.

And also, I think lastly, I think AI actually helps a lot on really deliver a hyper productivity tool for the individuals. Even today, you know, like we are still a startup company. We do a lot of the hirings.

The resume that we receive are personal websites already. Nobody actually give us the PDF anymore. They use a web coding product.

You know, people with zero knowledge of code before, but they can, you know, create this like a beautiful personal website to apply for jobs. So I see this like adoption is, you know, from different layers of drivers.

Guan Xin

That’s right. And Dowson Tencent is serving vast industries and you see the full spectrum. Just tell us how Chinese, you know, technology and industrial leaders are integrating AI and to drive tangible results.

Dowson Tong

Well, we definitely work with a lot of customers from different industries to deploy all forms of AI. When people talk about AI, I think we might tend to think of one big super system and give it a term AGI. But in fact, in reality, there are many different types of models that serve different purposes.

And more specifically, when you look at the different industries and different enterprises, they are trying to use AI in every single function within their operation. For example, at Tencent, not only many of our programmers are using coding tools extensively to turn out features much faster than before. We’re also seeing product managers, designers, accountants, basically, you know, from different roles are using these modern tools to automate their work, building tools that otherwise won’t be possible to enhance individuals’ productivity.

Some of the customers that we work with in the retail sector, for example, some of them use gen photo technology, gen 3D model technology to speed up their product design cycle. We help a lot of customers to use AI to get better ROI on their marketing dollars that they spend by better targeting, more personalized service that leads to better conversion. And not to mention, there are a lot of AI applications in healthcare.

We help some of the pharmaceutical companies with drug discovery. So I think we’re seeing a lot of energy, positive energy to endorse AI in the current environment. One thing I want to highlight, though, is that it’s pretty common in China that people are looking for good ROI.

They want to lower the cost of using AI. And for us, as part of our cloud technology, we always look for optimization for the cost of using AI so that it can be distributed, it can be diffused to a much wider community, leaving nobody behind.

Guan Xin

And Hisham, let me bring back your global investment perspective. Now you have heard about China’s national strategy and also these business dynamics. What do you think about China’s AI plus development?

And do you find it’s maybe, how do you compare it to global AI development?

Hisham Alrayes

Yeah, so my view is, one, it starts from the top. So it’s the discipline and direction, whether it’s a country or a region or even if it’s a company. So in our company, I went and we assigned 10% of the yearly objectives of all chiefs that they have to see how to adapt the AI.

And not to adapt AI as an interface, but how to transform that in actual efficiency. Now, if you take that example and move it to a country, you will see the compounded benefit onto the economy. Then you look at the open structure of the China AI philosophy, let’s call it, then you have the non-open structure.

And that signal that the benefit, they want to see it to trickle down into the economy, into the companies, so it’s more affordable. So it’s not the benefit of that company, of that product, the return of that individual. It’s not an individual, it’s an economy.

And this is very interesting. And they have applied that across the economy. Now, in our region, you will see also very leading models in UAE, for example, and how they want to make the AI as a part of mandatory learning process in schools, also in Bahrain.

So different countries, they are taking different commitment levels. And I think soon or now, we are seeing AI becoming as a part of the lifestyle. So it’s more of the lifestyle and how you utilize that.

But it’s very interesting to see what’s the endgame. I think China is looking to create value throughout the economy, very clear, with very specific objectives across the economy, not just as a benefit of those companies. And this is the difference in the philosophy, I think.

Guan Xin

That’s right. No, your observation is very right on the point, because China is really integrating AI into various aspects of the economy. And talking about the economy, it ultimately is about people’s economy and how we work.

Professor Guan, in your opinion, how is China’s AI push influencing workforce strategies from talent development to organizational design?

Gong Ke

Actually, back to the National Initiative AI+, there are six emphases, six focuses are identified. The first is to use AI to empower. choice to make this because that’s on the other side in China is to try to decrease the trade Supplers worldwide, so we will increase the domestic consumption Early this morning.

