THE DISPATCH
The Cookie-Cutter Trip AI Quietly Sold You
It looks personalized. It’s not. Here is the lens AI sees travel through, and how to push back
By Paul M. Rand, Founder and Editor
of Nomadic Spirit
You want to plan your next trip. Somewhere new. Somewhere you have wanted to go for years and have not yet figured out how.
You open ChatGPT, Claude, or Gemini and type: two weeks in Uzbekistan.
Ninety seconds later, you have a fourteen-day itinerary. Train times between Tashkent, Samarkand, Bukhara, and Khiva. Visa rules. A photograph of plov so you know what to order. Names of three small hotels on the road to Khiva, where the rooms are clean, and the breakfast is good.
Sounds pretty amazing, right?
Well, unfortunately, the hotel at the top of the list closed last year. The train between Bukhara and Khiva runs three days a week, not seven. The plov in the photograph is from a recipe blog, not from any kitchen in Uzbekistan. And the entire fourteen-day arc is the same arc, the same tool hands the next traveler who types the same prompt.
Beyond the noise
There is a lot of noise about AI right now. Pope Leo XIV recently warned about the technology in his first encyclical, “Magnifica Humanitas.” A Stanford study in Science this spring found that leading AI models are sycophantic, telling users what they want to hear even when the users are wrong.
What you need is something quieter: a working understanding of the lens through which AI sees travel. With that, you can use what AI does well without being at the mercy of what it does poorly.
The lens
The reason your Uzbekistan itinerary came back the way it did is structural. Three things shape what AI returns when you type a travel prompt.
Whose voices got fed in. AI is trained mostly on the parts of the English-language internet that have been most clicked and most read. Hotel marketing, TripAdvisor reviews, and the well-known travel blogs. Voices written in other languages, or in small local papers that no algorithm picked up, are weighted lightly or not at all.
What is not in the data. The bakery the locals walk to has never been written up. Nor has the bus driver’s shortcut, or the festival the village paper covered once.
Who profits from what surfaces. The recommendations that come back are the ones most aggressively optimized to surface: hotels with the largest marketing budgets, restaurants with paid placement on review aggregators, and tours run by companies that know how to play the algorithm.
The cookie-cutter trip
An entire industry is now forming around this. Just as Search Engine Optimization, or SEO, emerged to manipulate where companies appear in Google search results, a parallel industry is taking shape to manipulate where they appear in AI responses. Call it Generative Engine Optimization, or GEO. The recommendations you get from a chatbot today are increasingly shaped by professionals whose job is to make sure their clients are the ones it mentions.
Add it up. Thirty different travelers who type “two weeks in Uzbekistan” into the same tool will get roughly the same answer. The itinerary feels personalized because you typed the prompt yourself. It is not. It is a cookie-cutter trip shaped by the most-clicked-on parts of the internet, and by the people optimizing to surface there. The tool gave you the most-written-about Uzbekistan, not the one a friend would send you to.
What comes next
Knowing this gives you the language to push back. With the right follow-up question, the cookie-cutter trip becomes something closer to yours. The savviest travelers already use the tools this way.
This Dispatch was originally going to cover those tools, too, but it got too long. So we split it. Next week: which AI tools the savviest travelers reach for, with named apps and specific prompts for each.
