Artificial Travel
I saw an interesting article in the New York Times this morning. I suspect that its the sort of article we are going to read more and more often as the applications of artificial intelligence start to shine through and change our lives at the day-to-day level. This particular piece was about a reporter who was planning a vacation to Scandinavia and used some combination of three AI bots to plan her trip for her. I presume that meant that she put in certain parameters as to timeframe, cost and preferences and that the algorithms spit out a bespoke itinerary unlike any that she might have organized herself or any travel agent might have thought to organize. Her assessment of the AI itinerary was that it was materially unique and interesting because the AI bot selected a lot of activities and venues that the reporter was convinced she would never have heard of. When you think about it that is not so hard to imagine. If any of us were suddenly to plan a trip to an unknown locale,the first place we would likely head is Trip Advisor. What we would get would neither be specially designed for the traveller, nor that unique, given that sites like Travel Advisor use consensus appeal and polling to determine its favorite suggestions. What is needed and what AI can presumably do is create a Pareto optimal linear program, something invented in 1939 and taught in business schools for years. You develop a set of constraints and an algorithm figures out the best fit to maximize satisfaction of those conditions. What AI might be able to deliver like never before is the ability to consider much greater amounts of data to fit to those conditions. That would allow new and different attractions to get into the mix and the LP would allow the user to indicate as one of the conditions, the willingness or desire to try new and different things.
I am testing my own version of this in a different way. For the last 40+ years I have planned and organized our family vacations, pretty much on my own. Its true that for the last 19 years I have had Kim to help me with what we call the cruise director portion of the trip, which is to say that she would take the basics of my plan and overlay dining options and what I would call cultural activities, usually after dinner. I have been very proud of our family trips and i feel its safe to say that everyone has always enjoyed them, but I’m not sure anyone would ever grade them higher than a 4 out of 10 on the creativity index. When you have been doing this for as long as I have, its safe to say that you tend to find a formula that works and go with it. Therein lies the secret to introducing innovation into the planning, presumably like AI might. Get someone else involved and give them decider control.
My oldest son, Roger, asked to do this trip we are on to Virginia Beach. I know how much he likes beach communities and specifically the whole boardwalk scene. He has moved his and his wife Valene’s life to the Delaware shore, near Rehoboth Beach and Dewey Beach. He loves it and that’s great. Over the years, at his request, we have done various day trips to places like Atlantic City, Asbury Park, Ocean City and Cape May. But a day trip is not a family vacation. This time, centered around his birthday (which comes tomorrow) we have planned a five-day family vacation for ten of us (Kim and me, three kids and spouses, and two grandkids). And as a nod to Roger, who turns 42 this year, I have given him total control over the entire itinerary. What has resulted is a plan that is very much laid out, indeed published in an amazingly professional agenda complete with thematically asppropriate and fun graphics that colorfully and creatively lay out all the details including times and addresses. He went so far as to include”Fun Facts” for every day to add a customized aspect to it. Actually very impressive and not so surprising given his career in the family attractions business.
While we started our travel yesterday, the real schedule began today. Eight of us drove and met up at Chincoteague Island for a two hour boat tour of the wild ponies of Chincoteague. It seems that back in the days of the Spanish galleons, it is suspected that a galleon crashed into the jagged shore along the Delmarva Peninsula of Virginia and the ponies that they carried in the hold managed to survive to swim ashore, where they survived and thrived, roaming the island in the wild. I’m sure locals to the area for three hundred years might have occasionally dipped in to catch a few of these ponies and tame them for their own use, but mostly, they were left to themselves in this wild barrier reef island along the stormy Mid-Atlantic coast. Then, in 1926, the small town on the island needed to create a fire department and without other visible means, decided to use the ponies by capturing some and holding an auction in late summer once they knew how the year’s foals had fared. To get the ponies to the auction, the locals had to cowboy them across the Intercoastal waterway and that turned out to be a sight that created a monetizable event. This began the annual pony swim of Chincoteague Island. We were not there with 50,000 other visitors for the swim, but we did see where it is held and we did see some of the herd in distant marshes. This was not an attraction that I would have ever thought to attend on our way from Delaware to Virginia Beach, but Roger did.
Here in VIrginia Beach we will breakfast at the Pocahontas Pancake House (in honor of nearby Jamestown, the first established European colony on the shores of America as we know it). We will visit the iconic Neptune Statue and walk the boardwalk. Roger is very much into pop culture, much more than I am. He has found one local attraction after another to fill our day, his birthday, in a manner that suits his style. We get the pleasure of tagging along and thereby getting a glimpse at his life’s pleasures. He has noted on our daily schedule that Virginia Beach is deemed to be one of the ten best boardwalks in the United States as listed by National Geographic and Travel & Leisure. Tomorrow is all about nearby Norfolk and the attractions offered by this Navy town like touring the moored USS Wisconsin battleship and other attractions. We will also see Doumars, where the first ice cream making machine was invented. The point is simply this, we will see things and do things curated by Roger. Roger has done his curation not just with his own interests in mind, but with the interests of our full ten-person entourage, including an age range from 8 to 70 and the various levels of energy the group has at different times of the day, both for walking around town and going to restaurants nearby or far away. That is a complex equation that night make for a very complex linear program to its fullest extent.
The limitations of AI are that it can only use data to make its determinations. It can detect hard-to-see patterns and provide ideas accordingly, but what it cannot do is capture the nuances of what someone like Roger can know from a lifetime of experience with all ten people on this trip. Building a linear program that is effective requires the users of the end-product to give voice to their gut feelings, which they may or may not be able to do. I believe Roger knows more about us than we may know about ourselves. And then there is the unique blending of these interests that needs to happen. That may be the ultimate test. Can AI do a better job of capturing human sentiment than a human can? I think we all understand that the human brain is incredibly robust and facile, especially in detecting subtleties and nuancing personalities. For my money, I will vote for Roger over AI to plan our travel.