A：The primary applications of AI chat mentioned in the speech are focused on finding product-market fit, using AI to enhance user experience through features such as better search integration, improved speech recognition, easier access to multiple models of capabilities, and exploration around group chat. These enhancements are expected to increase usage and retention of the AI chatbot app in the coming months.

A：AI integration has enhanced a range of existing user experiences within Wasim, including content consumption, information retrieval, merchandise recommendation, and customer service. AI agents are designed to autonomously interact on behalf of users with in-app functionalities, particularly in media programs, and the integration of AI aims to improve computer usage capabilities and daily life while also highlighting the risks associated with unleased unsupervised AI.

A：The vision for AI agents is to deliver AI productivity that is beneficial to the general public as well as early adopters, which is expected to boost ecosystem activity and naturally generate revenue. AI agents are currently powered by a variety of foundation models, and it is anticipated that users will continue to have access to a range of models, with improved performance potentially enabling new unique capabilities.

A：The company has introduced a number of AI tools for enhancing productivity, including Work Body Q, claw, ten, and cloud lights. These tools are designed to be used through downloadable skills from a platform called skill up, and they facilitate autonomous workflows and continuous task execution.

A：The company is significantly investing in new AI products, with investments of seven billion RMB in the fourth quarter of 2025 and an additional eighteen billion RMB for the full year. The investments are expected to more than double in 2026, funded from increasing earnings from core businesses. These investments are considered strategically similar to investments in capital expenditures and are necessary to build the necessary foundation to unlock new value beyond ongoing operating expenses.

A：The case study about Tencents cloud highlights its development into a market leader with economic returns over time. Tencent Cloud started as a relative late entrant but committed to a patient and long-term investment strategy, gaining IT scale due to Tencent's status as China's largest single user of technology infrastructure. It has become the first cloud service provider in China to support AMD's recent generations of CPU, partnered with AMD as the largest partner in the country, and holds a leading position in the cloud video streaming service. Despite an initial focus on revenue growth, Tencent Cloud restructured to focus on high-quality services and achieved an operating profit, grew revenue, and sharply improved earnings in 2024.

A：The company expects to generate significant returns over the long run on its investment in new AI products. The initial investment in AI, such as AI chatbots and APs, is viewed as a fixed sum of cash investment necessary to incubate successful new businesses that will ultimately generate good economic returns similar to the review of the initial investment in new AI.

A：The latest business segment results show a nine billion RMB increase in value service revenue year on year to 99.5 billion RMB, with social network revenue growing three percent year on year to 31 billion RMB. Domestic games revenue grew 15 percent year on year to 49.9 billion RMB, and international games revenue increased 32 percent year on year to 20.6 billion RMB. Marketing services revenue increased 17 percent year on year to 21.1 billion RMB, and fintech and business services revenue grew 31 percent year on year to 24.5 billion RMB.

A：Tencent's major games have shown substantial growth in performance and receipts. Doubt, a force, sustained among the top three games in the industry with over 50 million peak daily active users and lifetime height monthly gross receipts. Battering P. C, a quality shooting game, achieved lifetime high growth receipts in 2025. Clash Real, a large mobile game, reached record DAU and DCR receipts in February. Warframe, an action game, launched major updates and achieved life time highs in DAU and MDAU in December 2025.

A：The increase in financial technology revenue was driven by wealth management and commercial payment services, with a sustained positive year-on-year growth in commercial payment volume supported by a higher number of transactions and improved value put transactions for wealth management.

A：Cloud services revenue increased due to a better pricing environment and the tight supply of memory and CPU industry-wide. Additionally, demand contributed to the on-year growth rate with short video platforms and AI video generation services increasingly using media processing solutions for streaming video and audio.

A：The operating profit for the quarter was RMB108.3 billion, which is an increase of 9% year on year.

A：Revenue from other games decreased to RMB13 billion compared to RMB25 billion in the same period last year due to lower subsidies and tax rebates.

A：Findings calls were RMB3 billion, up from RMB2.5 billion in the same quarter last year. Operating profit was RMB60.3 billion, up 17% year on year. Net profit after tax was RMB69.25 billion, up 17% year on year.

A：The gross margin for the quarter was 96%, which is a three percentage point increase from the previous year's margin of 60%.

A：Operating expenses and marketing expenses were RMB13 billion, up 26% year on year. This increase was attributed to higher promotional efforts to support the group's native application, with ID expenses rising by 20% on year.

A：The company had a personnel increase of 116,000 employees, up 5% year on year. The operating margin was 36%, which is a 11% increase year on year. The investment in severe infrastructure year-over-year operating capex decreased by 51%, and free cash flow increased over six times year on year.

A：The company proposed an annual dividend of RMB5.30 per share, reflecting an 18% year-over-year increase. The dividend will be payable to shareholders on the first of December 2022.

