字幕表 動画を再生する 字幕スクリプトをプリント 翻訳字幕をプリント 英語字幕をプリント Today, actually, is a very special day for me, 翻訳: Midori T 校正: Misaki Sato because it is my birthday. (Applause) And so, thanks to all of you for joining the party. (Laughter) But every time you throw a party, there's someone there to spoil it. Right? (Laughter) And I'm a physicist, and this time I brought another physicist along to do so. His name is Albert Einstein -- also Albert -- and he's the one who said that the person who has not made his great contributions to science by the age of 30 will never do so. (Laughter) Now, you don't need to check Wikipedia that I'm beyond 30. (Laughter) So, effectively, what he is telling me, and us, is that when it comes to my science, I'm deadwood. Well, luckily, I had my share of luck within my career. Around age 28, I became very interested in networks, and a few years later, we managed to publish a few key papers that reported the discovery of scale-free networks and really gave birth to a new discipline that we call network science today. And if you really care about it, you can get a PhD now in network science in Budapest, in Boston, and you can study it all over the world. A few years later, when I moved to Harvard first as a sabbatical, I became interested in another type of network: that time, the networks within ourselves, how the genes and the proteins and the metabolites link to each other and how they connect to disease. And that interest led to a major explosion within medicine, including the Network Medicine Division at Harvard, that has more than 300 researchers who are using this perspective to treat patients and develop new cures. And a few years ago, I thought that I would take this idea of networks and the expertise we had in networks in a different area, that is, to understand success. And why did we do that? Well, we thought that, to some degree, our success is determined by the networks we're part of -- that our networks can push us forward, they can pull us back. And I was curious if we could use the knowledge and big data and expertise where we develop the networks to really quantify how these things happen. This is a result from that. What you see here is a network of galleries in museums that connect to each other. And through this map that we mapped out last year, we are able to predict very accurately the success of an artist if you give me the first five exhibits that he or she had in their career. Well, as we thought about success, we realized that success is not only about networks; there are so many other dimensions to that. And one of the things we need for success, obviously, is performance. So let's define what's the difference between performance and success. Well, performance is what you do: how fast you run, what kind of paintings you paint, what kind of papers you publish. However, in our working definition, success is about what the community notices from what you did, from your performance: How does it acknowledge it, and how does it reward you for it? In other terms, your performance is about you, but your success is about all of us. And this was a very important shift for us, because the moment we defined success as being a collective measure that the community provides to us, it became measurable, because if it's in the community, there are multiple data points about that. So we go to school, we exercise, we practice, because we believe that performance leads to success. But the way we actually started to explore, we realized that performance and success are very, very different animals when it comes to the mathematics of the problem. And let me illustrate that. So what you see here is the fastest man on earth, Usain Bolt. And of course, he wins most of the competitions that he enters. And we know he's the fastest on earth because we have a chronometer to measure his speed. Well, what is interesting about him is that when he wins, he doesn't do so by really significantly outrunning his competition. He's running at most a percent faster than the one who loses the race. And not only does he run only one percent faster than the second one, but he doesn't run 10 times faster than I do -- and I'm not a good runner, trust me on that. (Laughter) And every time we are able to measure performance, we notice something very interesting; that is, performance is bounded. What it means is that there are no huge variations in human performance. It varies only in a narrow range, and we do need the chronometer to measure the differences. This is not to say that we cannot see the good from the best ones, but the best ones are very hard to distinguish. And the problem with that is that most of us work in areas where we do not have a chronometer to gauge our performance. Alright, performance is bounded, there are no huge differences between us when it comes to our performance. How about success? Well, let's switch to a different topic, like books. One measure of success for writers is how many people read your work. And so when my previous book came out in 2009, I was in Europe talking with my editor, and I was interested: Who is the competition? And I had some fabulous ones. That week -- (Laughter) Dan Brown came out with "The Lost Symbol," and "The Last Song" also came out, Nicholas Sparks. And when you just look at the list, you realize, you know, performance-wise, there's hardly any difference between these books or mine. Right? So maybe if Nicholas Sparks's team works a little harder, he could easily be number one, because it's almost by accident who ended up at the top. So I said, let's look at the numbers -- I'm a data person, right? So let's