"As machines process information increasingly well, what role will humans hold? What about us is impossible to copy?"
From the perspective of a systems thinking expert, the answer does not lie in a computational race against machines, but in understanding the shifts in organizational structures and the nature of decision-making.
1. The Truth About the "Pyramid": Hierarchy Is Actually a Way to Save Money
We often view the boss-employee system as a symbol of power. However, through an economic lens, organizations exist to optimize "Transaction Costs". This concept was introduced by Ronald Coase in 1945 to explain why firms are born instead of everyone trading freely on the market.
Take building a house as an example:
- In a perfect market, you could just snap your fingers and find the best craftsman.
- In reality, you must post advertisements, interview 10 candidates, negotiate salaries, and repeat this process every day.
The process of gathering information and negotiating is extremely expensive but does not directly lay a single brick. That is the transaction cost.
Companies are designed in a pyramid shape to minimize these costs through control and predictability. Specifically, according to Taylorism (Scientific Management), old organizations performed a harsh separation: Conception (Thinking) and Execution (Doing). The boss at the top holds the "thinking" role, while middle management transmits orders downward for employees to "do".
This model helps achieve "Economies of scale": the larger the organization, the lower the transaction cost per person. However, AI is now "invading" the very "Thinking" part that was once the privilege of management, forcing us to redefine this structure.
2. The Lesson from the Plate of Spaghetti: Distinguishing Between "Complicated" and "Complex"
To understand why AI is changing organizations, we must clarify a common linguistic confusion. In Vietnamese, there is often only one word ("Phức tạp") used for both Complicated and Complex. But in systems thinking, they represent two different worlds:
- Complicated: Think of a Boeing airplane with millions of components. Though massive, every part operates according to logical rules and is predictable. If a chip fails, we know the exact consequence. AI is extremely good at processing this "Complicated" nature.
- Complex: Think of a plate of spaghetti. The strands are interwoven randomly. When you stick your fork in, no computer can accurately predict how many strands you will pull up. This is a system that cannot be fully predicted by mathematics.
Modern organizations are no longer silent Boeing machines but volatile "plates of spaghetti". AI can solve complex logical problems, but humans are the entities who must face and make decisions among the "messy" (complex) strands of emotion and social context.
3. The Toyota Nudge: From "Supervision" to "Sense-making"
Management history has shifted power from the top of the pyramid to the "edge" of the organization. Toyota proved this through the "Andon Cord". At their factories, every worker has the authority to pull this cord to stop the entire assembly line if they detect an error. They are empowered to make decisions because they are the ones closest to reality.
This shift completely changes the role of the manager:
- From Supervision: Monitoring employees as if they were "cogs" in a machine (the Taylorism mindset).
- To Sense-making and Gardening: Modern managers do not "pull" a tree to make it grow faster. They are like gardeners, focusing on creating the right environment (soil, water, light) so that employees can make their own decisions.
The human task in the AI era is not to manage mechanical performance, but to create a "safe cognitive space" where individuals can understand and navigate the complexities of their work.
4. Every Decision Starts With an Emotion
We often pride ourselves on being rational creatures, but the story of Dr. Damasio's patient shattered that myth. After a brain injury caused the loss of emotional capacity, this patient — despite having sharp logic — was unable to make any choices at all, even something as simple as picking a meal.
In reality, the decision-making process is a "spiraling dance" within the brain:
- Emotion ignites a tendency.
- Reason is mobilized to build arguments and evidence to reinforce it (acting like a coat over the emotion).
- Emotion (belief in those arguments) drives the final action.
"Every decision starts with an emotion."
AI can support the "Thinking" part (data and logic), but it lacks the "emotion" to initiate a decision or take responsibility for its consequences. This is the only "Human" part that cannot be transferred.
5. The 2+3 Model: The More You Share, The More Value Grows
In an organization, there are always five transaction flows, divided into two groups with completely opposite natures:
- Market Flows (Financial, Material): These operate under the law of scarcity. If I give you 1 million VND or a pen, I no longer have it. These are "depleting flows" when given away.
- Non-market Flows (Information, Emotion, Trust): These operate under the law of abundance. When I share knowledge or empathy with you, that value is not lost; instead, it grows on both sides.
In the AI era, where information processing has become incredibly cheap, Trust becomes the center of every system. AI can generate infinite content, but it cannot "own" trust or reputation.
"Without trust, AI cannot help."
AI is liberating us from the rigid pyramid structures of the industrial age. It handles the "complicated" work so that humans can focus on meaningful "complex" (messy) strands. AI can support decisions, but humans remain the ones with the ultimate responsibility.
Summary from the "Organizing Work in the AI Era" workshop at UEH University — 13/01/2026 by Mr. Duc Ha Duong
#AI #FutureOfWork #SystemsThinking #Leadership #Trust