With the rise of machines, an important number of skills are bound to disappear. But the emergence of new issues also requires new forms of human expertise. Facing worldwide problems that we are yet unable to solve, we need to develop different forms of intelligence, learn to cooperate and achieve results that aren't possible for individuals alone. Will our education systems, fundamentally based on competition, meet this challenge?
Paris Innovation Review – Are our education systems adapted to a world changing at increasing speed, with less vertical, hierarchical and increasingly horizontal and collaborative organization systems?
François Taddei – Our educational systems our based on the resolution of traditional problems. If they want to achieve success within this system, students need to pass an entrance exam that consists basically in solving common problems. But there are other forms of intelligence which involve resolving new problems. For example, in startups, hackerspaces or IT communities, participants are judged on their ability to do something that others had never done before. This involves a very different set of skills than those required for copying what others have done before, only faster. A third form of intelligence, even higher than the other two, is the ability to define the problem for oneself. The creators of Amazon redefined what being a bookseller meant.
The problem with the first form of intelligence (solving classical problems) is that machines know very well how to apply it. They know the solutions to traditional problems, they can improve them day by day and are even able to implement these same solutions in ever changing contexts. For example, nowadays, machines know to drive cars. There is a whole bunch of trades that are being replaced. Manual jobs but also trades such as financial analysis, which required, and granted, a high level of education and income. According to a recent study by the University of Oxford, 47% of jobs in the United States are likely to be automated in the next ten or twenty years. So if we are only able to solve classical problems, we take the risk of being replaced by machines! The education system needs to change, because in today's world, we need creative and innovative people who know how to work collectively.
Why is it so important to learn to work collectively?
We are facing worldwide problems that we don't know yet how to solve. We must therefore develop different forms of intelligence and learn to cooperate together to achieve things we can't do individually. Typically, companies hire different and complementary profiles. As a team, they can achieve results that none of them could alone. Learn to work collaboratively as soon as possible is a crucial matter, especially because we can learn a lot from one another. The better we are trained to listen and take into account different opinions, the more chances we have to design a product or service that different people will alike. If we think in a single dimension, maybe we will have a perfect solution in this unique dimension but with a lot of side or negative effects in other dimensions. If, right from the start, we involve people from different worlds, with different ways of thinking, we will optimize our multidimensional results. Because reality is multidimensional, it's easier to achieve a result that will actually work in the real world.
Most interestingly, the Organization for Economic Co-operation and Development (OECD) will begin to measure collaborative problem solving from 2015 as part of its PISA survey. We will soon realize that some countries are better than others at developing collaborative practices and supporting young people in their questioning. The debate, which until now had triggered only a limited interest among parents and teachers, is finally emerging in the public sphere.
What are your recommendations to develop collaborative practices in educational systems?
Historical and contemporary models share a number of constants. Among historical models, the classical paradigm of Socrates: questioning, dialogue and support; and that of Alexander von Humboldt, the father of modern universities. According to the latter, universities should be institutions where people are free to learn, free to teach and free to conduct research. Humboldt also believed that teachers should not only transmit knowledge but also act as mentors who support students in their projects. Thus, teacher should not be there to tell them what they must do but to support them in what they would like to do. This philosophy is at work in the Massachusetts Institute of Technology (MIT) and other major US research universities, but has very little represented in continental Europe. The Socratic and Humboldtian ideals share common principles: dialogue, mentoring and coaching, all in a context of freedom. Nowadays, different institutions are based on these principles.
For example, I just came back from Petnica, Serbia, where these methods are applied in high schools with great success. Everything started thirty years ago with a meeting between students and teachers during a kind of summer camp. Today, there is a boarding school with 200 places, a physics laboratory with lasers everywhere, a biology laboratory with a confocal microscope, a computer lab with a supercomputer provided by the CERN (European Organization for Nuclear Research), a fab-lab, etc. All of these tools are available to high school students. They can spend periods of 15 days and come back several times, allowing them to pose a problem, experiment, go back home, continue their questioning, come back to re-test their hypotheses and progress by iteration. The smartest of these kids are encouraged to become mentors for the next generations. 30,000 young people were trained in 30 years. It is interesting to note that 50% of young Serbians who have published in Nature and Science have been trained in Petnica and that the current Serbian Ministers of Finance and National Education studied in this institution.
Another example is the Catts Pressoir High School in Haiti, where students are asked to look for problems and share them to find a solution together. The traffic light system in front of their home was broken during a long time because of the earthquake and had not been repaired since. The kids wanted to fix it. They finally found a cheaper solution than the original and proposed it to the public authorities. In the same high school, students invented an internal mobile telephony system to be able to send text messages to each other without paying money to mobile operators.
