In the case of OTAs (online travel agents: Jalan net, Rakuten Travel, Ikkyu, etc. in Japan), the top page had users enter "How much within the budget on January 0, how many people, and where do you want to stay?" to display a list of search results.
As a result, it is difficult for serendipity to occur between users and travel destinations, and the problem was that the polarization was accelerated, with winning tourist destinations winning more and losers losing more.
However, by utilizing ChatGPT, it will be possible to capture ambiguous needs that conventional OTAs have not been able to meet and match travelers with travel destinations.
There is a possibility that people will visit "unknown but actually good areas" where people have not come until now, and I expect that the current polarization trend between winning and losing tourist destinations will be corrected, contributing to the leveling of demand for destinations and the prevention of overtourism.

Expedia, an OTA from the United States, has already announced an itinerary function using ChatGPT.
On the other hand, there is a possibility that the ChatBot service, which existed as a SaaS service for hotels, will become obsolete. The recently listed Tripia is one of them.
The need for ChatBot services continues to be high due to the shortage of staff working at accommodation facilities, but there are currently eight of the above services alone, and they are crowded.
Behind the scenes, companies that seek API linkage of ChatGPT may survive, and companies that have developed AI on their own may be subordinated.