I heard from Jingdong JD. That’s the goose named with smart has doubled the last year and the the the fourth one is to increase the welfare of the citizen including education and Health care talking about education now in China we have very ambitious program to embed a I basically see from the primary school and Also in university the AI agents widely used is for example in my university.

There’s About thousand different kinds of agents is used by professors and students to help their studies their learnings So that’s the program I think it is very important for the young people to have the capability use AI We don’t know the future job, but we know definitely the future job needs the capability of using AI

Guan Xin

right, and I think China is definitely a Setting out these priorities and you Tom you’re in the industry and a lot of people are comparing the environment of AI development between China and other you know major countries When it comes to you know like infrastructure or energy or other?

You know key elements to drive the future AI development in your opinion What supports are proving the most critical in you know scaling the successful AI use I

Yutong Zhang

Think I think there are definitely differences that what we are trying to do I think We have built a comparable state-of-the-art performance model, but by only using 1% of the resource of the you know comparable frontier labs so I think I Think the difference is from day one that we know that we don’t have the luxury to just scale up the compute So I think our development approach is more about we do a lot of fundamental Research and innovations so the founder of the company is actually from academia You know all the technology of AI actually all from academia very Open community, and then we actually focus a lot on Doing the fundamental research to trying to increase the efficiency so I would say the the efficiency of Development is very important and not many company can do that.

I think a Lot of the research are just stay at the lab But we spend a lot of time to making all the research work In a production system with all the engineering mindset to make it work at scale You know like a moon the optimizers that we are using we are the first one to make it workable You know large-length model training and also we have the linear attention Kimmy we call it a Kimmy linear, which is faster than the full attention system So I think we did a lot of things to make sure that the efficiency is really really high And this is what you know has been defined when you are developing the AI systems from China, so I think that from the support perspective definitely Infrastructure is very important.

I think China always have a infrastructure first thinking we’re doing everything you know to build a highway and to build the Electricity plants and also the huge data centers in multiple cities, so I think that really makes the supply Very very cheap, so that will not you know unblock the innovations from frontier technology So I think that would be a very helpful and but the company also needs to be extremely efficient

Guan Xin

So efficiency is a key word here Dawson. Would you agree? Would you like to list some of the?

Factors, you know the most important to scale on AI development in China, and how is trying to doing that?

Dowson Tong

Well efficiency is definitely a very important drive in the China market but in addition to that I think one thing that’s of notice that The China AI ecosystems or technology ecosystems in general is very vibrant There are a lot of players In the model side compared to some other markets.

I think the number of model companies We have a lot more in China and in fact hundreds of them and and and they work together, too Yeah, right a lot of them are open source and two of the biggest IPO In the Hong Kong stock exchange this year where I companies with very different focus one focus on consumer side overseas market the other Open source, but more on the enterprise side and including one shot.

It’s also a partner of ours Even though we also have our own model called hi but the interesting thing is that there’s so many players in the ecosystem with open source being a very strong driver that helps lower the cost of Of Doing the inference using AI we’re seeing the the cost of using AI in China at least for the past 18 months continue to come down and And and for for Tencent we understand that customers want choices you know, there are many different use cases for company and They might want to use different models of different sizes for different purposes, so our focus as part of the cloud strategy is to provide tools products that are Model agnostic support different types of models.

I think that gives the the power of Choosing the right model for themselves back to the hands of the customers

Guan Xin

And no one has talked about energy and I like to ask a professor going about this because energy use is really one of the Top the crucial, you know elements in AI development in the future. Do you think China is well positioned to provide energy much-needed?

Gong Ke

Yes, actually, there’s a another very important Initiative relevant to the AI development is we called the the energy infrastructure built mainly with green energy in the western part of China because we’re rich wind power and solar power there and use the wide band transmission Network to help the companies in the western coast to use that computer power.

So we called it Dong Shu Xi Sun that’s the data in the west and the computer power Yeah, it can feel power in the west but data in the east So that’s a big issue. That means By the end of 2030 A large amount of power used by AI would be green, renewable power.

Guan Xin

Fantastic news. And we may have a few extra minutes to open this to the floor. Okay, please.

Please briefly introduce yourself and ask questions.

