Skip main navigation

Stupid smart devices

In this video Viktor Dörfler argues that smart technology and smart devices can often be stupid.
I find it really frustrating that SMART technology, SMART devices, are so incredibly stupid so often. For example, I bought a Harry Potter book, with softcover, on Amazon. I kept receiving recommendations to buy the same book with hardcover, with softcover, as well as hardcover with a different picture on the top. So they claim that they are recommending similar readings to you. Are these books similar? No. They are identical with the small difference in packaging. I obviously don’t want them. My friend, Zoltan– first time in his 62 years of life– went to Lima. For the next two years he was receiving about three offers a day to purchase a ticket to Lima. What happened?
What happened is that they have spied on his searches, but they did not know that he actually purchased the flight. So in both cases, we have bad input poorly constructed– input data– which is processed by a badly implemented algorithm. In the first case it was a simple statistical algorithm. In the second case, we don’t really know a simple data processing algorithm. That is small wonder that it leads to a bad outcome. Now don’t misunderstand me. I love SMART technology. I love if it makes my life easier. I love when it makes it faster to accomplish something.
And I also love even that it saves me for thinking, as long as it is about thinking about things that I don’t want to think about in the first place. However, the experiences, that purchasing a flight on the internet today takes me longer than when 15 years ago, in Budapest, I had to get into my car, to go to the other side of the city, to find my travel agent, and purchase the ticket. Now the first thing is– it is on the first thing that it is slower. But it also means that I cannot get such good deals or such interesting trips that he was able to recommend to me.
And, of course, he knew me very well, because I was working with him for several years, and he knew what I was interested in. So his recommendations were always relevant. But there is even a third problem– that I constantly need to be alerted when I am purchasing a flight– that I don’t purchase some additional insurance– as long as I am already covered twice, or some unknown membership, or to book a pair of ski when I am going for a summer holiday. So these are what I really don’t like. So I would love SMART technology if it was at least as smart as the not so smart people. Otherwise, it is major frustration.
Information systems are supposed to make our lives easier. However, we find far too often that their recommendations are incredibly bad. Why? Because their recommendation systems work like the shaving machine from the earlier video: they cannot handle customising it for real individuals and therefore it is customised for the average.
The problem is that the average does not exist. Sometimes you can calculate it, for instance, you could take everyone’s shoe size and calculate that the average shoe size is 8.37. Only there would be very few people who could wear those shoes. This also means that this is not an AI problem; it is not the AI that is not smart enough, the problem is that we use elementary statistics instead of using AI. Why? Probably because it is simply cheaper. For instance, there is no technical reason for why we would not have holistic dynamic traffic regulation. We could have cameras recording the traffic on all the roads, adjust the traffic lights, the speed limits, re-route the GPS (global positioning system) devices, etc. They could include car park availability, and this would work even better if it was linked to our calendars so that the traffic regulation could know how long I plan to stay in my parking spot. We could go even further. Each of us could have specific driving skill scores which would determine individual speed limits. Furthermore, if I am meeting my daughter’s primary school teacher because she misbehaved and my personality profile says that I can be nervous in such cases, my car simply would not start. Probably there are not too many things that would make me even more nervous then. Do we really want that? We need to figure out some sort of balance between intrusion into privacy and customising to the average, between having everything personalised and rules that are simple to apply. A machine cannot figure this out, only a human.
This article is from the free online

Understanding Information and Technology Today

Created by
FutureLearn - Learning For Life

Our purpose is to transform access to education.

We offer a diverse selection of courses from leading universities and cultural institutions from around the world. These are delivered one step at a time, and are accessible on mobile, tablet and desktop, so you can fit learning around your life.

We believe learning should be an enjoyable, social experience, so our courses offer the opportunity to discuss what you’re learning with others as you go, helping you make fresh discoveries and form new ideas.
You can unlock new opportunities with unlimited access to hundreds of online short courses for a year by subscribing to our Unlimited package. Build your knowledge with top universities and organisations.

Learn more about how FutureLearn is transforming access to education