Today I had lunch with a friend whom I haven’t seen in several years. We met in a part of the city I rarely visit so I wasn’t familiar with the restaurants. Our phone conversation setting up our lunch plans went something like this:
friend: “What are you in the mood for…”
me: “I don’t know, what is there…”
friend: lists several nearby restaurants – none of which I know
me: “umm…I don’t care, I’m fine with anything…”
Upon arriving we settled on one of the restaurants near the intersection where we met.
Upon going into the restaurant, our conversation went something like this:
friend to waiter: “are there any lunch specials…”
waiter points to lunch specials on the menu and then leaves us to make our decision.
friend to me: “want to share a pizza…”
friend: “want kind of pizza do you like…”
me: “any kind….lunch special sounds good”
friend: “no, too many meats. What about margarita or perhaps quattro formaggi…”
me: “sure, that sounds fine…”
waiter returns and takes our order.
In reality, you can see that these two decisions – where to eat and what to eat – didn’t take a tremendous amount of coordination. But it did take some coordination. More, it monopolized most of the initial conversation with someone I haven’t seen in two years. As we begin to (first) digitize broad swaths of information and (second) intermingle and intertwine these diverse streams of data, I can envision a variety of ways these series of decisions are handled very differently in the future.
Imagine this (not implausible) scenario:
First, my phone has stored locally everywhere I’ve eaten in the last two years or more likely my phone has access to a service or set of services which knows this information. Three or four years ago information like this wasn’t digitally available. Today, this information is stored across a variety of services. Foursquare knows where I’ve eaten and where I eat frequently. Facebook might know where I’ve eaten. Yelp might know what I like or Pinterest or Facebook photos or the tips I’ve left on Foursquare or the tweets I’ve sent through Twitter. I’ve left digital breadcrumbs across a number of Internet accessible services and the rate and magnitude of these digital bits of information are only increasing.
The second element of this future scenario requires sharing these diverse (and dispersed) streams of digital data. I believe cross pollination is the proverbial next big thing when it comes to digital data. We’ve started to capture (personal) information digitally, but we aren’t yet making predictions based what can be derived from looking at these streams collectively as opposed to individually. We are quickly entering an Age of Algorithms.
Here’s an example of how this might work. Not only does my phone have access to my personal digital data, but it also sees the digital data of my friend. This also isn’t that far fetched. Since we are connected on Facebook for example, this single service already knows a lot about each of us. Facebook could have recommended one of the nearby restaurants by mapping our profiles against each other and examining the intersection of our profiles while utilizing the GPS coordinates of my phone to identify high probability matches.
The second question – what to eat – could be solved in a similar way. Imagine our phones have embedded NFC or some comparable technology. As we enter the restaurant I opt to “share” my information with the restaurant. The digital streams of data I share will include obvious things like food allergies, but it could also include explicit or implicit eating profiles. I might share what restaurants I’ve ranked highly in Yelp or in other similar services. I might share my credit card information which knows not only every restaurant I’ve eaten at, but also what I ordered. In the future I might share my grocery store scanner receipts which contain a record of everything I’ve bought at the grocery store. Rather than asking the waiter what he or she would recommend and getting back their personal (and therefore biased) input, I could get a series of computer-generated, probabilistic recommendations based on the analysis of literally thousands of data points. These recommendations could provide options for what I like individually or what multiple people might like to share. In this way, we are increasingly turning over decisions to machines. Or at least, we are allowing machines to influence our decisions. In this we see how the digital world is going to increasingly influence what happens in the physical world.