see what were the sales for Nicholas Sparks. And it turns out that that opening weekend, Nicholas Sparks sold more than a hundred thousand copies, which is an amazing number. You can actually get to the top of the "New York Times" best-seller list by selling 10,000 copies a week, so he tenfold overcame what he needed to be number one. Yet he wasn't number one. Why? Because there was Dan Brown, who sold 1.2 million copies that weekend. (Laughter) And the reason I like this number is because it shows that, really, when it comes to success, it's unbounded, that the best doesn't only get slightly more than the second best but gets orders of magnitude more, because success is a collective measure. We give it to them, rather than we earn it through our performance. So one of things we realized is that performance, what we do, is bounded, but success, which is collective, is unbounded, which makes you wonder: How do you get these huge differences in success when you have such tiny differences in performance? And recently, I published a book that I devoted to that very question. And they didn't give me enough time to go over all of that, so I'm going to go back to the question of, alright, you have success; when should that appear? So let's go back to the party spoiler and ask ourselves: Why did Einstein make this ridiculous statement, that only before 30 you could actually be creative? Well, because he looked around himself and he saw all these fabulous physicists that created quantum mechanics and modern physics, and they were all in their 20s and early 30s when they did so. And it's not only him. It's not only observational bias, because there's actually a whole field of genius research that has documented the fact that, if we look at the people we admire from the past and then look at what age they made their biggest contribution, whether that's music, whether that's science, whether that's engineering, most of them tend to do so in their 20s, 30s, early 40s at most. But there's a problem with this genius research. Well, first of all, it created the impression to us that creativity equals youth, which is painful, right? (Laughter) And it also has an observational bias, because it only looks at geniuses and doesn't look at ordinary scientists and doesn't look at all of us and ask, is it really true that creativity vanishes as we age? So that's exactly what we tried to do, and this is important for that to actually have references. So let's look at an ordinary scientist like myself, and let's look at my career. So what you see here is all the papers that I've published from my very first paper, in 1989; I was still in Romania when I did so, till kind of this year. And vertically, you see the impact of the paper, that is, how many citations, how many other papers have been written that cited that work. And when you look at that, you see that my career has roughly three different stages. I had the first 10 years where I had to work a lot and I don't achieve much. No one seems to care about what I do, right? There's hardly any impact. (Laughter) That time, I was doing material science, and then I kind of discovered for myself networks and then started publishing in networks. And that led from one high-impact paper to the other one. And it really felt good. That was that stage of my career. (Laughter) So the question is, what happens right now? And we don't know, because there hasn't been enough time passed yet to actually figure out how much impact those papers will get; it takes time to acquire. Well, when you look at the data, it seems to be that Einstein, the genius research, is right, and I'm at that stage of my career. (Laughter) So we said, OK, let's figure out how does this really happen, first in science. And in order not to have the selection bias, to look only at geniuses, we ended up reconstructing the career of every single scientist from 1900 till today and finding for all scientists what was their personal best, whether they got the Nobel Prize or they never did, or no one knows what they did, even their personal best. And that's what you see in this slide. Each line is a career, and when you have a light blue dot on the top of that career, it says that was their personal best. And the question is, when did they actually make their biggest discovery? To quantify that, we look at what's the probability that you make your biggest discovery, let's say, one, two, three or 10 years into your career? We're not looking at real age. We're looking at what we call "academic age." Your academic age starts when you publish your first papers. I know some of you are still babies. (Laughter) So let's look at the probability that you publish your highest-impact paper. And what you see is, indeed, the genius research is right. Most scientists tend to publish their highest-impact paper in the first 10, 15 years in their career, and it tanks after that. It tanks so fast that I'm about -- I'm exactly 30 years into my career, and the chance that I will publish a paper that would have a higher impact than anything that I did before is less than one percent. I am in that stage of my career, according to this data. But there's a problem with that. We're not doing controls properly. So the control would be, what would a scientist look like who makes random contribution to science? Or what is the productivity of the scientist? When do they write papers? So we measured the productivity, and amazingly, the productivity, your likelihood of writing a paper in year one, 10 or 20 in your career, is indistinguishable from the likelihood of having the impact in that part of your career.