Platforms, algorithms and AI: Issues and hypotheses in the mediatization perspective

Interactional digital algorithm 339 As the algorithm encourages its interlocutors to remain installed in the same proposed original dimensions, the likelihood of renewal is reduced. The algorithm tends to be a closed learning system. As social participants learn the culture from the envi- ronments in which they participate, we can consider that frequent interaction with digital algorithmic devices also stimu- lates a reinforcement of environmental sameness, of remaining in a single niche, stimulating incompetence to learn new situations. The more the participants (passively) learn to interact with the algorithm – obeying the conformation it makes of the participants themselves and of their previous choices and their way of interacting in society – the less available they will be, when faced with new situations, to creatively learn how to address them through their own experimentation, whether on an individual level or in the composite terms of any type of so- cial action through which they are inscribed in the broad soci- ety. Waiting for a new automatic solution, it risks turning social participants into colonized by the interests of the algorithm and their own previous habits, now reinforced. The challenge posed by this issue is how to make the algorithm perceive/accept innovative angles to problems and accept personal or group selections in invigorating or critical perspectives that stimulate new learning about the world. Or more precisely: how to develop social interactions that circumvent the algorithm. d) Simplifying and conforming compositions of diversity The algorithm not only reinforces preferences – but, for each user, it selects those that favor an articulation with group patterns – without the user being aware of the group cat- egories to which they are being subscribed to. In other words: the algorithm generates a niche not consciously chosen, and not even perceived by users, but to which they will be adjusted and conforme H d u . man interactions that are strongly organized based on automated algorithms tend, due to this very automation, to be exercised in a way that is not very flexible, making it difficult

RkJQdWJsaXNoZXIy MjEzNzYz