Social and cultural contexts in the humanization of artificial intelligence
Nowadays, as soon as someone says “I want to study Anthropology”, “I want to study Sociology”, the following questions (and in almost the same order) are sporadically heard: @what are you going to do? ” “What can you work on? “Is it easy to get a job? “, “Should you make money doing this? “And you, a little insecure, maybe respond to things like” in social projects “,” with the mayor’s office or with vulnerable communities”, “there are always connections “and well from the last question or speech, if you studied Social and Human Sciences, it is because his aspiration in life was not exactly becoming an Arabian Sheik.
It is almost impossible to think about Social Sciences outside the field of social intervention, the academic field or the public / private field, because even at the moment we are in, we continue to live under traditional conceptual constructions that limit knowledge to specific fields of action, which although they are highly relevant, they are not the only ones. So, if you talk about social professions in the world of technology and even more in the development of AI, it will seem as crazy and crazy as if you were talking about the Earth being geocentric (round) in the 2nd century.
By definition, Social and Human Sciences are responsible for understanding social universes, describing them, questioning them and building a network of political, social, cultural, historical, communicative and ethnic contexts; to be able to build archetypes or social stereotypes, which allow to predict the action of a person for his life in community.
Now, if what I just defined does not help to clarify the relevance of this knowledge to socialize and even humanize an AI, let me explain further:
In essence, the social scientists who ask (for not delegitimizing years of struggle to be recognized as science), are in charge of investigating the behavior of a person or group of them; the reason for their actions, what motivates, mobilizes and governs them, what are the established social parameters of interaction and how they behave in these interactions; the last (the how) is the action, that is, it is the result of the why, what and what, and this action has different ways of expressing itself according to (and here comes a magic word to understand all this) o context, people act and communicate in different ways, according to the need of the moment, that is, the context; for example, you will not act and communicate the same way at a business meeting and with your closest circle of friends.
All social settings generally establish roles in interactions. Now, when you call a customer service line looking for a solution, you don’t start the call with a “Good morning, thanks for calling … what can I do for you”.
To teach Artificial Intelligence, there is a possibility that it can be trained under the characteristics of young children’s learning and program its algorithms based on basic human instincts, such as common sense.
Think of an Artificial Intelligence as a child, that you will begin to teach how to name things and people. So you teach him to identify that these things and / or people play a role, the toy is for fun, my mother is the one who takes care of me. Now, everything and every person is related to a feeling, my toy gives me happiness, my mother shelters me, the nurse who gives me the vaccines and scares me. Later, you teach her the ability to assign context: my toy is at home and I can use it when I have free time.
Now let’s do the same exercise applied to what concerns us: AI is a child who already knows how to write, however, by itself, it does not associate the meaning of the words; therefore, you must teach it the definition and teach it according to the context the words can be related to positive, negative or neutral feelings; you teach how to recognize people and different roles: agent, analyst, collaborator, salesperson, user, customer, affiliate, in short, how we want to name them and, finally, teach how to interpret this information so that she can live in society and interact naturally, as a social and cultural entity.
Science has shown that cognitive skills are developed in early childhood and that, because of these skills, humans can think abstractly, so it is unreasonable to include these training resources in an AI development program. Therefore, it is possible to perceive AI as human, as it has learned to the point of coherently reproducing our codes of relationship and behavior.
That is why understanding the social and individual behaviors and the forms of communication that people have in everyday life is a valuable knowledge to transmit to Artificial Intelligence, as they will learn and understand different contexts or types of business to identify roles. In the interaction, for example, who is the client and who is the agent, he will systematize and characterize the interactions according to the entire cultural and social framework mentioned above, with the aim of simulating human reactions in response to the need of the moment.