The first AI Native from Dr. Kai-Fu Lee When “Intelligence” became something accessible that anyone has access to

AI Native: Kai-Fu Lee's Vision from Techsauce Global Summit 2026

Anyone following recent developments in technology will likely be familiar with the name Dr. Kai-Fu Lee—AI researcher and investor, former President of Google China, and bestselling author of AI Superpowers. At the latest Techsauce Global Summit 2026, he took the stage with an opening line from his new book, AI Native: The Mandate to Transform Your Company. Below is a summary of his talk and how AI is reshaping the business world.

AI isn't just another industrial revolution—it's making intelligence nearly free

Kai-Fu Lee began by comparing AI to the Fourth Industrial Revolution, but argued that framing undersells AI's true potential. He laid out the progression as follows:

• Machines replaced manual labor
• The information age made knowledge more accessible
• The AI age is making both intelligence and execution remarkably affordable — in many cases, free

When intelligence becomes something everyone can access at little to no cost, innovation accelerates. Lee identified four reasons AI is growing faster than any technology before it:

• Knowledge-based first: AI automates white-collar work before it moves into blue-collar work
• Rides existing infrastructure: It scales on top of the internet and mobile internet, making adoption even faster than platforms like TikTok or Instagram
• Massive investment: Models with the same capability have gotten roughly 175 times cheaper in just two years
• AI teaching AI: AI systems can now train and improve other AI systems without waiting on human instruction

The Agentic Era and why so many organizations are getting it wrong

We are now entering what Lee calls the Agentic Era or the era of AI agents, which differ fundamentally from ordinary chatbots. Rather than simply generating answers, agents can write code, use external tools, and act autonomously in the real world, much like a personal assistant.

The problem, Lee found after speaking with CEOs around the world, is that most organizations don't actually know how to use AI effectively.

Many CEOs proudly claim their companies have built 200 to 800 AI applications. But when asked how any of that has increased profit, most can't answer—because what they've built are mostly surface-level integrations, like a chatbot that answers customer FAQs.

So why does deploying AI often end up costing companies more than it saves?

Many companies purchase AI models and see no meaningful results after months of use. The core issue: the AI lacks context. It doesn't understand the company, and it doesn't understand how the business actually operates day to day. Feed it too little information and it becomes unreliable; feed it too much and it becomes useless. This is the classic problem of "garbage in, garbage out."

The solution: an Enterprise Universe Model and "Boss AI"

To solve this, Lee proposes that companies build what he calls an Enterprise Universe Model, built on three key pillars:

• Ontology: Organizing a company's scattered information into a structured map that AI can actually understand
• Multi-agent systems: Deploying multiple specialized AI agents that work together to solve problems, similarly to how a CEO assembles a team of experts from different fields
• Boss AI: A system that gives the CEO real-time visibility into what's happening across the company including what's discussed in meetings, where key decisions are actually made to solve surfacing problems that would otherwise stay hidden beneath the surface

AI vs. human: left brain vs. right brain

Many worry that AI will eventually replace the CEO. Lee's answer was simple, it won't.

• AI as the "left hemisphere": excels at analysis, logic, and processing information, often even better than humans
• Humans as the "right hemisphere": bring aesthetic sense, empathy, vision, and the ability to make decisions about things that have never happened before. Imagine Steve Jobs creating the iPhone, or Elon Musk building an electric car from scratch
• What AI cannot do is bear responsibility. When AI makes a mistake, you can't fire it, cut its pay, or hold it accountable in any meaningful way. That burden must remain with a human.

Restructuring companies around the DRI (Directly Responsible Individual)

Lee predicts the traditional corporate pyramid will gradually disappear, replaced by the DRI (Directly Responsible Individual) model — essentially, a "mini-CEO" structure.

What is a DRI?

In the old hierarchy, a new project would flow from CEO → VP → Director → Manager → individual workers. Along the way, details get lost, time is wasted in meetings, follow-ups drag on, and instructions get diluted at every handoff. Most workers at the bottom have no real authority to make decisions.

In the new model, the CEO assigns a single person as the DRI with a clear goal (for example, "increase sales of Product A by 30%"). That DRI is given decision-making authority, a budget, a set of AI agents (for coding, marketing, data analysis, etc.), and one or two human collaborators. This dramatically shortens the workflow. There's no more hierarchy to navigate—the DRI owns the outcome end-to-end, taking credit for success and full responsibility for failure.

As companies cut out middle layers of management, people currently in those roles face four possible paths:

1. Become a DRI (mini-CEO): Someone with vision, leadership instincts, and the ability to direct AI effectively can take ownership of a project and be accountable for its success, much like a CEO.
2. Become a "Super Coach": Since many DRIs are skilled at working with AI but not necessarily skilled at working with people, this role focuses on people management—supporting teams, resolving conflict, and providing the human touch DRIs may lack.
3. Become a transitional bridge: Someone who helps the organization move from the old system to the new one—running meetings, updating documentation, adjusting workflows. The catch: this role has a shelf life and will likely disappear once the transition is complete.
4. Get flattened out: People whose job was mainly to relay instructions, follow up on tasks, or manage meetings, works that AI agents can now do on their own, will find themselves quickly obsolete.

Epilogue: this isn't a smaller slice of the pie—it's a bigger cake

Lee closed with a reframe: we shouldn't think of AI and humans as competing for the same fixed slice of economic value. If AI handles the "left-brain" work and humans focus on the "right-brain" work, the overall pie—GDP, could grow 5 to 10 times larger than it is today.

Work that depends on human connection and genuine creativity will always belong to people. The most important thing right now, Lee argued, is to stay open-minded and start learning to work alongside AI to build that future together.