Window or aisle? It's one of life's great either/ors, right up there with coffee or tea, surf or turf, Mac or PC. It's also an example of a deceptively straightforward question that reveals the challenges and potential of artificial intelligence in travel.
Hidden Complexity in Travel Planning
You make thousands of decisions in the course of planning a trip, not always consciously.
For instance, you might have a preference for aisle seats, but an even stronger preference to avoid aisle seats close to the lavatory. So when selecting your seat on a flight, you intuitively look for ones that are on the aisle and far from the back of the cabin.
It's a relatively simple thought process, one that you've probably gone through hundreds of times before. Things become more complicated if someone else is booking the flight for you. In this case, you'd have to translate an innate preference into a formal set of instructions, such as "book the aisle seat, unless the aisle seat is too close to the lavatory."
But how close is too close? One row? Two? Pity the executive assistant who has to guess. A decision you can make without a second thought turns out to have hidden dimensions. The time you save by outsourcing your travel booking is negated by the time you have to spend giving detailed instructions to the person making the booking.
It's hard enough for one human to think on behalf of another. For machines, mimicking the subtleties of cognition is impossibly difficult. But that's changing.
The Promise of Artificial Intelligence
The newest virtual assistants use artificial intelligence (A.I.) to learn your preferences with remarkable accuracy. These apps use information you enter into a profile - preferred airline, for instance - but also patterns discerned from your past booking behavior.
Less advanced automated booking systems can make frustrating, seemingly unaccountable decisions, like booking a middle seat when there are aisle seats still available. Legroom is a meaningless concept to a computer, unless it is specifically programmed to "know" otherwise. Artificially intelligent systems can avoid these errors by taking into account the fact that you've never chosen a middle seat before.
Seat choice is an illustrative, though admittedly minor aspect of trip planning. A.I can also help with other decisions that have a far greater impact on your travel experience: what flight to choose, when to book, where to stay, how you should make alternate arrangements in case of a disruption.
A.I. is particularly useful for scheduling. Virtual assistants can scan your calendar to proactively identify events that might require travel, then suggest the most convenient itineraries.
Apps such as Pana and HelloGbye combine machine learning with a fluid user interface built on in-app messaging, voice command, and perhaps somewhat surprisingly, the ability to connect with a real human.
Unlike the executives of old, who might have relied on a secretary or dedicated corporate travel agent to search for and book their flights and hotels, younger business travelers are largely comfortable making their own arrangements. Virtual assistants are designed to optimize travel management for the self-sufficient generation.
The appeal is clearly there for individuals looking to cut down on the time it takes to coordinate their trips. Employers will also be intrigued from a perspective of expense management. If intelligent travel planning apps are configured to recommend the most cost-effective itineraries, they could have a huge impact on how much employees spend.
The Robots Are Coming
Though the software behind these new apps is highly sophisticated, it's less obviously impressive than the kind of walking, talking robots that science fiction has taught us to associate with the words "artificial intelligence."
At one hotel near Hilton corporate headquarters in McLean, Virginia, science fiction has become reality. This March marked the debut for Connie, a robo-concierge powered by IBM's Watson who can, among other things, make restaurant recommendations and give directions to the nearest bathroom.
We've traditionally equated quality service with a human touch. Think of how many people you interact with in the course checking into a luxury hotel: valets, desk clerks, bellhops, all of them quick to remind you that should you need anything, they're only a call away.
But preferences might be changing. More and more hotels are doing away with front desks in favor of digital kiosks or even advance check-in using guests' mobile phone. Self-service check-in is already standard for airports. How far until the same is true for hotels?
Technology and Travel Costs
This all fits into a larger trend of technology changing the cost structure of travel. Uber has turned thousands of cars into de facto taxis, while Airbnb has hugely expanded the supply of short term accommodations. Both services lower costs and upend the expected service model for travelers.
Even something as commonplace as WiFi fundamentally alters how we travel. When you can work as easily from an airport as from an office, cheaper flights with layovers aren't off limits. Though A.I. and robots generate the most interest, simpler technologies could end up having the most impact.
So is travel A.I. a novelty or a game changer? That's another question that's open for debate. But for the record, I'm a window guy.