AMD's Dr. Su in Conversation with Kai-Fu Lee: AI Transformation Must Be CEO-Driven, Future "DRI" (Directly Responsible Individual) Will Be the Core of Enterprises | Live from the Event

Key Themes and Vision

1. AI as a Universal, Transformative Force

Lisa Su's keynote frames high-performance computing and AI as technologies that "benefit humanity", capable of solving the world's most significant challenges. This is not merely corporate rhetoric—it signals AMD's strategic positioning. Su's statement that AI should be "everywhere—for every person, every workload, every device form factor" reveals a critical insight: the future of AI is not confined to data centers. It spans the full computing stack, from cloud to edge to personal devices.

• AMD touches billions of users daily across data centers, PCs, and edge devices.
• The company is investing in open-source ecosystems and full-stack innovation to democratize AI access.

This vision implies that AI's next growth phase will be driven by diversity of application, not a single dominant model or platform. Su cites projections that global AI active users will grow from over 1 billion to 5 billion, emphasizing that varied models and workflows—not monolithic solutions—will fuel this expansion.

2. China as a Strategic AI Hub

Su's emphasis on China is both strategic and substantive:

• AMD has operated in China for over 30 years.
• Its Shanghai R&D center is one of AMD's largest globally, covering chip design, AI software, and platform engineering.
• Multiple AI Centers of Excellence have been established, alongside partnerships with leading Chinese cloud providers.

Su's characterization of China as a leader in "open innovation" is noteworthy. It suggests that China's ecosystem advantages—scale, speed, and collaborative openness—are recognized by global tech leaders as essential to accelerating AI development. This framing positions AMD not as a foreign vendor but as an embedded partner in China's AI ecosystem.

The Dialogue: From Generative AI to Agentic AI

3. Two Critical Tipping Points (Kai-Fu Lee)

Lee identifies two transformative shifts that define the current moment:

First: AI coding has crossed the threshold.

• A year ago, AI could write a single function.
• Today, it can build entire functional modules and end-to-end products.
• This matters because, in the digital world, everything an agent does is fundamentally code. When AI can write complete code autonomously, self-governing agents become feasible.

Second: Multi-agent architecture is the next breakthrough.

• A single agent, no matter how large the model, has inherent limitations.
• Lee draws a powerful analogy: just as investment committees and boards of directors achieve what no individual can, AI agent teams can surpass any single agent's capability.
• A multi-agent system might include agents specializing in planning, review, execution, and risk control—debating and refining each other's work.

4. The "Agentic AI" Paradigm

Lee articulates a clear evolutionary timeline:

This progression culminates in what Lee calls the "one-person company" model. With modular multi-agent frameworks, a single developer can act as a macro-architect, launching and operating a fully functional business. This is not science fiction—it is the practical implication of Agentic AI.

The key paradigm shift is this: you no longer give AI a prompt; you give it an organizational goal. The agent system then autonomously coordinates, executes, measures, optimizes, and closes the feedback loop. This represents a fundamental change in how humans interact with AI—from instruction-following to goal-delegation.

5. The CEO as the AI Transformation Driver

One of the most provocative points from the article's title and dialogue is that AI transformation must be CEO-driven. Lee mentions ongoing conversations with global CEOs about how to execute this transformation. The implication is profound:

• AI transformation is not an IT project or an engineering initiative—it is a strategic, organization-wide transformation that requires top-down leadership.
• The concept of DRI (Directly Responsible Individual) as a future enterprise core suggests that in an AI-augmented organization, accountability and decision-making will be restructured around individuals who directly own outcomes, supported by AI agent teams.
• For developers in the audience, this represents a massive opportunity: helping CEOs and enterprises make this transition is itself a business opportunity.

6. Implications for Developers and the Industry

The dialogue between Su and Lee is ultimately a call to action for the developer community:

• Build the platforms: Lee's company, 01.AI, is focused on creating platforms for developers to build multi-agent systems.
• Think architecturally: The role of the developer is shifting from writing code to designing agent ecosystems.
• Embrace openness: Both speakers emphasize open ecosystems and collaboration as accelerators for AI progress.

Deeper Meaning

The conversation reveals a shared belief between AMD and 01.AI that we are at an inflection point comparable to the birth of the internet. The combination of powerful hardware (AMD's compute platforms) and sophisticated software (multi-agent AI frameworks) is creating conditions for AI to move from tool to teammate—and eventually to autonomous organizational operator. The challenge is no longer technical feasibility but organizational readiness and visionary leadership. This is why the CEO must drive the transformation: it requires rethinking the very structure of how enterprises operate.