Audience

Thank you so much for sharing insight. I am Sayaka Tanaka, and I am from Japan. I am the founder of Waffle, which’s mission is to grow the gender gap in the industry and also engage in policy recommendation.

And I have a question to Gon-san. I’d like to understand the level of the content of AI education in Chinese elementary and junior high and high schools. Is creative education such as developing product using AI not just AI literacy including in compulsory education?

And also, how are the teachers trained or supported so that they can teach AI effectively?

Gong Ke

There’s a nationwide program for teaching the teachers, training the trainers to use AI. And also, we carefully adopted the United Nations framework for the competence of teachers and students in AI. So if you could see, China is the first country to have the white paper on the AI application in education.

So if you have chance to read that white paper, you can get more information.

Dowson Tong

Can I add to that? I think the new generation learns AI not only from schools and the traditional education institutes. By having free tools, in our case, like what we call Yuanbao, many young generations can go to these JPT-like app to ask questions, to meet their curiosity needs.

I think encouraging the new generation to be curious and use these freely accessible AI tools would be the best way to develop the habits of using AI as part of the learning tool.

Audience

Thank you for an interesting session. My name is Saeed Sakri. I’m an economist from Oman.

I am curious, anybody can answer actually, about the impact or the negative impact of using AI on education especially and on labor market. Adopting technologies has always been also part of adaptation, has had negative impact on education, on economic activities, especially the labor market. I don’t know if you have assessed or looked into that and what was your conclusion and how you have already adapted.

And I’m very happy to see that you have a white paper for use of AI in education system. But maybe you have already an experience and have seen some of the negative impact of adopting AI and the impact on people’s education abilities and so on. And also on the traditional jobs and the markets that usually everybody can do and now it is being done by AI.

Gong Ke

Thank you very much. Let me shortly answer your question about education because there’s a negative side. So the challenge is how to help students to use AI to empower their deep thinking, not satisfied by the instant answer.

That’s a very big challenge. And also talking about the workforce, the employment, by far that’s an early phase of AI adopting. The employment is increasing last year, the year before last year.

And in the Chinese job market, there’s deficits of 5 million workers required to have the AI capability. But along with deeper adoption of AI, some jobs will be replaced. So there’s a nationwide plan to upskilling the workers to be able to use AI and to find new tasks which is empowered by AI.

So I stop here. Thank you.

Yutong Zhang

Yeah, I think I just want to add some echo to Professor Gong’s comments. I think it’s not necessarily a negative effect, but I do see change in terms of the people that we want to work with, new type of AI native organizations. So we really emphasize on having the people have general intelligence capabilities rather than specializations.

I think that’s a trend because AI can provide some on-demand expertise in the domain. And also I think for the organizations, the historical function-based organization structure will also be changed. So I think secondly, I feel like humans are creating new knowledge at a faster speed than before.

For the past two years, I sleep very little and we have a lot of progress. I think every day we’re feeling that the learning ability is very important than the past experience because the past experience and the past knowledge may get expired sooner than before. So I think that is also very important.

So actually, I think the education system, if they can adapt to develop more general thinking and general knowledge, and also really good learning ability, really good AI proficiencies, we still think we are still finding difficulties in hiring rather than the opposite.

Dowson Tong

I think we need to learn how to ask the right questions. That’s a very important, yeah.

Guan Xin

Maybe no time for one, maybe just, can you do it very briefly? Very briefly.

Audience

So last year it was noteworthy. In the green room, American supremacy in AI was celebrated. And three weeks later, there was the deep sea moment.

What’s the likelihood we have something similar this year? How far are we from another deep sea moment this year?

Yutong Zhang

Don’t know, don’t know. We will launch a new model very soon. Your model is quite good.

Kimi, this is a good model.

Guan Xin

Okay, let’s wait and see. There are always new excitement on the horizon. We have this open-minded participants in this industry, and let’s hope for more progress in the sector.

And thank you so much for joining us. And if you continue with your thoughts about the conversation, you can use the hashtag AM26 on social media. The session is concluded.