In all of these situations, the principle is the same: the system allows students to ask themselves questions, do things together and by themselves – the teacher plays the role of the mentor. Could these collaborative methods could be applied in primary school? The question is worth examining: after all, why aren't children never invited to work in this way when they are perfectly capable of doing so? In fact, it mostly depends on their environment and the ability of adults to accompany children in their questioning. Often, when a child asks questions, he exhausts his parents and his environment and ultimately his own curiosity because of his inability to find partners to accompany him in his approach. Adults therefore need supervise the children that are being trained. As part of our project for the Centre for research and interdisciplinarity with forty classes, we trained doctoral students who accompany children in their questioning.
To summarize, all these different places have been reinventing what I believe Humboldt has formulated best: freedom to learn, teach, do research under the guidance of mentors to make progress. Humboldt thought of university students but the same principles are being applied to high school students in Serbia and we are trying to apply them to elementary students in the CRI with a strong emphasis on horizontality, cooperation, sharing, refining assumptions and ideas through discussion and experimentation.
Does this scientific approach work with everyone?
Alison Gopnik, a psychologist who teaches at the University of Berkeley, says that we were all born researchers: the youngest babies are already able to observe the world, to be surprised, to derive hypotheses from experiments, make mistakes and learn from their mistakes, dynamically revise their assumptions and attract the attention of others on what they have done. These components are precisely the essential components of any scientific process. Creativity is not just a matter of a few geniuses who are visited by a muse overnight. Everyone is born a creative researcher. However, without practice these innate skills quickly fade away. Creativity is a process that needs to be studied, accompanied, encouraged and learned: the same way we improve every time we make a backhand shot in tennis, we become increasingly creative each time we take the opportunity to exercise our creativity.
What matters most is being able to transform naive questions into a scientific investigation. By pushing scientific inquiry to the limits, we quickly reach the frontiers of knowledge. For example, when asked why water wets, a good physicist knows the answer. If asked to explain the details of his answer, a very good physicist can still provide an explanation. But when reaching the third or fourth level of this questioning chain, even a Nobel Prize winner will give up. When they realize that have reached the frontiers of knowledge, children are much more motivated. They won’t solve complex problems on their own, obviously, but they are very happy to participate and it motivates them a lot to learn what is already known and to try to go beyond.
Do cooperation and collaboration necessarily need to be opposed to competition and selection? Could we perhaps find a middle way?
I think that cooperation is much better than competition. Today, school systems are mainly based on competition. In France, for example, we tend to select the elite while discarding the others whereas as cooperation has much more benefits. However, there is an intermediate model that seems to work fairly well: “coopetition.” It consists in creating teams and putting them in competition. What’s true in soccer also works in student competitions. At first, I wouldn’t have necessarily believed it could work but I’ve seen students become world champions of synthetic biology at the MIT because they were motivated enough by this type of coopetition.
Most importantly, in these forms of coopetition, like in the MIT, all students must document in real time everything they do. The information they produce is therefore available to the other students, especially to the students from one year to another who can use it to go even further. It’s the creation of what I call “cooperative learning and innovation ecosystems”: each time someone learns something, someone else will be able to learn the same thing more easily; each time someone innovates in one dimension, someone else will be able to innovate in the same dimension more easily. Therefore, the purpose of coopetition is not to repeat the same test ever faster, as in competition, but to go further. It’s less about running a 100 meter sprint every year that about climbing new mountains. I'm not sure whether this model is appropriate for everyone but I do believe that coopetition is better than traditional competition.
To what extent does digital technology change things?
The advantage is that creating with digital technology isn't relatively cheap. Today, thanks to digital technology, we can afford making mistakes, an essential component of creativity. Digital technology also allows copying: we can recover files created by others, modify, mutate and recombine them very easily. As a biologist, I can assure you that mutation and recombination are the engines of evolution! Digital technology allows us to learn from what others have achieved to go one step further. That’s how collective creation emerges: we use different looks, different approaches and achieve previously undiscovered results.
Do you have any examples of the power of this type of creativity and collective intelligence?
The online collective encyclopedia Wikipedia is the classic example. In the field of programming, GitHub is also a good example: millions of coders trade millions of programs, collectively. The MathOverflow or PhysicsOverflow websites allow to ask questions and provide collective responses to these questions. There is a voting system to bring out the best questions and best answers, not a single answer like on Wikipedia but a cloud of solutions. The system highlights the reputation of solutions and the reputation of the people who have contributed to these solutions. Some people even include the number of points they earned on these platforms on their resume. Even geniuses sometimes need a group to move forward. Timothy Gowers, a professor at Cambridge who received the Fields Medal (considered the Nobel Prize of Mathematics), admitted on his blog that he had been stuck on a math problem for over two years. In three months, his problem was solved thanks to the contributions he received!
In recent years, many playful forms of collective creation have emerged. For example, the Fold-it game was created to understand protein folding, a very complex problem. All the knowledge on physical chemistry was incorporated into the game rules. Players were invited to fold proteins according to these rules and by using their ability to see in 3D and being creative. Some players solved problems that neither biochemists nor supercomputers had been able to solve! This is a striking example of the superiority of collective intelligence when it is channeled well: collective teams are often more efficient than individual players.