Abstract

This work-in-progress study addresses three paradoxes in the relationship of humans and intelligent machines, i.e., machines with infused artificial intelligence (AI). The paradoxes refer to (1) the dependence of people on AI, (2) the long-term effects of human-AI relationship, and (3) the virtualization of objects. As per the first paradox, it is reasonable to assume that AI is expected to help people be more autonomous in their daily routines and professional duties, but reality shows an increasing dependence of people on AI tools for all sorts of interests and needs, to the extent that Sam Altman, OpenAI’s CEO, foresees that people will prefer (or be bounded to) to rely on AI decision-making rather than on themselves in the future. There is, therefore, a paradox involving the current intents and the future prospects of the use of AI by people, making it appropriate and relevant to apply the Access-Behavior-Cognition model of technology use effectiveness (Bellini, 2018) to understand the environmental and personal configurations that facilitate or limit an individual to use AI in purposeful, safe, enjoyable, and autonomous ways. As per the second paradox, we are studying whether a future where all human needs are satisfied by the agency of machines is ecologically viable, given the Schopenhauerian transitory-satisfaction model that describes human motivation and behavior. Finally, as per the third paradox, while the sophisticated representational technologies (Burton-Jones & Grange, 2013) resolve numerous physical constraints, those same technologies serve as a medium to spread fabricated data and even fabricated selves. Such a possibility makes emerge the paradoxical ontological situation where physical processes (e.g., face-to-face meetings) are needed to validate certain digital objects (e.g., one’s impersonation) that were precisely created to eliminate the need for their material instantiation (e.g., in an online job interview). We are studying the three paradoxes through multiple theoretical framings (e.g., phronesis, mindfulness, ELIZA effect, and Collingridge dilemma) and behavioral experiments with Turing tests and Oxford-type debates.

Download

Share

COinS