Inhale. Hold it. Hold it. Hold it. HOLD IT!

Ninety-five percent — your colleagues, your competitors, your industry — are still sitting in rooms holding their breath for things to work out.

This is about the dangerous illusion of job security in an AI world.

Someone told me recently that I’m in the top five percent of AI users globally. I didn’t know whether to feel encouraged or unsettled. Because if that’s true, it means the room holding its breath isn’t an abstraction. It’s full of real people — smart, capable, experienced people — waiting for someone to tell them it’s safe to breathe. And last week, one of the largest industries in the world showed us exactly what that looks like in public.

What the Room Heard

San Antonio last week. The hospitality industry’s largest annual technology conference. Six thousand professionals. Four hundred exhibitors. Eighty-five thousand square feet of the latest technology the industry has to offer.

And five takeaways that were, at their core, things people hoped to hear. Maybe not what they need to hear.

That’s not a criticism of HITEC. It’s an observation about what happens when an entire industry arrives at the same moment of uncertainty at the same time.

The conference’s own summary said it plainly: AI has moved from experimentation to implementation. The future guest experience will be defined by choice. Digital workers are entering the conversation. Data is both the biggest opportunity and the biggest obstacle. And — perhaps most tellingly — human connection still drives the industry forward.

Yet related insights had the feel of textbook reassurances.

We’re moving forward!

Guests will still choose us…

The robots aren’t taking over.

The data problem isn’t our fault, and

People still matter.

A robot named Nylo was asked directly whether machines were coming for human jobs. Its answer — “no jobs stealing, just job boosting” — drew laughs and, if you were watching carefully, something that looked a lot like relief.

That’s not a punchline. That’s false relief. The room exhaled — and nothing changed.

The Cost

Here’s what I keep thinking about.

The accepted practice in every industry conference — not just hospitality — is to frame the moment as progress. Where we’ve been. Where we’re going. We congratulate ourselves for having the conversation. Then we go home feeling better than when we arrived.

That’s not wrong.

Humans have always managed uncertainty through community and shared narrative. It’s one of the things we do best.

But there’s a cost.

When the primary function of the conversation shifts from understanding the problem to managing anxiety about the problem, the real problem stays unsolved. It just gets better-named.

And the real problem — the one nobody in that room named directly — is this: AI isn’t coming for your job the way the movies suggested. It’s doing something quieter and more permanent. It’s shifting the ground beneath the people who never saw it as their problem to understand.

Only 6 percent of hotels appear in AI search results. That number surfaced at HITEC, briefly, and the room’s practical response was: make sure your information is consistent, generate reviews, respond to guests. Good advice. But it’s the advice of damage control, not of someone who has actually grappled with what it means when the discovery layer of your entire industry quietly shifts underneath you.

That’s the dangerous illusion.

Not that AI is coming. But that paying attention to it occasionally — attending the conference, nodding at the keynote, updating your information — is the same as understanding it. It isn’t. And the gap between those two things is where job security goes to die quietly, without announcement.

The Comfortable Answer

Either/or thinking has a long and comfortable history in every industry. Technology versus people. Efficiency versus experience. Human versus machine.

HITEC 2026’s mature, sophisticated answer to all of it was: both. Human plus machine. Efficiency and experience. Technology that serves people.

Which is true.

And also insufficient. Because both/and thinking, without clarity about which both and which and, is just a more comfortable version of the same confusion. It lets everyone leave the room feeling like they participated in the solution. Most of them didn’t.
They participated in the exhale.

The Ones Who Weren’t There

There is a different kind of person who was not in that room last week. Tech engineers.

They didn’t need to be. They don’t typically attend the super events. They don’t require the shared narrative or the false relief. They were never anxious in the same way — because they were never managing the same fear.

They build the infrastructure that the room depends on. They understand what the system actually does, not what the conference says it does. When ninety-four percent of hotels vanish from AI search results, they don’t reach for a reassurance. They reach for the architecture. They want to know why the pipe is broken — not how to feel better about the leak.

They have always been this way. Quietly. Consistently. Without needing a keynote to tell them the moment has arrived.

The moment, for them, is structural. It always was.

I know a few of them. They are squarely facing the challenges that San Antonio spent four days dancing around. And they are building something — not from the anxiety, but from the actual shape of the correct problem.

It’s early. But the direction is different from what I saw coming out of that room.

More soon.

Meanwhile, I’ll ask you this question: What is your industry managing instead of solving?

Human nature is about believing we know.

Most people don’t.

What we actually crave is certainty.

Clarity that could lead to truth tends to arrive as discomfort.

Why? Because it forces a choice.

(Allow me to interject into my own train of thought... I love the
old English literary references to dyspepsia. Dickens’ Scrooge,
Podsnap and Gradgrind each faced uncomfortable moral
choices culminating in dyspepsia. Belch. ‘scuse me.)

