Mind-Reading Travel AI
Recently, I was entertained by a skit from comedian Ronnie Chieng, a correspondent for the Daily Show. He humorously proposed that Amazon Prime should transform into “Amazon Today,” eventually culminating in “Amazon Yesterday” – a service that identifies and delivers our purchases before we even realize we need them.
Travel is no exception. I demand solutions to travel problems I didn’t even know existed, before I even realized they’re problems.
Mind-Reading Travel AI
In the context of travel, it dawned on me that what I truly desire is for Travel AI to read my mind. It’s a simple yet ambitious idea – just reading my mind. No one can accuse me of thinking small. 😉 I observe this as a former travel technology developer & CTO, former boutique OTA owner, former travel agent (retail and corporate) and now travel venture capitalist investor.
Autonomous Action
To be more precise, I want my Travel AI to mimic my actions; to consider all relevant options, weigh the pros and cons, identify the optimal solution, and act autonomously. Initially, Travel AI will be a digital extension of me, but eventually, it’ll surpass the capabilities of a hundred versions of me.
Travel AI’s Knowledge
For Travel AI to think, evaluate, and act like me, it needs to automatically identify, reverse engineer, and implement my travel habits, preferences, and rules. My exceptional personal assistant or travel agent would be familiar with these. Travel AI can now empower them to perform better than ever by identifying new latent habits, preferences, and rules (HPRs) that no one realized I had, while adhering to known ones and tracking changes along the way.
Sample Habits, Preferences and Rules
Every traveler, including myself, has a unique set of habits, rules, and preferences that guide their travel decisions, often without them even realizing it. For instance,
- I tend to search for a select few airline options influenced by my loyalty and preferences, rather than the dozens universally available between city pairs.
- My schedule is carefully planned to ensure I leave on a Sunday late night and return on Friday late night, maximizing my time with family.
- During peak traffic hours, I prefer to leave earlier and utilize the lounge, airport conference room, or even switch to a train instead of an Uber car.
- When heading to New York, I aim to fly into JFK for meetings or conferences, and if possible, depart from EWR to visit friends and family.
- If meetings conclude earlier than expected, I’ll check out early from NYC, take a train, car service, or helicopter out of the city, and opt to stay with friends or at an EWR airport hotel.
I share the above as a sampling of nuanced habits, preferences and rules Travel AI can address.
» Comment below on what quirky “habit, preference or rule” do you have? «
Proactive Problem-Solving
Some problems are completely unforeseen, but many more are predictable.
Like any exceptional travel agent or personal assistant, Travel AI will anticipate potential problems and act before they become actual problems.
Proactivity is invaluable. It’s one thing to solve a problem after it occurs, it’s another to prevent it from happening in the first place. Travel AI will assist my travel agent or executive assistant in searching, evaluating, and executing tasks, effectively acting as hundreds of helpers working simultaneously.
One Answer
One aspect I appreciate about Gemini (Google’s Search AI engine), which has spoiled me compared to the dozens of links provided by traditional Google searches, is receiving a single answer (and supporting links) when I ask a question. I save *so* much time — ask, answered, move on with life — and I’m finding the % of times I ‘fall back’ to traditional Google.com searches is rapidly decreasing as AI becomes smarter and faster.
I expect Travel AI searches to yield one optimal solution as well. It’ll sift through a myriad of data points and come up with one “best case” answer versus dozens of options to decide amongst.
Isn’t that what a very good executive assistant or travel agent already does?
Ramifications of One Answer
I propose that the desire of travelers for a singular answer (or perhaps the top three answers) presents both a potential advantage and a risk in contrast to the current state of travel search. I suggest that travelers, particularly those from Generation Alpha and beyond, who have been raised with “native AI search” results, will expect 1-3 search results from their travel providers, just as they would from their search engines. When this shift occurs, what becomes of the 90% of airline or hotel rates that go unseen? Could it be that travel search results will evolve into a format of three sponsored results paired with three organic results? How will “time on site” and “engagement” be redefined? Thankfully we have time on our hands.
» Share your thoughts below if and when we’ll see travelers wanting 1 – 3 options versus the dozens of results today? «
Endless Everchanging Travel AI Data Points
Travel AI will scour a Travel LLM trained just on my entire travel history, it’ll ping the GDS for available options, it’ll ping NDC data, it’ll ask LLM’s trained on hundreds of thousands of itineraries of similar cohorts to me, it’ll access 3rd party API’s, it’ll apply proprietary OTA/TMC specific algorithms and recommendations, it’ll filter for applicable corporate policies and rules, consider costs and budget, it will selectively ask and verify it’s observations, it’ll apply all my personal “habits preferences and rules” (amongst many other data points and sets of training data).
Ending with one “best case” answer versus dozens of options to decide amongst.
Excited Anticipation
The possibilities that AI has to offer fill me with childlike excitement. The journey won’t be easy, and it will require a few steps forward and occasional steps back, but I posit it will yield a positive ROI, a greatly improved traveler experience — you know, the person that pays all the bills in travel — and be well worth the wait.