Can Artificial Intelligences Suffer from Mental Illness? A Philosophical Matter to Consider

Abstract

The potential for artificial intelligences and robotics in achieving the capacity of consciousness, sentience and rationality offers the prospect that these agents have minds. If so, then there may be a potential for these minds to become dysfunctional, or for artificial intelligences and robots to suffer from mental illness. The existence of artificially intelligent psychopathology can be interpreted through the philosophical perspectives of mental illness. This offers new insights into what it means to have either robot or human mental disorders, but may also offer a platform on which to examine the mechanisms of biological or artificially intelligent psychiatric disease. The possibility of mental illnesses occurring in artificially intelligent individuals necessitates the consideration that at some level, they may have achieved a mental capability of consciousness, sentience and rationality such that they can subsequently become dysfunctional. The deeper philosophical understanding of these conditions in mankind and artificial intelligences might therefore offer reciprocal insights into mental health and mechanisms that may lead to the prevention of mental dysfunction.

Legal, regulatory, and ethical frameworks for development of standards in artificial intelligence (AI) and autonomous robotic surgery

“…Professional groups may also lobby against the use of AI and autonomous robotic surgical devices (as seen in the case of J&J's FDA approved automated sedation system; this was intended to replace anaesthesiologists). Furthermore, it may potentially be too early to discuss whether robots can have psychiatric problems, as currently that question does not apply to anything on the horizon for surgical robots.…”

Addressing the ethical issues raised by synthetic human entities with embryo-like features

“…But we have already noted that synthetic biology methods might be used to generate post-embryonic human cerebral and neural tissue organoids that likewise present complete and active pain pathways, possibly even in childhood or adult forms. Beyond sentience and pain, ethical concerns have been raised in both biological ( Greely et al, 2007 ) and non-biological ( Ashrafian, 2016 ; Coeckelbergh, 2010 ) contexts with the possibility of generating entities with consciousness and self-awareness, and these concerns could potentially also be triggered by very sophisticated brain organoids. Moral concerns could also arise with other types of human entities, as illustrated by our ESATE experience above.…”

Correcting the Brain? The Convergence of Neuroscience, Neurotechnology, Psychiatry, and Artificial Intelligence

“…Yet AI has no interests beyond the completion of these actions, and so nothing in this process can be said to be better for the AI itself than anything else from an imagined AI point of view. We pass over the question of whether AI could exhibit something like irrationality or mental illness, though perhaps reflection on this idea could aid in understanding HI and the potential disorders of humans in general (Ashrafian 2017 ).…”