Does Modern Technology Actually Make Our Lives More Efficient?
LAUREN GILGER: Conventional wisdom is technology can make us more efficient. Dishwashers save us time in front of the sink and algorithms can help us quickly find a compatible mate and help employers weed out unqualified applicants. But our next guest isn't so sure that the latest and greatest tech is making us more efficient. Edward Tenner is with the Smithsonian Lemelson Center for the Study of Invention and Innovation and author of the book “The Efficiency Paradox: What Big Data Can't Do.” Our co-host Mark Brodie recently spoke with him.
MARK BRODIE: So let me ask sort of a defining-our-terms question. First off, how do you define efficiency?
EDWARD TENNER: I define efficiency very simply as getting more for less. And you know we can't get a free lunch, if we can't get something for nothing, then we can get more outputs for the same inputs or use fewer inputs for the same outputs.
BRODIE: And it seems as though that is something that technology is trying to increase, right? Like trying to make everybody work faster and more efficiently and be able to do more in less amount of time.
TENNER: But everybody wants it for different reasons. People who think green, of course, want more sustainability. They want lower carbon emissions. They want more miles per gallon, or if possible, all electric cars that will use energy even more efficiently. So everybody wants efficiency for reasons of their own.
BRODIE: Does it seem as though we're getting it?
TENNER: Well, in a lot of ways we are getting it. For example, if you look at miles per gallon, in the way cars are driven, or if you look at how convenient it is to order books or groceries or anything else online, life is much more efficient than ever.
BRODIE: It sounds like there's a “but” though.
TENNER: Of course, there's, you know, there's always a catch and there's always a cost. I'm all for efficiency but I'm pointing out some of the problems with efficiency and not abandoning the idea of efficiency, but promoting the idea that sometimes, lower efficiency in the short run will be good for efficiency in the long run. And conversely, if we become too efficient in the short run, we may be overlooking opportunities to become even more efficient in the long run.
BRODIE: Well, how do you see it that lower efficiency now might be better off down the road?
TENNER: The great strength of artificial intelligence and big data and algorithms is that they can really detect what have been the most efficient patterns in the recent past, and they can give us more of the same, so they can learn from experience and they're really terrific at that. The problem is that real efficiency in the long run depends partly on happy accidents. It depends on serendipity, it depends on our peripheral vision. And the problem, for example, of an efficient algorithm that you're using to find the kind of people who've been most successful in an organization is that it will learn very quickly what backgrounds, what personality traits and so forth have made people most effective. But the problem with that is that the environment is always changing, and if an organization doesn't have enough diversity, and that includes diversity of all kinds, then it probably won't be as responsive to changes in the long run. So you're probably better off having some people who have unusual backgrounds who might not be typical of your most productive employees but have something else to offer and they are a kind of insurance policy against change.
BRODIE: It sounds like what you're saying is almost that when people are looking at decisions they're making, they need to really put thought into it and not just try to emulate what somebody else has done or replicate what somebody else has done, but really try to think about what somebody else has done or really think about how that can be applied to themselves as opposed to just really trying to copy and paste from somebody else.
TENNER: That's right, there are quite a few studies that show how intuition or gut feelings can lead us astray and they're very right. But there are other studies that show that our intuition can really be a very good check on what automatic pilot is telling us. And speaking of automatic pilot, of course, we've seen what happens when for example aircraft companies are trying to build vehicles that are even more efficient in fuel economy and also easy to operate. But they don't take into account unusual circumstances that might be only once in a thousand times, or even less often. But if that once in a thousand times is a disaster, then it's really important to have pilots who have the skills and training to be able to respond.
BRODIE: Well is part of the problem here, do you think, sort of operator error, for lack of better word, in the sense that we're almost on autopilot ourselves and not really paying attention and not really thinking about what algorithms or what a formula might be telling us?
TENNER: Well I think that it's really good to have as much as we have automatic. If you compare the general safety record before and after that kind of automation, there's no doubt that autopilot generally makes things safer, provided that you have two things, provided that you have a redundancy in the system so that one glitch is not going to be fatal, so there is another part of the system that can that can kick in or that can warn you if something's going to go wrong, and you also need people who have the right training. So the argument in the book is really that artificial intelligence, machine learning and the kind of skills that people have are really complimentary, and instead of saying either or, instead of saying well we should go back to the old way, or we should kind of press for with complete automation, I'm saying that something that's really obvious that there is a really good balance and that it pays for us to explore that balance in our own lives.
BRODIE: All right well, Edward thanks a lot for taking the time to chat. I appreciate it.
TENNER: Thank you very much.
BRODIE: That's Edward Tenner with the Smithsonian Lemelson Center for the Study of Invention and Innovation. His latest book is called "The Efficiency Paradox: What Big Data Can't Do."