AI Acceleration: Three Stages of Transformation with Distinct Focuses
"Each stage has its unique characteristics," he explained in detail to the reporter. In the enablement stage, the focus is primarily internal, centered on reducing costs, optimizing internal processes, and managing risks. "At this stage, companies should concentrate internally, aiming to automate repetitive tasks, minimize human errors, and enhance operational efficiency—essentially solving how to empower employees to work faster, better, and more cost-effectively. The technologies adopted here are usually mature and market-proven."
The integration stage is externally driven: the core focus shifts to leveraging technology to create new revenue streams, enhance products/services, and improve customer experiences. Companies pivot from internal optimization to external markets and clients, addressing the question, "How can we deliver greater value to customers?" Here, AI and similar technologies become integral features of products and services rather than just internal tools.
The evolution stage is ecosystem-driven: the goal is to use technology to fundamentally transform business models, redefine industry standards, or even create entirely new markets. Companies at this stage no longer just sell products but build platforms that attract developers, businesses, and users to collaborate, forming a robust ecosystem. This phase also requires embracing uncertainty and investing heavily in forward-looking, high-risk R&D.
Jiang Liqin emphasized that companies should treat these three stages as a dynamic, potentially overlapping strategic portfolio. To keep pace with technological adoption, they must focus on multiple fronts: "Companies must honestly assess which stage they’re in—whether they’re still struggling with efficiency, seeking growth, or nurturing ambitions as ecosystem builders. At the same time, they should manage new technology applications like an investment portfolio. For instance, allocate most resources to the enablement stage to stabilize core operations; dedicate a portion to the integration stage to explore new growth opportunities; and invest minimally in the evolution stage to future-proof the business."
Moreover, regardless of the stage, high-quality, easily accessible data is the foundation. Building a robust data platform and governance framework is a prerequisite for technological application. Companies must also form agile, cross-functional teams to enable rapid testing, learning, and iteration. Ultimately, technology is driven by people, so businesses must also focus on talent acquisition, development, and retention.
The reporter learned that KPMG will soon release its report, "New Intelligence Ignites New Quality: Generative AI Empowers Industrial Transformation—Practices and Pathways," during the 2025 World AI Conference. Jiang Liqin noted that from the perspective of front-, middle-, and back-office functions, generative AI’s potential applications are nearly ubiquitous. Based on current adoption trends across industries, these use cases fall into four categories: front-office business enhancement, middle-office decision-making, back-office management, and general-purpose tools. In practice, generative AI’s core functionalities include content generation, reasoning and Q&A, data augmentation, and interaction innovation.
He also pointed out that many companies hesitate to transform due to a lack of a clear, actionable, and long-term roadmap—from top-level strategy and scenario planning to seamless collaboration between technology and organization, all the way to execution.
