During the, the summer, 2008, I remember quite well, Wired magazine produces this cover story, The End of Science. So you can imagine my reaction. How is it possible to publish such a, such a paper? And we go back to the beginning of this lecture about the end of the end of the end of, and also this, a bit before there was this book, The End of History. The, it was, the thesis was, now that we have democracy and free market economy, it looks like we reached the end of history. [COUGH] Honestly I cannot understand how it's possible to write something like, the end of. And you can choose because if you go on Amazon and you just type, the end of, you will have hundreds of proposal, like everything is finished. Okay. The End of Science. So Wired is not a stupid magazine. So, why did they cover this type of thesis? It's the idea that the power of the machine could develops theories, concepts, et cetera. I told you, the former video, I don't think this is true. But I'm going to talk about the same question on a different mode. So a theory is built on causality, on causes. And we spoke about causes, about Jung. There is a strange paradox about cause. It's everywhere. We think constantly in terms of causes and effects. But we cannot see the cause everywhere. Its not accessible to, to, to the sense. That's, that's the paradox. So, on one hand causality is the cement of the world, it puts things together. It gives sense to what we have around us. And on the other hand, we cannot see it. We cannot see it. So now, Big Data, causality, the end of science, how does it fit together? Lets put it this way. Lets take two event, A and B. You have five possibilities in terms of connection. Number one, it's possible to have a coincidence. Imagine yesterday night the news on TV, like do you, on the headlines, Toyota just launched a new electric car. And in Paris, the French soccer team won against Morocco by 1-0. A and B, A and B, no connection at all. This is called coincidence. But in the end, it's quite an exception, because we always try to see connection. Imagine another first headline. Yesterday on Paris, thunderstorms, heavy rain. And France won against Morocco by 1-0. Slowly, we will try to see a connection. Say, maybe in Morocco, they have less rain, they are not used to play under the rain so much. And we try to find some explanation. And this is human being. We try to find cement, sense, connections. So real coincidence is in the end, exceptional. We always try to make connection. And what's the connection, is the second level. It's called a correlation. A correlation is a statistic link between two events. And we work a lot in terms of correlations. For example, [COUGH] unemployment is a huge problem. You can connect stas, statisti, statistically with events far away like apps sold on the iPhone. And you can have a connection say, if there are more of those apps sold, you can see more unemployment. This is correlation, correlation. It's statistical link. Another example, countries with lots of Nobel Prize are countries where people eat lots of chocolate. It's a fact, it's a correlation. If the chocolate consumption per inhabitant is high, the chance to have a Nobel Prize is high. It's a link, it's a fact, but it's not a cause. It's not because people eat chocolate, that you have Nobel Prizes. Nevertheless, there is a correlation. And at this stage, you can imagine like, a third reason why you enter the world of, of Big Data. I remember 20 years ago, it was already common in the marketing to connect the owner of some cars with some first name. Correlation. It's not a law, but it's a correlation. And you can do a lot about correlations. And Big Data is definitely a fantastic tool to find correlation in possible results or with a small machine. Sometimes you have situations like this A and B are the thunder and the lightning, or the lightning and the thunder. A and B. This is particular case why? Because there is a correlation with 100%, 100%. When you have lightning, you have the thunder. So it's not like the chocolate and Nobel Prize, it's 100%. And this has another name. It's called either a conjunction, or sometimes implication, implication. The thunder implies, is implied by the lightning. The lightning implies the thunder. But it's not the cause, it's not the cause. Implication was already analyzed by Leibniz. Leibniz, he was talking about two clocks. If you look at two clocks, they move exactly the same speeds. So there is a conjunction, an implication. But 100%, definitely, but there is no causality at all. So the first step, coincidence, correlation, conjunction or implication are a world where Big Data can help, can help. And eventually be useful to develop some new theory. But now we're going to go the next step. Causality. If you take the sound of a rooster, of a cock, and the sunrise. This is more than an implication. There is a kind of causality, of causality. So we enter one step further in a world where you can find a cause. So this is the beginning of the theory. But it's the end of the Big Data. Because how do you know exactly who causes what? Is it the rooster singing that cause the sunrise? Or is it because the sun is rising that the animal start to sing? It's not that easy, it's not that easy, this relationship. Who causes what? Who causes what? And to connect with the the topic of this lecture, remember, about children using PCs and technology. I read a book a long time ago called Mindstorms. And in the end, the author said, if you look at a child facing a computer, the question is, who is programming who? Is the child, while moving his mouse, programming the computer? Or is this message on the screen that in the end programs the child? So the causality, there is a causality but it's hard to see which is the causes, which is the effect. Which is of course, the last the last step of this hierarchy. Why a pyramid? Because you can come with a question. Do Egyptians buy, build pyramids? Or did pyramids build the Egyptian? You can answer bo, both are, are possible. It's very hard to know the direction of the cause. And of course, and then the ultimate step is, which is sufficient and what is necessary? So you have to summarize like six levels between A and B. From the coincidence to the end, the net, the cause and the direction of the cause, you have like a sophistication of the theory. I'm convinced Big Data can help and computer can help up to a certain stage. But the end, and we go back to the former video, to conceptualize, again, it's in your hands.
