Sujeito e decisão na sociedade de dados

• Autor(a) principal: Brandão, André Martins
• Data de Publicação: 2017
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da PUC_SP
• Texto Completo: https://tede2.pucsp.br/handle/handle/20603
• topic: Sujeito; Decisão; Sociedade de dados; Inteligência artificial; Subject; Decision; Data society; Artificial intelligence; CNPQ::CIENCIAS SOCIAIS APLICADAS::DIREITO

The investigation of the problems of the subject and of the decision making in the present society guided this research. The systemic method was chosen for this study, specifically the systemic constructionism. The researched subject can not be considered out of his time. Therefore, a model of history founded on the human capacity for creation, the poetic, and on a spiral vision of recurrence of things was approached. It was assumed that in the current time the world is going through a technological revolution. The industrial society, based on the production of material goods, changes to a data society, founded on information flow, control, and automation. Massive amounts of data are produced daily, generating big data, and technologies capable of decision-making - artificial intelligences (AI), created for task automation, are also developed. They have been used to make decisions in the place of human beings or to provide relevant information to decision making. The meaning of artificiality was questioned. It was observed that artificial is all that is created by the human being, as opposed to what is natural. In this sense, the contemporary world is more artificial than natural, considering it was largely created by human ingenuity. The subject was modeled as a simple behavioral system, the apparent complexity of its behavior is determined by the complexity of the external environment in which it is found. It is a minimal model of characteristics of the subject’s internal environment, which limit the process of adaptation to the external environment. In addition to these minimal features, everything else is artificial - created by the human. The subject was treated as a semantic-pragmatic atmosphere, a symbolic system that accumulates meanings about the world and acts according to them. Intrinsic differences between humans and machines, such as the boundaries between emotion and reason, still prevent a total mimicking of the functions of the former by the latter. In spite of the impossibility of complete mimicry, artificial intelligences can be useful tools for human decision-making, providing unseen or unimagined patterns and correlations. However, risks such as depersonalization and objectification of the human being may arise from this use. Regarding human decision-making, a model based on game theory was investigated, demonstrating that, in the field of law, every legal decision is also political, and that there is no right or true decision, just the possible one for the case. As for the machine’s decision, it was investigated modes of decision-making and consequences of AI and big data technologies. It was verified that current technologies use empirical techniques, mainly the inductive method, speaking the language of probability. The predictive potential of AI and big data technologies, including the possibility of recognition of patterns in large data sets, is shown as a possible legal tool. It does not consist in the complete automation of decision, in a substitution of the human, but in a combination of human and machine. Such technologies can not be expected to solve old problems inherent to the application of law. It is just a new way to work, which can generate as many problems as other ways, or even more. It is necessary to know the operation and structure of AI and big data technologies in order for them to effectively assist in human decision making. Ignorance in this matter may lead to the opposite effect