A：Profit from existing business is expected to grow to cover incremental AI investment. The gap between revenue and profit growth into 2026 should be considered with long-term AI investment in mind. The company has been actively staffing up with high-quality talent and has a plan in place for resource allocation and monitoring progress in AI deployment, including a focus on the development of a state-of-the-art language model and user engagement. The company plans to recruit selectively to address the constraint in GPU availability and is progressively increasing computing power, especially in the second half of the year.

A：The two main areas of investment for supporting business growth are the financials with existing businesses and the investment into AI for supporting these businesses.

A：New AI products might initially be investments with a later return because there is not yet a market in China for consumers to pay for AI products or companies to pay for AI agents at a high cost. However, over time, it is expected that these new AI products will be able to generate revenue and provide an attractive return.

A：Tensor cloud is initially positioned as a business where investments include losses, but over time it is expected to turn into a property business and generate revenue for the company.

A：The company prefers to buy due to its strong balance sheet rather than paying additional margin for leasing, although sometimes due to supply chain constraints and regulations, they might have to rent.

A：The company's approach to the different layers of the AI stack is dynamic and they invest in all layers, recognizing that it is difficult to determine which layer will be more important than the others in a fast-moving market. However, the company has the resources and a team to invest in all these layers, and they have built strong teams for specific tasks such as model capability, orchestration capability, and the ability to go across devices.

A：The company views the parallel emergence of different AI models as an exciting concept where multiple strong models specialize in different kinds of activities. They believe that there is not a single model that will dominate, and instead, a decentralized regime will allow for many companies to have their own models, supported by cloud infrastructure, to cater to unique user needs and win in the market.

A：The company's strategy for leveraging the agent and API capabilities is to deploy them in a way that aligns with their existing capabilities and interests. They believe that their diverse range of centralized and decentralized ecosystems positions them well for the new agent and API services. In terms of preventing other large language models from undermining their foundation models' value, the company suggests that the choice is available to use models with varying performance and price points, and they have the confidence in their team's capabilities to advance and improve their offerings.

A：The company intends to be one of the players in the physical AI market but not the only one. They believe they are in a unique position to provide the data to train models for physical AI applications due to their expertise in 3D models and their involvement in video game development.

A：The speaker believes that AI will accelerate the demand for the usage of world models, such as 3D models, and enhance capabilities in industrial design, architecture, and video. They view AI as a significant but not the biggest opportunity ahead of them, with many large and more immediate opportunities currently existing.

A：AI is currently being used to enhance the content creation within existing games, as discussed at the Game Developers Conference. However, the speaker notes that AI has not yet reached the capability to create games completely from scratch due to several reasons that were not detailed in the transcript.

A：Tencent acknowledges the surge in demand for AI compute power and the resulting price inflation for server AI servers. They note that the industry is responding with pricing adjustments and that they aim to deliver more value to customers. This includes potentially passing on the cost inflation to customers, with a focus on optimizing and enriching the services they provide to achieve better pricing and margins.

A：Tencent's strategy is to enrich their services by delivering more value through optimization, which could involve renting out computing power at better prices, subdividing and virtualizing IT resources, and potentially bundling services into platform as a service or software as a service offerings to improve pricing and margins. This aligns with their journey from录 losses to substantial profits, with the intention to continue moving towards better monetization and profitability.

A：The strategy to remain competitive in AI involves treating AI as a new frontier with multiple opportunities rather than a single game. This allows for innovation in various aspects of AI packaging from models to products to agents. The existing services will also have new agents' capabilities on mobile and PC. Fundamental capabilities, diverse infrastructures like security and cloud, and the ability to innovate quickly are crucial for success in these new AI opportunities.

A：Apple's recent announcement will affect revenue sharing with game developers by potentially increasing the revenue share based on gross revenue rather than net revenue after Apple's cut. This is especially relevant for a small minority of game publishers where revenue share is usually calculated on net revenue. Moreover, if the profit is reinvested into new AI products, the incremental profit stream will be subject to corporate income tax, which could be a factor for Apple to consider in its financial strategy.

A：The key takeaway from Apple's announcement is that the App Store will effectively offer developers in China the same lower rate that is offered to developers in other parts of the world. This is in line with industry trends and is expected to lead to a normalization of reduced take rates across different regions, including China. This is viewed as a positive development, marking the start of a multi-step journey that could eventually lead to more uniform take rates globally.

A：The significance of the upcoming WWDC event for Apple is to address the interest in AI chips and discuss the priorities concerning in-house chip development. The focus at this moment is not considered to be on developing in-house AI chips, as the emphasis is on utilizing the best training chips available to train the best models. The market for inference chips is seen to have many options and lower margins, with an expectation that the margins will be manageable. Apple's strategy is to leverage these chips effectively and not to concentrate on reducing the cost of inference at this stage. Instead, the focus is on utilizing the best training chips and iteratively improving model training, with the goal of eventually reaching state-of-the-art (SOTA) models. The priority is on product development and integration capabilities to design exciting AI products for users.