And the method of testing difficult ...

The rational method of trial and error leads to a better result than impeccable intellectual planning .

David Kelly , founder of IDEO ( www.ideo.com )

Being as a listener, and sometimes acting on various forums devoted to innovations, science, business, I often hear from experts the following thought: "The state has ceased to set large -scale tasks." We are talking about the tasks with a capital letter, about some large-scale goals, for which you should go at risk. About those very noble goals, the achievement of which brings not only material and moral satisfaction, but also benefits to all of humanity or its significant part. Of course, when they begin to talk about such purposes, they always recall the space program - Soviet and American. However, we forget about an important historical lesson and an important tool for the implementation of incredible creative energy, which is released when the level of talent for the very advanced part of society corresponds to the level of the task - competitions aimed at solving specific, useful problems. The essence of such competitions is simple. An interested organization or institution (often in this capacity, the state in the person of its agent performed) announces a competition for solving a complex problem.

The history of such competitions originates in the 17th century, but they gained real popularity only in the XVIII century. I will give only one example: the task of accurately determining longitude during long -term sea travels was considered so difficult and important that the British parliament in 1713 declared a monetary prize with the size of 20 thousand pounds (about 5 million dollars at the current rate). The English watchmaker John Harrison managed to solve this problem and presented the first chronometer in 1735. A little later he was awarded a well -deserved prize.

Since that time, historians of science and technology have more than 200 significant competitions for solving specific problems. The speed of scientific and technological progress is growing rapidly-the Moore law allows you to create all new technologies that give a solution to problems that seemed insoluble yesterday. The US government is very actively supporting the initiatives of the federal authorities and private organizations to make certain significant problems to solve the public-now there are about 400 different scientific and technical competitions on the Challenge.gov website.

Among the most significant and recent, I note the Darpa Grand Challenge to create a vehicle that allows you to drive through the desert Mohava (USA) 211 km in completely autonomous mode. Thus, the Darpa agency intended to receive an autonomous medical transporter that would evacuate the wounded fighters from the battlefield. In 2007, several teams successfully coped with the task. And although Darpa did not achieve its goal - the tow truck was not created - the agency solved a much more significant problem (such accidents are often called the good English word Serendipity) - founded a whole industry of autonomous vehicles.

Conducting competitions to solve specific problems has a number of important advantages over the “beauty contests” of innovative projects beloved in our country. First of all, it is a focus on strategic goals without the need to choose a specific team that will reach them. The customer, often the state, pays for the actual result. At the same time, the country - the author of the competition - independently strengthens the competencies and knowledge in the field of the competition. The creation of a community of developers - soliders of similar problems - becomes the most important by -product of such a competition. Private investments are stimulated, exceeding the size of the prize funds in terms of value. As mentioned above, Darpa Grand Challenge led to the creation of a multi -billion dollar industry of autonomous vehicles.

In our country, in the context of competitions, they often recall the flights of Chkalov and other Soviet aviators. However, the internal closeness and economic altruism of Soviet society did not allow fully realizing the technological potential from such competitions-none of the then-record holders entered the series, and strategic aviation was long a weak link in the USSR defense. The reason for this, in my opinion, was that the broad masses were not involved in the very solution of problems - they assigned the role of “enthusiastic viewers”.

Among other things, competitions attract public attention to solving the selected problems, to the best people of the country who solve them. In the end, the broadcast of such competitions on television is better than “Dom-2” or “Voice”. Their main result will be in our confidence that Russia "can, if it wants." Generalization of international experience allows you to formulate a number of important principles for conducting such competitions. First of all, this is the maximum involvement of citizens who themselves believe that they can solve such a problem. In addition, the organizers of the competitions should show openness and honesty when setting the task and determining the winner. The completed competition does not mean that the work is completed - post -competition with the community created is no less value than the solution to the problem itself. But the business, the industry can be created only when all intellectual property remains with the winners, and the state receives an unescruded indescribable license.

Among the ideas that are currently being worked out as part of various initiatives, I highlight the following. "Winter route." The essence of the problem is that at the moment the task of the autonomous management of the vehicle in bad weather conditions (snow, rain) is almost unresolved. The task is to achieve the most safe driving in autonomous mode in the winter season, taking into account the possible lack of road marking, low distinguishing between the roadway.

"Formula X". Many have heard about the performance of Turing test, in which the computer impersonates a person. But the same idea can be applied to robot cars. Why not create a racer-robot who will ever overtake a person? We can conduct races of unmanned, autonomous vehicles. Similar tasks can be developed and implemented using our engineering community and in other areas.

Artificial intelligence technologies received a new start due to a huge breakthrough in the creation of deep learning neural networks. For the first time in history, artificial intelligence managed to solve better than a person not only of chess tasks, but also purely practical things, such as the recognition of images. In this area, our country has a good tank. In a recent prize competition, held by the Institute of Artificial Intelligence Paul Allen (co-founder of Microsoft), the Russian team took seventh place ( www.kagghe.com/c/the-allen-scieence-Challenge/leaderboard ) of 170 international participants. In such a significant area, we could also hold interesting competitions by raising domestic linguistics and computer science to a new level.

Life, logic of scientific and technological progress and the fourth industrial revolution give birth to new challenges that require new approaches. The old approaches based on the state order and the choice of winners based on the “best papers” do not meet the task of the technical and technological modernization of Russia.

Albert Efimov ,
Head of the Robotics Center Skolkovo