They were the excel whizzes and the PowerPoint masters, those who could synthesize a rambling meeting into a sleek deck before lunch, or provide the minutes of a meeting held in English, but explained in your native tongue.
In my years overseeing leadership talent reviews for MNCs in Asia, managers would often identify, perhaps over-emphasize those kind of qualities when nominating early-in-career “high potentials”. Articulating complexity quickly were, for many, initial signals of leadership.
AI has arrived and is turning those signals into noise.
Now, with AI, anyone can now write in foreign tongues or summarize 100-page documents in seconds. But it has also made mediocrity effortless. We are drowning in AI slop, overwhelmed by overly-lengthy “summaries”, and endlessly checking AI output for inaccuracies.
The New Bottleneck
The bottleneck has shifted from producing to selecting. When anyone can generate a neat deck, a fluent email or a confident summary in seconds, those outputs stop signaling quality or leadership. The scarce skill is deciding what deserves attention, what standard is acceptable and what consequences you will accept when the output is wrong.
This is critical thinking. This is judgment. Good leaders are the ones with sharper critical thinking skills AND astute judgment. Critical thinking without judgment is analysis paralysis. Judgment without critical thinking is impulsive action. But a harmonious blend of the two lead to sound decision making.
Ever since foundation models put rapid analysis and drafting on free-flow tap, the bottleneck in many workflows has shifted to higher-level human capabilities. It has moved from finding and organizing information to choosing a direction, making tradeoffs, and taking action.
If you are strong at thinking critically and making sound judgment calls, you can open up new strategies, build new products, slash workflow inefficiencies. If you are not, you become the chokepoint, drowning in data and indecision.
The Human Premium
When intelligence becomes instant, abundant, and cheap, the “average” worker stays average, but a smaller part of the population – the AI-Augmented Leaders – begin to scale output significantly. They aren’t working harder. They are leveraging the available tools more effectively.
As the framework below illustrates, the AI-Augmented Leader uses AI as an “amplifier” for analysis and drafting, but differentiates on human capabilities. AI can process infinite data, but it lacks “skin in the game.” AI doesn’t feel the weight of a mistake or the heat of a crisis. Ultimately, we only trust those who share our risks and offer a level of authenticity that an algorithm does not have.
The framework identifies eight core human capabilities. In addition to the previously mentioned Critical Thinking and Judgment, three more stand out as the new bottlenecks of leadership: Direction, Influence and Agency.
- Direction: Direction is the capability to clarify what actually matters, and to commit to a long-term path when the data is ambiguous. AI can provide a map, but humans must choose the destination. And the unique, bold and especially the “unreasonable” goals are not going to be found in the AI model’s training set. Those goals are set by those Steve Jobs called “the crazy ones,” those betting against the status quo.
- Influence: AI is very skilled at persuasion techniques: active listening, non-judgmental responses, cognitive behavior therapy. But persuasion is not the same as trust. We instinctively reject “personal” expressions when we discover the source is an algorithm, because we know an algorithm doesn’t share our fate. Influence in the AI era goes hand-in-hand with authenticity. Humans want to be led by humans.
- Agency: Agency is the engine. It is the most critical tool in the AI-Augmented Leader toolkit because it is the rarest. Many people look like leaders on paper. They have the strategy, the synthesis, the spiffy slide deck. But Agency is what separates the people who can explain what should happen, from the people who make it happen. It is the grit to own a decision under uncertainty, knowing the tradeoffs, and doing the unglamorous follow-through when the “perfect” AI plan hits the messy reality of the physical world.
Mastering these human capabilities at the corporate level can lead to competitive advantage and market differentiation.
AI-Augmented Leader as Orchestrator
When people talk about “AI transformation,” they often imagine a tool roolout. A company buys software, trains staff, and expects productivity to rise. In practice, the tool is the easy part. The hard part is changing how work moves through the organization. Without that change, AI often creates extra noise. Different teams generate different summaries of the same issue. Everyone has a different recommendation. No one is sure which answer is trustworthy. Progress is slow.
AI Augmented Leaders solve this by turning AI into a practical workflow. They break work into steps and decide what belongs to machines and what belongs to humans. AI does the heavy lifting where speed matters, like scanning for patterns, gathering evidence, drafting first versions, producing options.
Humans do the parts where responsibility matters. They decide which problems are priorities, choosing between imperfect options, giving the go-ahead on high-stakes actions, and dealing with exception cases.
Then these leaders make the workflow work across different teams. They build shared definitions. They establish rules for escalation. They define how decisions get recorded. They make it clear who owns outcomes.
The impact of the AI tool or systems will vary greatly depending on whether a company is orchestrated by AI-Augmented leaders or not.
Here are two Asian corporate cases in point, where leaders in these two companies have blended AI and human capabilities in harmonious concert to ensure customer’s needs are met. Both SHEIN and Grab reached scale through disciplined orchestration of human and machine strengths.
Two Cases in Effective Orchestration
SHEIN: Founded in China in 2008, SHEIN sells clothing directly to shoppers through its app and website. Their simple promise to customers is the latest, coolest fashions at low prices, with new styles appearing constantly. Behind the scenes, AI and data systems scan what people are paing attention to online. Then teams translate those signals into new designs and small initial production runs. If an item performs well, the supply chain ramps up quickly through a network of suppliers set up for short runs and fast reorders. If an item performs poorly, they stop early and move on.
The human part shows up in “taste,” what counts as “cool” and “fashionable” in the first place. Social media posts, influencer content and customer reactions shape what looks desirable, which items feel current, and which designs seem worth repeating. Leaders make the calls on which signals to trust, how aggressive to be on speed versus quality, and how to manage reputation and compliance risks, while keeping the AI-enabled business model moving.
Grab: Establsihed in Malaysia as a ride-hailing service, Grab grew into a regional super-app offering rides, food delivery and payments across Southeast Asia. In day-to-day operations, one recurring problem is account takeover and fraud. A bad actor gains access to someone else’s account and starts doing things the real owner did not authorize, like ordering food, taking rides or changing account details.
When this happens, Grab’s operations team have to figure out what happened quickly and accurately. The evidence is spread across many places, including login history, AI-supported device and network signals, transaction records, GPS traces, and customer support messages. Some of these messages and naotes are in different languages, becuase cases span multiple countries and teams. Grab has built an AI agent framework that follows a standard investigation playbook, pulls the relevant evident from internal systems and drafts a clear case summary for the human team.
When the case reaches a point where the AI system should not decide, such an exception, an escalation, or a high-stakes action, the workflow pauses so a person reviews, approves, or routes the case tot he right specialist. The AI tool enables evidence gathering, translating and summarizing at super-human speed, leaving the judgment calls, customer remediation, and accountability to the humans.
Three Chords and the Truth
In an age of instant intelligence, “being in the loop” is not enough. AI speeds up sensing, analysis, and coordination. Humans orchestrate, leveraging our capabilities to determine what to pursue, what standards to apply, which risks to accept, and how to follow through when conditions change. The SHEIN and Grab cases show the same pattern in different settings.
Someone once said that the fundamentals of Country Music are “three chords and the truth.” The AI-Augmented Leader model is similar: blending the logic and math of musical patterns with the everyday emotions and motivations of human experience, and living a life that sings.
Originally published on CKGSB Knowledge.
