On Humans, Artificial Intelligence, and Oracles

Abstract

Our epoch is fascinated by human-like minds and by the fantasy of intelligent machines, which might surpass and substitute human intelligence. Computational machines were initially invented for unburdening humans from tedious and unrewarding tasks, often doing repetitive tasks better and quicker than bored human. Today, things are changing, and the mounting trend is to produce intelligent machines imitating human skills, such as intuition, emotions, sentiments, capacity for perceiving atmosphere and context, capacity for sense of humor, and so. The potentiality of next-generation artificial intelligence (AI) is thus in the limelight of the scholarly discussion and raises public concerns. In this chapter, I will develop a new argument, based on Floridi's Diaphoric Definition of Data, to demonstrate that computational machines might replicate human intuitive skills but they cannot exactly duplicate them. The chapter does not aim to grade natural vs. artificial intelligence, not yet to raise any ethical consideration on AI, rather it only aims to show the inherent limits of AI and its applications. Finally, I will conclude that the current debate about the hypothetical risk that AI might one day surpass human intelligence is largely misplaced.