The AI they cannot build

I recently returned from Singapore, where my daughter-boss, Rachel and I attended SuperAI, Asia’s largest AI conference at Marina Bay Sands. The event hosted 10,000 attendees, 150 speakers and nearly every major technology company, including Google, OpenAI, Microsoft, Arm and Mistral. Companies that once seemed futuristic are now actively networking on the convention floor.

I attended keynotes on AI agents, frontier models, robotics, planetary intelligence and the economics of compute. Experts discussed the rapid pace of change, the shortening relevance of skills and the emerging divide between the AI-literate and the AI-inexperienced. After two days, my key takeaway was not a technical insight, but a reminder of something fundamentally human.

Matt Abrahams, a professor at Stanford’s Graduate School of Business, has taught strategic communication to executives for nearly 20 years. He does not dismiss AI or downplay its impact, but he has observed a notable trend in his classroom. His MBA students frequently ask: With AI advancing, what will remain for us?

His answer is always unexpected.

He refers to it as Old School AI — the original AI: Genuineness and Influence.

This is not simply advice to “be yourself.” Abrahams makes a more nuanced point.

Content creation is now faster and more affordable than ever. Large language models can generate reports, presentations, marketing materials, meeting summaries and business proposals more efficiently and cost-effectively than humans. As a result, we are inundated with technically accurate but emotionally empty output, creating overwhelming noise.

When everything sounds the same, when every memo has that same smooth, frictionless AI-generated finish, the one thing that cuts through is the thing a machine cannot manufacture.

Are they real?

Do they care?

Can I trust them?

Those three questions are as old as human communication. They have not changed. And AI, for all its capabilities, cannot answer them convincingly because trust is not built by accuracy. It is built by presence. And presence is the one thing you cannot prompt.

Think about what AI genuinely cannot do. It cannot lean in when a colleague is struggling and make them feel acknowledged. It cannot sense the room and change when the energy shifts unexpectedly. It cannot sense the fear beneath the confident answer, the grief just under the professional surface or the doubt hiding behind the polished slide deck.

The leaders people trust, the colleagues people rely on, the speakers people remember — all share one pattern. Their communication is real. Imperfect, sometimes. But real.

That gap between the real and the generated is where your advantage lies.

Abrahams offers two skills he believes will become more valuable, not less, as AI becomes more prevalent.

The first is authenticity.

And he is careful to define it because authenticity is not a personality type. It is not something you either have or don’t. It is a practice. A discipline. Something you can learn and improve the way you learn an instrument or a language.

His pragmatic framework is three questions you ask yourself before any high-stakes conversation:

What do I want my audience to know?

What do I want them to feel?

What specific action am I requesting?

Know. Feel. Do.

Simple enough to remember. Powerful enough to change how you show up.

The second skill is influence through structure.

Our brains are not wired for lists. They are wired for sequences, for stories, for messages with a shape.

Abrahams uses a framework he calls:

What: What is the idea?

So what: Why does it matter to your audience, not just to you?

Now what: What is the concrete next step?

Three questions.

Works in a pitch, a feedback conversation, a team briefing or a difficult one-on-one.

The structure forces you to begin with substance and finish with movement. Because influence, he says, does not exist without someone making a shift.

Many people are nervous. I fully sympathize with them. Will they be replaced by AI? They wonder whether the skills they have spent twenty years building still matter.

Here is what two days in Singapore confirmed for me.

You do not need to out-AI the AI. That race is already lost.

The more AI generates content, the more visible the human behind the content becomes.

The more machines can produce information, the more valuable becomes the person who can make that information feel real, relevant and worth acting on.

I flew home from a room full of the smartest machines on earth, and the loudest thought in my head was the oldest one in the book:

Be human. Nobody in that room had built a machine that could do that. And nobody ever will.
So be human. It’s working out so well for us; AI is trying to copy it.

Grateful for the encouraging response and requests for another run of my one-day Level Up Leadership: Agile. Able. Adaptive. seminar-workshop. Join us on Aug. 26, for a practical and inspiring learning experience designed for leaders, managers, entrepreneurs, business owners and professionals who want to lead with greater clarity, confidence, courage and competence in a disrupted world. Seats are limited. For inquiries and registration, contact April at +63 928 559 1798 or Sylene at +63 976 638 8974. Visit www.levelupleadership.ph.