Thank you so much. Thank you. Thank you.

Hisham Alrayes

Speech speed

103 words per minute

Speech length

692 words

Speech time

401 seconds

AI is not in a bubble and offers substantial value across the full spectrum from power generation to capital markets

Explanation

Alrayes argues that AI represents a genuine opportunity rather than a speculative bubble, citing a poll where 60% of people agreed it’s not a bubble. He emphasizes that substantial value and returns can be extracted from this market-changing invention across multiple sectors including power generation, data centers, chips, technology, software, and capital markets.

Evidence

Disagreed with

– Gong Ke

Disagreed on

Primary driver of AI adoption success

Gong Ke

Speech speed

118 words per minute

Speech length

500 words

Speech time

252 seconds

The AI Plus action plan focuses on diffusion and adoption rather than AGI or chips, aiming to move AI from chat to products and services

Explanation

Professor Gong explains that China’s AI Plus action plan, announced in August, deliberately avoids mentioning AGI or chips to emphasize diffusion, adoption, and penetration of AI into production and daily life. The main direction is transitioning AI from conversational applications to actual products and services that create real value.

Evidence

Agreed with

– Yutong Zhang

Agreed on

Infrastructure development is fundamental to AI advancement

The “Dong Shu Xi Sun” initiative places data processing in the west using renewable energy while serving eastern markets

Explanation

Professor Gong explains this specific initiative that translates to “data in the west and computer power in the west” but serves eastern markets. This approach ensures that by 2030, a large amount of power used by AI will be green, renewable power.

Evidence

Specific timeline that by end of 2030, large amount of AI power will be green and renewable

Major discussion point

Energy Infrastructure and Sustainability

Topics

Infrastructure | Development

AI education is being embedded from primary school through university with nationwide teacher training programs

Explanation

Topics

Future of work | Development

Agreed with

– Yutong Zhang

Agreed on

AI is transforming workforce requirements toward general intelligence and learning ability

Yutong Zhang

Speech speed

143 words per minute

Speech length

Zhang argues that the rapid pace of AI development means humans are creating new knowledge faster than before, causing past experience and knowledge to become obsolete more quickly. This shift makes learning ability and adaptability more crucial than accumulated experience.

Evidence

Personal example of sleeping very little and making significant progress over past two years; company still has difficulty hiring despite concerns about job displacement

Major discussion point

Impact on Education and Employment

Topics

Future of work | Online education

Dowson Tong

Speech speed

123 words per minute

Speech length

692 words

Speech time

335 seconds

AI is generating tangible results across industries through productivity enhancement, marketing optimization, and healthcare applications

Explanation

Tong describes how Tencent works with customers across different industries to deploy various forms of AI, emphasizing that AI consists of many different types of models serving different purposes. He provides concrete examples of AI applications from programming and design to marketing and drug discovery that are delivering measurable business results.

Evidence

Tencent programmers using coding tools for faster feature development; product managers, designers, accountants using AI tools; retail customers using gen photo and 3D model technology; AI for marketing ROI and personalized service; pharmaceutical companies using AI for drug discovery

Major discussion point

AI Investment and Economic Value

Topics

Economic | Development

China’s vibrant AI ecosystem has hundreds of model companies with strong open-source culture driving down costs

Explanation

Tong highlights the exceptional vibrancy of China’s AI ecosystem compared to other markets, noting the presence of hundreds of model companies with many focusing on open-source development. He points to two major AI company IPOs in Hong Kong as evidence of this ecosystem’s strength and diversity.

Evidence

Hundreds of model companies in China; two biggest IPOs in Hong Kong stock exchange this year were AI companies with different focuses; one focused on consumer side overseas market, the other on open-source enterprise side; AI costs in China have been declining for 18 months

Major discussion point

China’s Unique Market Advantages for AI Development

Topics

Economic | Development

Disagreed with

– Yutong Zhang

Disagreed on

AI development approach: Resource scaling vs. efficiency optimization

Cost optimization is crucial for AI diffusion to wider communities, leaving nobody behind

Explanation

Tong emphasizes that Chinese customers consistently seek good ROI and want to lower AI usage costs. As part of Tencent’s cloud strategy, they focus on optimizing AI costs to enable broader distribution and diffusion across wider communities, ensuring inclusive access to AI technology.