The good news is that attracting clarity is mostly achieved by removing what doesn’t belong. Embarrassingly, we almost always—already—in every way—know what doesn’t belong. But humans cling to the familiar, and to putting things off for... (Oy)

So here it is.

Clarity is the interface of neuroscience, quantum physics, and theology. But it’s easier to discover than it might sound.

Clear thinking can be encapsulated into five simple commandments.

Clarity is not a gift.

It is a discipline.

Cut the noise.

Ask the right question.

Face the truth.

Listen.

Act.

Most of life becomes surprisingly simple after that.

You are welcome here for more. I appear to have a “jones” for clarity. But I mostly share the best of getting to the right problem to be solved. You might like that. ‘Sfun. ‘sprofitable. Engineers welcome here! Poets? Not so much.

I thought this Men’s Health piece was worth sharing because it frames anxiety in men in practical, recognizable terms — not just as worry, but as behaviors people often normalize. It’s not about weak men. It’s about outside-of-community demands in the desire to be a visible success.

That connects directly to a larger theme I keep coming back to: solving the right problem starts with seeing the right problem.

If we misread avoidance, irritability, overdrinking, or social withdrawal as character flaws, we may respond with judgment when the real issue is signal, stress, fear, depletion, or untreated anxiety.

Read the full article here:

Read Men' Health piece about Stress

In 2018, when 12 boys and their soccer coach were trapped inside Thailand’s Tham Luang cave system, the visible problem was obvious: they were underground, behind floodwater, with monsoon rain closing off the way out.

That was terrifying enough. And visible problems created visible reactions.

Pump the water out.
Drill from above.
Wait for the rain to stop.
Teach the boys to dive.
Send in more equipment.
Try something.
Try anything.

And as hours passed, these each proved wrong. Lost time is the cost of confusion.

False opportunity is motion without clarity. It is noise. And that is all.

The world saw water, darkness, distance, rain, oxygen levels, and time running out. All of those mattered. But none of them was the decisive problem.

The decisive problem was panic.

A terrified child, underwater, in total darkness, inside a narrow cave restriction, could drown himself and possibly kill the rescuer trying to save him. Not just one child. Twelve.

That was the signal.

The rescue did not succeed because the cave became safe. It did not. It succeeded because the rescuers stopped trying to solve the cave and started solving the actual failure point.

They could not remove the danger.

They could reduce the chance of panic.

That changed everything.

The Real Problem Creates Leverage

The Thai cave rescue is often told as a story of courage.

It is that.

But more useful, it is a story of a few individuals reaching a separate diagnosis.

The apparent problem was:
The boys are trapped behind floodwater.

The real problem was:
The boys would not consciously self-rescue through a lethal underwater route without panicking.

Once that distinction became clear, the rescue strategy changed.

This is the really-real lesson leaders miss.

Most teams are not failing because they lack effort. They are failing because their effort is aimed at the wrong problem.

They hold more meetings.
They add more process.
They buy more software.
They hire more people.
They push harder.
They move faster.

But speed doesn’t help if the target is wrong.

The visible problem creates activity.

The real problem creates leverage.

The difference is everything.

When Authority and Expertise Align

This rescue also worked because power and authority came together with rare expertise.

Formal authority mattered. Thai officials, military leaders, local coordinators, and Governor Narongsak Osatanakorn helped organize the operation, control access, mobilize resources, and make the rescue legitimate.

But authority alone could not solve the cave.

The decisive insight came from people who understood the environment: elite cave divers and medical experts who knew what panic, breathing, sedation, and airway control meant inside a flooded cave.

That combination mattered. The collaboration was fresh. The cooperation was swift.

Power could mobilize.
Authority could approve.
Expertise could diagnose.
Leadership aligned all three.

In business, the inability to achieve a fresh approach is where the Friction Tax appears.

The people with authority do not see the technical constraint.
The people with expertise do not have decision rights.
The people closest to the real problem are not in the room.
The people under pressure reach for the most visible fix.

That is how organizations waste time, money, and credibility solving the wrong thing. And destroy reputation, goodwill and value.

The Thai cave rescue shows a better pattern:

Find the real failure points. Exploit new ideas.
Put expertise close to authority.
Act before confusion becomes the default strategy.

The Business Lesson

In business, the cave is rarely literal.

But the pattern is everywhere.

Sales are down, so everyone blames marketing.
The product is slow to grow, so everyone demands more features.
The team misses deadlines, so everyone talks about discipline.
Customers complain, so everyone rewrites the messaging.
A project stalls, so leadership asks for more updates.

Sometimes those are the right moves.

Often they are reactions to the visible problem.

The better question is:

What problem, if solved, changes everything else?

That is the Thai cave lesson.

Not every problem deserves equal attention.

Some problems create activity.
Some problems create leverage.

Leaders earn their keep by knowing the difference.

And this is where the practical value begins.

Because every leader eventually faces some version of that cave: too much noise, too little time, incomplete information, competing advice, and no perfect option.

