Ethics and Mathematics: why cod morality

We have already found out how important Cyberetics in the modern world is, and even presented a world in which a person and a car merged together. The future is not defined unambiguously, but we are building the present here and now. And today there are a number of problems that relate to everyone who is present in the virtual space.

One of the key questions that gets up during digitalization is how to encode ethical issues at the level of mathematics. In China, the state boldly answers this question - the algorithms of building a social rating are taught in accordance with the fact that the authorities consider it correct. In Russia, these issues are practically not discussed.

Also on the topic: Are your rights real in the virtual world?

Ivan Ninenko , a member of the Future Foundations Expert Council, gives an example from personal experience: one official dealing with digitalization issues has received the task of increasing the fundraising on alimony. For this, you can, for example, limit the debtor's access to public services. However, the question immediately arose: to what level to limit them? It is not clear who should solve this problem and bear responsibility for the result.

The problem of the trolley

“With the spread of automation, we begin to delegate decisions to autonomous agents. For example, self -governing machines , ”the expert comments. During the moderation of ethical problems, we have the necessary tools (limit access to technical benefits, for example), but we have no idea about the boundaries of these sanctions.

Another example: optimization of the Moscow transport system. We have the value of the average speed of the entire flow of machines as a whole, part of it falls on public transport. We can regulate the situation on the roads, starting from traffic lights (they are connected to a single system), ending with markings. Suppose we need to increase the average speed of cars. What should be the speed of public transport? Mathematically can be determined, but a person should solve the issue of priority, because the computer only calculates the function.

“Algorithms will make more and more decisions, which optimizes the system. But it depends on us what parameters to enter. For example, what priority should an ambulance get? This is a question that requires a clear answer. This is the complexity of ethics during digitalization. It requires to encode the system that is now between the lines, ”summarizes Ivan Ninenko.

We understand what is good and what is bad. But in the coefficients (namely, algorithms work with them) this is difficult to express. A classic example is the problem of a trolley that is now interested in everyone in the context of programming unmanned vehicles.

A heavy uncontrollable trolley rushes along the rails. On the path of its following are five people, tied to rails with a crazy philosopher. Fortunately, you can switch the arrow - and then the trolley will go on a different, backup path. Unfortunately, there is one person on the spare track, also tied to rails. What are your actions?

Should autopilot evasive a hare? Each driver himself makes this decision. But unmanned cars need to be programmed to solve in such situations. We are approaching to unify decisions that were previously the ethical choice of each individual person.

Law, Order and Ethics

According to Lilia Zemnukhova sociologist, a researcher at the Sociological Institute of the Russian Academy of Sciences and the Center for Science and Technology of the European University in St. Petersburg, ethics in the digital world should be considered in two planes: legal and moral and moral.

Legal is related to how technological solutions will be implemented in practice. As soon as something goes wrong (and this always happens with new technologies), it should be clear who is to blame and what to do. Either the technologies have already been inscribed in understandable social relations, and they become intermediaries (for example, financial transactions and their hacks where banks are responsible), or they produce new entities like robots or unmanned vehicles where there are a lot of participants, and the responsibility is distributed (developers, corporations, users, state, insurance companies).

In the legal plane of ethics in the digital world there are many unresolved issues. For example, Sberbank, launching persons for accessing banking services, hardly asked users permission to use personal data. But in this case, he takes on all responsibility. Trusting the information about himself to the state or corporations, we seem to subscribe to their responsibility for “correct” decisions. Lilia Zemnukhova

Another example that is understandable to everyone - the Roskomnazor is black lists , depriving us of access to information. While we can bypass its restrictions, we do not feel a big threat. We must be aware that with digitalization, we quietly find ourselves in the power of someone's decisions. “I want to emphasize that the legal side of the case is always regulated by the state, and we need to participate in this process, because outsourcing ethics is impossible, and ethical issues will never be resolved by themselves and should become the destiny of users no less than developers in every new case,” the expert summarizes.

As an illustration of this situation, Lilia Zemnukhov cites the Corporate Report of Microsoft about the development of artificial intelligence (AI). It focuses on the fact that the technologies do not solve our problems - they are in the area of ​​responsibility of not only developers, but also all users (including institutional players - business, state, society). If one of them decided to use artificial intelligence for their own purposes, he is obliged to think about the consequences. If this actor is not able to evaluate risks, it is necessary to gather expert teams, including representatives of social and humanities.

And then the second problem that we indicated above arises: moral-moral. It deals with ideas and developments, and not with technology. Already at the design stage in technology, principles related to ideology, culture, and values ​​are invested. It is then that breaks and displacements begin: prejudices, stereotypes, normative and rational picture of the world are fixed. This can be an uniform issue for developers and absolutely unacceptable for users. To “configure” this sensitivity, the Institute of Electric and Electronics Engineers, IEEE), for example, formulated ethical recommendations for specialists in the field of AI development.

“In this document, a separate much attention is paid to non -Western values ​​and ethical concepts (for example, Buddhism or Confucianism). This in a good way is similar to a high-quality, point crowdsourcing, in which it can be taken into account how different ethical issues are resolved in different societies. The technocratic approach does not contribute to the creation of a complex and inclusive environment, and technology development is not only an engineering sphere. Therefore, ethical decisions should not be made by individual officials, ”said Lilia Zemnukhova.

Obviously, digitalization is the next stage in the development of modern society. We translate the document flow into a digit , give eternal life in the virtual world of cultural and scientific heritage, and also strive to make friends with computer algorithms. We do not yet understand what formulas can describe our ethical principles, but we understand that solving moral problems will not be simple.