As the domestic debut of the AIGC long-form drama "Journey to the West: The Sequel" heats up the market, investor focus is shifting from booming compute hardware to downstream AI applications with real commercial potential, signaling a pivotal transition from concept hype to business validation in the AI industry. So, what narratives deserve attention in AI applications, and how can investors identify targets with true commercial viability? Great Wall Fund's international business department deputy general manager and fund manager Qu Shaojie shares his analysis below.

Qu Shaojie notes that large models and compute hardware form the technical foundation, but only when AI is genuinely embedded in business processes, solving real pain points and delivering value users are willing to pay for can the entire industry establish a positive feedback loop. If AI remains confined to technical demos and pilot concepts without deployed applications, the investments in compute and models cannot close the commercial circuit, making industry expansion unsustainable. The rotation from hardware to applications essentially reflects the market's shift from speculating on technology supply to verifying the actual value technology creates. In practice, AI has already left the lab, achieving scale deployment in coding assistants, video generation, pharmaceutical research, and intelligent office tools—from full-chain support in code generation and debugging to industrialized production of digital content, and from drug target discovery to automated document processing and business workflows. Many scenarios have moved beyond PPT concepts into enterprise-level deployment.

Recent US earnings reports show AI applications are no longer just stories; they are translating into revenue. At the same time, the four major cloud providers are sustaining high growth in AI cloud services, with related cloud revenue climbing rapidly—a side reflection of genuine enterprise demand for compute and AI services. The prosperity of downstream applications is feeding back to boost cloud infrastructure momentum. Qu Shaojie emphasizes that annualized recurring revenue (ARR) is currently the most important validation metric for assessing AI application value. ARR represents sustainable, renewable annualized income, distinct from one-off projects or short-term pilot orders. Targets capable of generating consistently growing ARR are likely the best direction in today's AI application space.

On the recent surge of AI short dramas, Qu Shaojie observes that these productions have already achieved a complete commercial loop. AI dramatically compresses costs and production cycles for scripts, visuals, and audio, with real revenue-sharing and paid subscriptions emerging both domestically and overseas. Some companies are already reporting revenue increments from AI short dramas in their financials, with overseas paid models performing particularly strongly. However, the explosion in AI short drama supply has not significantly expanded total user time; instead, it is largely competing for existing users within the pools of short video, online fiction, and traditional film and television, diverting audience attention and consumption hours. Media companies that proactively embrace AI and treat short dramas as a new business line have the opportunity to unlock second revenue streams and achieve dual improvements in valuation and profitability. Traditional media firms that resist change and fail to build AI content capabilities face a double squeeze: AI-generated content captures user time, while AI-driven cost restructuring amplifies the disadvantage of traditional live-action production models. These companies not only miss out on AI dividends but also suffer from technological disruption, further pressuring their earnings.

When comparing the two main AI application tracks—domestic ToB enterprise services and vertical industry transformation versus overseas AI applications such as tools and content—Qu Shaojie believes it is not a binary choice. ToB vertical AI transformation targets government, manufacturing, and financial clients, emphasizing industry expertise, private deployment, and quantifiable ROI from cost reduction and efficiency gains. Revenue often comes from project-based or subscription models, with longer implementation cycles but stronger customer loyalty. ToC applications, particularly in consumer and entertainment sectors, should focus on rapidly acquiring end users to scale quickly. From a market perspective, both domestic and overseas markets deserve attention. China's AI model capabilities can compete on par with the best overseas models; leading domestic models have significantly narrowed the gap, showing strong competitiveness in long-document processing, multimodal generation, and reasoning cost efficiency. The domestic market's advantages lie in local industry scenarios, data compliance, and government-enterprise client resources, while overseas markets offer stronger willingness to pay and mature subscription business models. Companies should avoid staking everything on a single market; using the domestic base as a foundation while expanding overseas for incremental growth is likely a more prudent strategic choice.

On valuation increases and compliance issues in AI application investments, Qu Shaojie advises against following speculative trends and instead focusing on real ARR growth rates, gross margins, and financial sustainability. Priority should go to targets where AI-driven incremental revenue is verifiable, customer payments are genuine, and unit economics are improving—while steering clear of concepts lacking financial data support.

Disclaimer: The information in this communication is sourced from channels the company deems reliable and the personal judgment of research staff, but the company makes no direct or implied warranties as to its accuracy or completeness. This communication does not constitute a full description or summary of any securities or markets, and any opinions expressed may change without notice. It should not be used by recipients as a substitute for independent judgment or as a basis for investment decisions. The company, its affiliates, employees, or agents assume no liability for any actions taken based on this content or any losses arising therefrom. Without prior written permission from Great Wall Fund Management Co., Ltd., no one may distribute, reproduce, or publish this communication in any form, nor make any deletions or modifications contrary to its original intent. Fund managers remind all citizens of their duty and right to report money laundering crimes and to strictly comply with relevant anti-money laundering laws and regulations. Market risk exists, and investment requires caution.