What travel AI still hasn't learned from Hipmunk

Skift reminds us of an uncomfortable lesson about travel AI: Hipmunk, the metasearch engine that sixteen years ago revolutionized how search results were displayed, built its pages around the psychological reality of the traveler, not around what was technically easy to show. Today, with all the machine learning and generative AI power available, most travel tools still haven't replicated it.

The difference was stark. While other search engines showed flat lists of identical options, Hipmunk presented intuitive visualizations, semantic filters, and flows that recognized that traveling isn't just comparing prices. A traveler who slept poorly on planes could instantly find morning flights. Someone who hated long layovers had a clear visual map. It wasn't pure AI, it was design obsessed with the real problem.

Today we see travel tech startups building chatbots that sound fluent but still force users to navigate generic or incomparable options. AI generates smooth text, but doesn't solve the fundamental problem: understanding what the traveler wants to decide and showing them only that. It's the trap of confusing technological capability with clarity of value. Some hoteliers see these tools as the future. I see many of them as Hipmunk without the Hipmunk. The challenge isn't more sophisticated AI. It's teams that understand travel as well as they understand code.