IT NewsFacts and forecastsTechnologies
Max Wolf | 07/11/2020
According to the publication ScienceDaily, researchers from the University of Warwick,
Imperial College London, EPFL (Lausanne) and Sciteb Ltd have developed a “principle of unethical optimization” that helps artificial intelligence identify and reject
“Unethical” schemes for working with consumers and the general public.
In a scientific paper published by the Royal Society Open Science, which is called the Principle
unethical optimization ”, scientists described the details of building the principles of selection of“ ethical ”, consistent with accepted moral standards and formal laws of strategies that can be applied to
commercial and other work with users of certain services, consumers of services and buyers of goods.
An example of unethical schemes is, for example, the following, described by a British newspaper, which conducted a study in 2018 and found out that if you enter data into the insurance calculator
company, then when entering the name Mohammed Smith, the conditions are more favorable than when entering the name John Smith.
Judging by their description, the prerequisite for conducting this kind of research was not so much the love for simple, not too burdened mathematical, that suddenly broke out in the scientific community
educated people, and the demands of regulators and businesses. The latter already today, or in the very near future runs the risk of running into gigantic fines – for “unethical” schemes that are used to
continuous increase in sales. Such schemes today are often selected not even by the sales manager, but by artificial intelligence, based on data on activity and behavior
The goal of manufacturers of goods and service providers is to offer them as quickly as possible, that is, at the right time, exactly what a specific person or organization wants to buy. Legal development
services (also quite a good seller) allows you to initiate a lawsuit for the “unethical” of such behavior of the seller. The consumer, learning about how the supplier of goods or services spent
about his preferences, it may be trite offense. But it may also turn to the regulator, who, in turn, has the right (if it is stipulated by the norms of the law) to set a serious fine.
It is obvious that the current size of fines and the likelihood of their application made regulators, business and scientists pay close attention to this problem and initiated the creation of the above
principles of “ethical” filtering, allowing commercial structures to rely only on harmless and law-abiding schemes.
Given the current trends to intensify the struggle for a wide variety of rights, as well as for equality and justice, the principle of filtering out any suspicious schemes that have a chance to be named
consumers “unethical”, may soon become one of the basic in the work of artificial intelligence.
The developers are confident that the application of the principle of optimization they created and the identification of unethical schemes will constantly improve the work of AI in the process of self-learning.
Magazine: IT News Magazine, Magazine Subscription