Читать книгу Societal Responsibility of Artificial Intelligence - Группа авторов - Страница 11

A practice that corresponds to appealing to the general public or consumers to propose and create elements of the marketing policy (brand choice, slogan creation, video creation, product ideation/co-creation, etc.) or even to carry out marketing services. Within the framework of crowdsourcing, professional or amateur service providers can then be rewarded, remunerated or sometimes only valued when their creations are chosen by the advertiser or sometimes simply for their participation effort. Crowdsourcing has especially developed with the Internet, which favors the soliciting consumers or freelancers through specialized platforms.

AI appears as an essential evolution in the processing of digital information. It represents the part of computing dedicated to the automation of intelligent behaviors. This approach is the search for ways to endow computer systems with intellectual capacities comparable to those of human beings. AI must be capable of learning, adapting and changing its behavior.

The idea of elaborating autonomous machines probably dates back to Greek antiquity with the automatons built by Hephaestus, reported notably in the Iliad (Marcinkowski and Wilgaux 2004). For Brian Krzanich, President and CEO of Intel (the world’s leading microprocessor manufacturer), AI is not only the next tidal wave in computing, but also the next major turning point in the history of humankind. It does not represent a classic computer program: it is more educated than programmed. It is clear that the AI lawsuit has mixed fantasy, science fiction and long-term futurology, forgetting even the basic definitions of the latter.

Thus, the concept of AI² is to develop computer programs capable of performing tasks performed by humans requiring learning, memory organization, and reasoning. The objective is to give notions of rationality, reasoning and perception (e.g. visual) functions to control a robot in an unknown environment. Its popularity is associated with new techniques, such as deep learning, which gives a program the possibility to learn how to represent the world because of a network of virtual neurons that perform each of the elementary calculations, in a similar way to our brain.