Deep learning through artificial neural networks enables machines to develop their own learning methods in order to respond to very specific problems. In other words, this technology allows artificial intelligence (AI) to learn how to learn, which raises the question of the limits of AI’s learning capabilities and, by extension, of whether human intelligence itself can be digitalised.
This article supports the view that human learning and the knowledge that results from it will always be the sole domain of humans. Regardless of AI’s technological complexity and computing power, it will never have access to the same knowledge as human cognition. Only a conscious subject can have an intentional and qualitative relationship with the object of their learning and with the process of learning itself. This special relationship therefore allows the subject to ascribe meaning to their learning experience. Conversely, AI must resort to mimicking this process or to developing its own learning process without ever being able to make sense of its knowledge.
This leads us to conclude that human and social sciences have an important role to play in the development of digitalisation, as by defining and challenging meaning, they ensure a better distinction between (and thus a better understanding of) human and artificial intelligence.
The whole essay can be read on our Italian page…
About the author
Kevin Reymond, graduated in Philosophy and Social Anthropology at the University of Lausanne, is an intern at the ASSH as a specialization within the Master “Discourse analysis and public communication.”