Android Authority writer says Claude organized, not planned, a three-week Japan trip

The account said the writer does not favor large language models as decision-making tools but finds them useful for logistics. Asking Claude, Gemini, or ChatGPT to plan a multi-day trip to Tokyo tends to produce a generic list of attractions, the writer said, including recommendations that can be outdated or that ignore travel time across the city. The writer attributed that outcome to models drawing on broad internet knowledge rather than knowing a traveler's preferences, walking limits, or how an individual stop fits an itinerary.

To avoid that, the writer curated the trip first. The account said the writer collected ideas from Instagram, YouTube, blogs, restaurants, shops, photography locations, and prior research, then asked Claude to turn the material into a usable plan. A dedicated Claude Project held a master document with the trip information. At the top, the writer kept a decision log for major choices such as dropping a side trip, changing hotels, or setting the number of days in a city. Keeping that document updated allowed new conversations within the project to start without re-explaining the entire trip, according to the account.

The writer also used Chrome MCP to give Claude access to live web pages. When the writer found a blog post about a photography location, they could provide the URL and have Claude extract useful details, such as a train schedule or admission fee, and assess whether the location fit the plan. The account said more filtered information made Claude more useful.

Personal reference documents played a large role. The writer created a guide for a severe seafood allergy, covering safe Japanese dishes, ingredients to avoid, and phrases for restaurants. Other lists included camera lenses, vintage watch stores, and Japanese streetwear brands. In one case, the writer gave Claude a long list of local designer brands. Instead of returning another long list, Claude grouped the stores by geographic area, which the writer then turned into a Google My Maps file. The writer said the task would have taken hours of opening addresses and moving pins but took minutes.

The account's main conclusion was that human curation cannot be removed from travel planning. Claude and similar AI agents operate better with better filtering, the writer said, and human taste remains the best filter. The writer also said the trip changed how the itinerary was structured across three weeks in Tokyo, Kyoto, and several small villages, moving away from planning every minute of every day.

Editor's Summary

A writer for Android Authority described using Claude to organize a three-week Japan trip, not to choose destinations, and said the method saved time and produced a smoother vacation. The account emphasized that travelers should curate their own preferences and itineraries first, then use AI to extract details, cluster locations, and manage logistics. It illustrates a practical but limited role for large language models in travel planning, where human judgment remains central.