Four days ago, 🌏 Peter Syme 🌍 published a very elegant, almost poetic piece about the “fog of war” in the travel industry, now facing the invasion of AI. I enjoyed reading it. But there is one important point on which I disagree with Peter. Or, perhaps, not completely disagree.

The fog probably does exist. We don’t know what Google will look like in five years, which platforms will control the customer, how people will search for travel, what AI agents will be capable of doing on our behalf, or even whether some of the companies we consider unassailable today will still matter.

But I am beginning to wonder whether we actually need to see through that fog.

Because if we stop trying to predict the unpredictable and instead look without fear at the things that really matter to a tour operator, the landscape becomes remarkably clear.

I think every tour operator should look at their business today and ask themselves a simple but rather uncomfortable question: “What is there in my business that AI cannot do?”

The answers will hurt. But this is healing pain. The kind of pain that clears away not only the fog of war, but also the much thicker fog of our own illusions and unfounded hopes.

Take your business apart, brick by brick. Hold each brick in your hand, look at it from every side and give yourself an honest answer: can this part of my work be done entirely by AI? Yes or no.

Do you book accommodations for travelers? Ciao. People have been perfectly capable of booking hotels without a tour operator for many years, and AI will make this process so personalized that trying to compete with it as a human being will become increasingly pointless.

Do you arrange transportation? Flights, trains, cars with or without drivers, motorcycles, cabins on ships, tuk-tuks or bicycles? It is almost boring to explain why AI can do all this easily and quickly. And if you are brave enough to look a little further ahead and imagine the approaching invasion of driverless cars, buses, taxis and all sorts of other vehicles that will need passengers but no drivers, it may hurt even more.

Insurance? Largely automated years ago, even without AI.

Perhaps your value is in putting all these things together? Finding the hotel, choosing the transportation, understanding the customer’s wishes and turning everything into one logical itinerary? This fortress does not look particularly safe either. AI can already produce more versions of an itinerary than any traveler has the patience to examine.

Guides are a more interesting case. I wrote about them before, and I still believe that a really good guide will remain extremely difficult to replace for a long time. But not every traveler wants a guide. Some don’t want to pay for one. Others simply aren’t interested enough in historical or cultural content. And many experiences, unless they require professional supervision, can happen perfectly well without a guide walking beside you.

It is not particularly pleasant to look at your own business this way. But once you do, the part of the landscape that actually matters starts to look much less foggy.

AI bombardment can destroy or devalue many layers of your travel business. But the layer at the very bottom is much harder to hit: the things that depend on local access, personal relationships and physical experiences that have never been turned into standardized digital inventory.

And there is an important distinction here.

Suppose you organize a visit to Castello di Brolio. It is a wonderful place, but from the point of view of a tour operator, it is becoming a dangerous product to build a margin around. The castle has its own website. Opening hours are there. Visits and tastings are described there. Booking information is there. AI knows about Brolio and can increasingly do the organizational work that once justified the intermediary.

So perhaps you should find another winery.

And I don’t mean another famous winery whose website just happens to be less sophisticated.

I mean a tiny family winery run by people you have known for years. A place that may barely exist online. You taste their wine, and the owner’s mother brings out homemade local products that you cannot buy in stores, including some strange onion or tomato jam that turns out to be ridiculously good. There is no “experience” to select from a drop-down menu and perhaps no booking system at all. The owners welcome your customers because you have known each other for years, perhaps even become friends. And that relationship has now become one of the most valuable assets in your business.

Maybe AI knows that this winery exists. Maybe. I am not actually sure.

And even if AI knows about the place, knowing about something and making it happen are two very different things.

The same applies elsewhere. A private cheese producer who normally doesn’t receive tourists. A table at a local restaurant that may not appear on any reservation platform for years. An underground crypt with unique frescoes whose key is kept by the priest. A working salt mine. An old forge where somebody will actually put a hammer in your hand and let you strike glowing iron.

AI may know about all of them. But somebody still has to open the door.

And this, I think, is where the real value of a local tour operator is moving. Not toward knowing more than AI. That battle is probably already lost. The value is in being able to make things happen in the physical world that cannot simply be selected, combined and purchased from existing digital inventory.

Perhaps, then, there is an even better question to ask ourselves: “What does AI still need someone in the real world to make happen?”

For me, this distinction is particularly important because this is exactly the problem we have been trying to solve with DGtalguide.

The relationship with the winemaker belongs to the tour operator. We don’t want to replace it. We want to remove the phone calls, spreadsheets, confirmations, schedules, payments and all the other manual work required to turn that relationship into a product that can be sold again and again.

I don’t think our future is in competing with AI over who can choose a hotel better, write a more interesting itinerary or tell a tourist more facts about a church. AI will almost certainly win those competitions.

The interesting problem is what happens afterward.

AI can create the journey. AI can sell the journey. But eventually the traveler arrives in Tuscany, Naples, Iceland or Morocco, gets out of the car and stands in front of a real door.

Somebody has to make sure that door opens.

That is where I see the role of DGtalguide: organizing everything around those unique local relationships so they can become reliable, scalable travel products.

So perhaps Peter is right about the fog. There are plenty of things ahead that none of us can see.

I am simply no longer sure that we need to see them.

When I look at the things that really matter for the future of a local tour operator, the road seems surprisingly clear.