The question is not whether confusion will appear.

It will.

The question is whether you can find the one problem that actually changes the outcome.

That is where the difference shows up.

Next I’ll show you how that works.

Read more

Everyone is trying to solve the AI problem.

Companies want better models.
Investors want better exposure.
Governments want better competitiveness.
Managers want better productivity.

But what if that is not the real problem?

What if the real problem is that AI is being discussed like software while being built like heavy industry?

That distinction matters because software scales differently than infrastructure. Software can be distributed almost instantly. Infrastructure has to be financed, permitted, powered, cooled, insured, and operated. It has physical constraints. It has debt cycles. It has bottlenecks. It has exposure to energy markets and geopolitical instability.

Three stories today point in the same direction.

Oracle shocked investors with an AI infrastructure spending plan that could reach $95 billion in fiscal 2027, well above expectations. KKR launched a $10 billion AI infrastructure platform with Nvidia, Vistra, and the Kuwait Investment Authority. At the same time, renewed Middle East conflict pushed oil near $95 and reminded markets that energy security still governs inflation, confidence, and operating costs.

These are not separate stories.

They are one story.

The Story

Oracle is spending aggressively to expand AI and cloud infrastructure. Investors reacted negatively, not because demand disappeared, but because the cost of meeting that demand is becoming enormous. The company reported strong revenue and large future obligations, yet the market focused on debt, capex, and free cash flow.

KKR’s Helix Digital Infrastructure points to the same shift. AI capacity now requires private capital, chip design, power provision, sovereign investment, and experienced operating leadership. Nvidia contributes design expertise. Vistra brings power. KKR brings capital structure.

Meanwhile, Middle East instability and oil-price pressure show that the global economy still lives inside the physical energy system. Digital ambition does not float above oil, gas, shipping lanes, and central-bank reaction functions.

The Noise

The noise is easy to identify.

Oracle is reckless.
AI is a bubble.
Private equity is piling in.
Oil is up because of war.
Markets are nervous.

All of that may be true. But none of it is sufficient.

The market loves simple labels because simple labels are easy to trade, argue, and amplify. “AI bubble” is emotionally satisfying. “Tech selloff” is easy. “Geopolitical risk” is familiar.

But labels are not diagnosis.

The Signal

The signal is that AI is moving from a software-margin story to an infrastructure-capacity story.

That means the important questions change.

Not just: Who has the best model?
But: Who has access to power?

Not just: Who has the fastest growth?
But: Who can finance capacity without destroying returns?

Not just: Who owns the customer?
But: Who controls the compute layer underneath the customer?

AI demand may be real, but real demand does not guarantee attractive economics. Railroads had real demand. Telecom had real demand. Energy infrastructure has real demand. The hard question is always who earns the return after the capital cycle turns.

The Real Problem

The real problem is not whether AI matters.

The real problem is whether the market is applying the wrong mental model.

If AI is treated as pure software, investors will emphasize growth, adoption, and user engagement. If AI is infrastructure, they must also evaluate capital intensity, debt structure, utilization, power contracts, grid constraints, and regulatory risk.

That is a different discipline.

It is less glamorous.
It is less viral.
It is more useful. Practical. Realistic.

What If Everyone Is Solving The Wrong Problem?

What if the AI race is not primarily a race for intelligence?

What if it is a race for capacity?

The company with the best model may not win if it cannot access enough compute. The company with demand may not win if it finances expansion badly. The company with capital may not win if it cannot secure power. The region with ambition may not win if its grid cannot support the load.

That reframes the entire conversation.

Three Alternative Paths

First, define the problem as a power-access problem. In that case, utilities, grid operators, and energy developers become central to AI strategy.

Second, define it as a capital-structure problem. Then investors should examine debt, free cash flow, customer commitments, and asset utilization before celebrating growth.

Third, define it as a coordination problem. AI infrastructure requires alignment among technology companies, power providers, financiers, regulators, and customers. The bottleneck may not be brilliance. It may be orchestration.

What Changes If This Is True?

Leaders stop treating AI adoption as a software rollout.

Investors stop buying every AI-adjacent company as though all exposure is equal.

Policymakers stop assuming innovation alone will carry national competitiveness.

Managers stop promising AI transformation without understanding the infrastructure behind it.

The right question becomes: what system must exist underneath the promise?

Better Decisions

Separate AI software exposure from AI infrastructure exposure.
Treat power availability as a strategic constraint.
Analyze AI companies through cash flow, capex, utilization, and financing quality.

Those decisions help avoid the obvious mistake: treating every AI company as if it has the same economics.

But the more valuable question is downstream.

If AI is becoming infrastructure, then the consequences will not stop at valuation. They will reshape capital allocation, energy markets, regional competitiveness, and the kinds of companies investors reward.

That is where the second-order effects begin. And where the money starts moving. Let’s look at how.

Read more