Evidence

Audience

Speech speed

133 words per minute

Speech length

299 words

Speech time

134 seconds

Questions about negative impacts on traditional jobs and education abilities need ongoing assessment

Explanation

An audience member from Oman raised concerns about the negative impacts of AI adoption on education and labor markets, noting that technology adoption has historically had adverse effects on economic activities and traditional jobs. The question seeks assessment of these impacts and adaptation strategies.

Evidence

Historical precedent that adopting technologies has had negative impact on education and economic activities, especially labor market

Major discussion point

Impact on Education and Employment

Topics

Future of work | Online education

Guan Xin

Speech speed

139 words per minute

Speech length

906 words

Speech time

389 seconds

AI is expected to contribute $15 trillion to the global economy by 2030, with China capturing about a quarter of that value

Explanation

Guan Xin presents the massive economic potential of AI development globally and China’s significant expected share. She emphasizes that AI is transitioning from a technological frontier to a core driver of global growth, positioning this discussion within the context of substantial economic transformation.

Evidence

Agreements

Infrastructure | Development

Economic | Development

Unexpected consensus

Focus on AI diffusion over AGI development

Speakers

– Gong Ke
– Yutong Zhang
– Dowson Tong

Arguments

The AI Plus action plan focuses on diffusion and adoption rather than AGI or chips, aiming to move AI from chat to products and services

Chinese companies focus on fundamental research and efficiency, achieving comparable performance with only 1% of resources used by frontier labs

AI is generating tangible results across industries through productivity enhancement, marketing optimization, and healthcare applications

Explanation

There is unexpected consensus among speakers from different backgrounds (academia, industry startup, and large corporation) that practical AI implementation and widespread adoption is more important than pursuing AGI. This represents a pragmatic approach that prioritizes immediate economic value over theoretical breakthroughs.

Topics

Economic | Development

Positive view of AI’s impact on employment

Speakers

– Gong Ke
– Yutong Zhang

Arguments

AI adoption is creating 5 million job openings requiring AI capabilities, with nationwide upskilling programs planned

Learning ability is becoming more important than past experience as knowledge expires faster

Explanation

Despite common concerns about AI displacing jobs, both academic and industry perspectives converge on AI creating new employment opportunities and enhancing human capabilities rather than simply replacing workers. This optimistic view contrasts with typical dystopian narratives about AI and employment.

Topics

Future of work | Development

Overall assessment

Summary

The speakers demonstrate remarkable consensus across key areas: AI’s substantial economic potential, the importance of efficiency and cost optimization, the need for infrastructure development, the transformation of workforce requirements toward general intelligence, and China’s unique advantages in AI development through scale and ecosystem vibrancy.

Consensus level

High level of consensus with strong alignment on strategic priorities. The implications suggest that China’s AI development approach, emphasizing practical implementation over theoretical advancement, widespread diffusion over concentrated development, and economy-wide benefits over individual company gains, represents a coherent and widely supported strategy among diverse stakeholders from academia, industry, and finance.

Differences

Different viewpoints

AI development approach: Resource scaling vs. efficiency optimization

Speakers

– Yutong Zhang
– Dowson Tong

Arguments

Chinese companies focus on fundamental research and efficiency, achieving comparable performance with only 1% of resources used by frontier labs

China’s vibrant AI ecosystem has hundreds of model companies with strong open-source culture driving down costs

Summary

Zhang emphasizes efficiency through fundamental research due to resource constraints, while Tong focuses on ecosystem diversity and open-source collaboration as the key differentiator

Topics

Economic | Development

Primary driver of AI adoption success

Speakers

– Hisham Alrayes
– Gong Ke

Arguments

China’s approach emphasizes creating value throughout the economy with specific objectives, not just benefiting individual companies

The AI Plus action plan focuses on diffusion and adoption rather than AGI or chips, aiming to move AI from chat to products and services

