AI For Every Day Decision Making — Drive or Get a Lift to the Airport?

Image for post
Image for post

Should I drive to the airport and park or arrange an Uber / taxi?

This question was riddling me for a few days. Not so much that I couldn’t live with either decision but that I was thinking about whether AI could actually help me decide.

First, what was the problem? An upcoming two week trip with the family meant traveling with kids, spouse, and enough luggage for four in a car to the airport. The airport is about a 25 minute drive from our house on a toll highway. The cost of hiring a large enough vehicle to lug us all with gear to the airport is a little bit less than the cost of budget parking at the airport.

I was going through the pros and cons and was wondering if the time would come when an AI assistant could help me decide. If I were to assume that an AI would assess the situation similar to the way I would, here’s what the factors looks like.

Cons against taking a taxi/Uber and pros for driving

  • The chaos of travel day means not being able to reliably schedule a ride within a 15 window anytime in advance. In the past, we’ve had taxis cancel on us for having them wait too long… rightly so.
  • Driving means being able to load up a car as we pack and not rushing to pack a taxi as it arrives. Less chance of forgetting items. Oh… and what if we forget an item on our way to the airport? Driving means some fallback plans.
  • We like to make the time on the car ride to the airport a special time together celebrating being off on our trip. We’re a bit more repressed when a stranger is sitting with us. It’s not as much fun.
  • Speaking of which… I’m waiting for the days of autonomous driving. Most people are good but again, we like the private celebration of just the family being in the car.
  • Taking an Uber / taxi means we also need to bring along the booster seat for our older daughter. It also means checking that in and taking it with us on an overseas trip. We have a booster for our car rental at the destination for a nominal fee but in this scenario, it means having to lug around an extra item to check in and pick up both there and on the way back.
  • We have to setup the booster seat for the taxi ride there and then for the taxi back.
  • We’d need to arrange a ride for the way back and there’s uncertainty what the situation will be (and what condition two young children will be in when we arrive after a long flight)
  • Driving provides some certainty on when we’ll get there and where we’ll park and all of the steps getting to and from the airport.

Cons against driving / pros for taxi / Uber

  • It costs less, by about 10% vs driving and parking.
  • The drop off is directly at the terminal, saving about 10–15 minutes getting there and allowing us to go directly to check in.
  • Same on return. The time waiting for an Uber / taxi at the airport is still likely less than the commute to the parking garage
  • The discount parking garage requires walking on an incline with two kids, luggage, and a stroller, going up an elevator, coming down an elevator, and taking a monorail (“what’s that word?”)
  • I’ve already determined that the other non-airport parking lots aren’t feasible — not with two kids in tow.
  • It can be arranged in advance and no hassle with Uber. Can also be canceled if need be.
  • Leaving the car in an publicly accessible garage for a few weeks isn’t the best idea.

Factors that could affect a decision

There are a few options that could affect the decision making:

  • A taxi / Uber option with a booster seat. This reduces the annoyance of having to lug around the booster seat and install / uninstall it.
  • Dropping off the family first at the terminal and then parking.
  • Using an off-lot valet service that will park the vehicle.
  • I don’t travel on long trips with the family to the airport long enough to have built up experience of getting there with them (once or twice a year). I travel often on 1–2 day trips but the economics of different travel options to / from the airport and parking there for that length of trip are different.
  • Getting a friend / family member to give a lift
  • The distance of luggage carts in the parking garage
  • The ability to leave the car for one minute to bring items into the terminal if I’m using the drop off and park scenario

Information sources for decision making

Assessing which mode to use means collecting information from a few sources…

  • How my family feels about these options
  • Comparing price
  • Understanding how my schedule on the day of travel is shaping up
  • Understanding how prepared I am for the trip (how much packing have I completed 12 hours before the planned departure time)
  • Comparing time
  • Looking at the stress impact caused by the scenarios (relying on people we haven’t met yet, working around a pre-arranged schedule, having a driver wait for us, amount of distance to walk, spending unnecessarily, lugging luggage, loading a vehicle)
  • Looking at the reviews of alternative services and how they could affect the above

Where AI is helping today

AI is helping in a very subtle way today and already plays a role…

  • Cost estimates from Uber are based on machine learning
  • Estimated times and directions from Google Maps are based on AI
  • How reviews for alternative services are presented from Google or Yelp is algorithmically determined, with the idea that the most impactful review will be displayed first

Without this information, the accuracy of the estimates would be lower and the effort required to attain this information would have been more work (maybe 2–3 minutes of calls to different taxi companies as one example). For reviews, having them presented as a number helps me make a quantified decision as opposed to asking around for opinions from friends who might have interacted with a service and then weighting their response.

What we’d want from AI (or an assistant in general)

Really, what we’d want would be for an AI assistant is to assess what’s important to us, gather the information for us, and present a straightforward conclusions. Even better, it would come up with creative alternatives we might not have thought of.

On top of this, for us to actually follow the recommendation, the AI needs to back up its assertion with reasons. We’d want something like this:

“Leor, you should order an Uber Black SUV to the airport and leave at 12:30 PM. Worst case, if something comes up after you book, you can just cancel. If you’re concerned about your Uber rating in that scenario, don’t be — the max they will ding you is .1 stars. Don’t book in advance… I’ll just set a reminder for you at 12:15 PM to book one from the app and you can make a decision then. Also, just send a note to Uber Help to see how to book with a booster seat as that option might save you having to lug around the booster. You can sit in the back with the family and have fun on your way to the airport. Or better yet, work on your ability to talk to strangers.”

“Leor, just drive and park at the airport in the discount lot. You can load the car at your own pace and you’ll have one less thing to worry about. You’ll also be tired from a lot of travel when you arrive back and you won’t need to worry about a finding a lift when you get back or setting up a car seat with screaming exhausted kids. You can leave at 12:45 PM and have lots of time to get to the airport. You don’t need to worry about dealing with anyone else or whether or not you have to check the booster seat. I’ll set a reminder for you to start loading the car up at 12:15 PM.”

Maybe we’re expecting too much?

From the above scenario, maybe I’m expecting too much from AI assistants. In the end, we need to know our preferences more than anyone else. It could be that an AI assistant could help elucidate these preferences through question asking so that we can better understand what we want and reveal to ourselves what’s driving us (pardon the pun). For this scenario, AI could affect our decision making if it allowed for the creation of alternatives:

  • Autonomous driving
  • Autonomous parking
  • Creation of new businesses that could turn a trip to the airport into an experience

In the end, with all the data available, especially for more or less inconsequential decisions like the one above with binary options, I just need to pick and move on to more important things.

Written by

Independent daily thoughts on all things future, voice technologies and AI. More at http://linkedin.com/in/grebler

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store