Winemakers On Fire

Greetings from Cape Town, South Africa!

I've been thinking about what AI cannot do.

Not what it does badly. Not what it hasn't learned yet. What it structurally, fundamentally, and by its very nature cannot do, regardless of how fast it improves or how much compute it consumes.

Two things keep surfacing.

It cannot ask a question that runs against everything it was trained on. It can only synthesise what already exists. The genuinely uncomfortable question, the one that questions the consensus rather than confirming it, requires a human being willing to be wrong in public.

And it cannot build genuine trust. The slow, specific, accumulated trust that comes from showing up honestly before anyone is watching, investing in relationships before they return anything measurable, and being changed by the people you're in a relationship with.

These are not small things. In a world where AI is replicating almost everything else, they are everything.

The six posts in this issue explore both from different angles. They come from my forthcoming book, Own Position Zero: The Human Advantage No Algorithm Can Fake, which I've been quietly writing while these posts were finding their shape.

Read them slowly. The ideas that follow took twenty-five years to learn.

The questions you've stopped asking.

Still Asking?

AI knows everything that has already been thought.

Every pattern. Every best practice. Every consensus opinion in every industry ever documented.

That is also its ceiling.

The question that runs against everything the consensus believes, the one that makes your peers uncomfortable and your industry defensive. That question requires a human being.

It always has.

The Expensive Silence

Every industry has a question nobody is asking.

Not because the question is unanswerable. Because answering it honestly would implicate too many people who have built their careers around something incomplete.

The most valuable thinking in any field almost always happens at the edges.

When someone asks the question, the consensus decided it wasn't worth asking.

Ask It Anyway

The assumption you've stopped questioning is doing more damage than the ones you never held.

Not because it's wrong. Because you've stopped testing whether it's still right.

Markets move. Consumers shift. The world changes.

Assumptions don't update themselves.

The question you stopped asking three years ago deserves a second look.

Today.

The table AI cannot build.

What Algorithms Miss

AI can map your network.

It can identify connections, suggest collaborations, optimise your engagement, and tell you who you should know.

What it cannot do is sit across a table from someone, say the uncomfortable thing, and have the relationship survive it.

That specific, costly, irreplaceable act is where trust actually lives.

No algorithm reaches it.

Slow Currency

Trust doesn't compound like interest.

It compounds like soil. Slowly. Invisibly. Through thousands of small acts of showing up before anything is returned.

The wine businesses built to last didn't acquire trust through a campaign.

They accumulated it through decades of decisions nobody was watching.

That's not a strategy.

That's character.

The Only Table Worth Building

Five Swartland winemakers chose collaboration over competition.

They shared a stage. Shared audiences. Shared risk. Built a category none of them could have built alone.

No algorithm recommended it. No platform optimised for it.

Five people made a human decision to build something larger than themselves.

Twenty years later, the world studies what they built.

Wine Futurist Mike Carter, based in Cape Town, South Africa, believes the path to future success isn't found in perfect predictions but in perfect positioning. Connect on LinkedIn for more contrarian thinking about wine's future.

If this issue resonates, please share it with one person who needs to hear it.