Summary

Alrayes emphasizes top-down economic philosophy and discipline as the key driver, while Gong Ke focuses on specific technical diffusion targets and practical implementation metrics

Topics

Economic | Development

Unexpected differences

Emphasis on AGI vs. practical applications

Speakers

– Gong Ke
– Yutong Zhang

Arguments

The AI Plus action plan focuses on diffusion and adoption rather than AGI or chips, aiming to move AI from chat to products and services

Chinese companies focus on fundamental research and efficiency, achieving comparable performance with only 1% of resources used by frontier labs

Explanation

Unexpectedly, the academic representative (Gong Ke) emphasizes practical diffusion over advanced research, while the industry representative (Zhang) stresses fundamental research and innovation. This reverses typical expectations about academic vs. industry priorities

Topics

Economic | Development

Overall assessment

Summary

The discussion shows remarkably low levels of fundamental disagreement, with speakers generally aligned on China’s AI development direction. Main differences center on implementation approaches rather than core objectives

Disagreement level

Low to moderate disagreement level. The speakers share common goals of AI diffusion, economic value creation, and inclusive development, but differ on specific strategies and emphasis areas. This suggests a mature, collaborative approach to AI development in China with room for multiple complementary strategies rather than conflicting visions

Partial agreements

Economic | Development

Takeaways

Key takeaways

AI is not in a bubble and offers substantial economic value, with China expected to capture about 25% of the projected $15 trillion global AI contribution by 2030

China’s AI Plus action plan strategically focuses on diffusion and adoption across industries rather than AGI development, targeting 70% AI agent adoption by 2027 and 90% by 2030

China’s unique advantages include huge manufacturing/retail sectors providing scalable environments, exceptional openness to new technology (85% support autonomous driving), and a vibrant ecosystem with hundreds of model companies

Efficiency is critical for Chinese AI development – companies achieve comparable performance with significantly fewer resources (1% compared to frontier labs) through fundamental research and engineering optimization

AI integration spans all business functions and industries, from programming and accounting to healthcare and retail, with emphasis on productivity enhancement and cost optimization

China is building sustainable AI infrastructure through the ‘Dong Shu Xi Sun’ initiative, using renewable energy in western regions to power eastern data processing needs

AI education is being embedded nationwide from primary school through university with comprehensive teacher training programs based on UN frameworks

The workforce impact shows net positive employment with 5 million new AI-capable jobs created, though traditional roles may be displaced requiring upskilling programs

Future organizations will prioritize general intelligence over specialization as AI provides on-demand domain expertise, making learning ability more valuable than past experience

Resolutions and action items

Nationwide upskilling programs planned to help workers adapt to AI-enabled roles and find new AI-empowered tasks

Comprehensive teacher training programs implemented to support AI education in schools

Infrastructure development continuing with green energy focus in western China to support AI computing needs

Moonshot AI planning to launch a new model soon as mentioned by Yutong Zhang

Unresolved issues

How to help students use AI for deep thinking rather than accepting instant answers remains a significant educational challenge

The full extent of negative impacts on traditional jobs and education abilities requires ongoing assessment and monitoring

Uncertainty about when the next major AI breakthrough (similar to ‘deep sea moment’) might occur

Long-term effects of AI adoption on workforce displacement and the effectiveness of retraining programs

Balancing AI efficiency gains with maintaining human cognitive abilities and critical thinking skills

Suggested compromises

Model-agnostic approach allowing customers to choose different AI models for different purposes rather than forcing single solutions

Open-source culture in China’s AI ecosystem to drive down costs while maintaining innovation

This insight recognizes that AI education is happening organically through accessible tools rather than just formal education systems. It suggests a democratized learning approach that complements institutional education.

Impact

This comment expanded the education discussion beyond formal systems to include organic, curiosity-driven learning. It reinforced the theme of accessibility and diffusion that ran throughout the discussion, showing how China’s approach enables grassroots AI adoption.

Overall assessment

Explanation

The specific measurement criteria for these ambitious diffusion targets were not detailed and would require further research into policy implementation.

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.