Tuesday, 12 February 2013

The Black Cat: Existentialism in Cognitive Science



We've all seen the videos—YouTube’s library of “adorable kittens” is famous and infamous among students and teachers everywhere. While most of these cats love life and everything in it, one in particular has a more grim outlook. His name is Henri.

Henri puts a new spin on a classic question of philosophy of mind and cognitive science: what is it like to be a cat? “My thumbs are not opposable. Yet I oppose everything… I am like a pendulum that does not swing.” “I am free to go. Yet I remain… I alone feel this torment.” “I try to tell the doctor about my depressive state and my growing disillusionment with the world… They pronounce me healthy, as always.” While undeniably bleak, the musings of “le chat noir” provide an interesting framework for the analysis of the mind: the “not-human.” The implications of this statement go beyond mere animal cognition; while it is easy enough to anthropomorphize Henri, the existential philosophy behind his ideas is much more interesting when considered in light of artificial intelligence and neural networks.


What role should existential phenomenology have in the construction of artificial intelligence? Some scholars argue that philosophy provides the essential constraints for the construction of an accurate theory of mind in cognitive science (ex. van Gelder).  Others, however, say that philosophical concepts simply must be considered during artificial intelligence and neural network research so that these systems can accurately approximate all aspects of the human mind, not just the computational (ex. Wrathall and KellyDreyfus). 

According to this line of thought, if the mind is the brain plus “something else,” any model or network for the mind must account for both the biological hardware and the more abstract “something else.” In this way phenomenology serves as both a regulator and a catalyst for model development; it forces models to remain within the constraints set by the actual experiences of the human being, while simultaneously demanding they go beyond the traditional analytical parameters of computer science and approach some problems, such as experience, in novel ways. 

As is always the case when combining science and philosophy, these theories are controversial. It is unclear what role phenomenology should actually play in the construction of artificial minds. Nonetheless it is difficult to dispute that it must contribute on some level, even if it is only an ideological one.

For those interested in better understanding the non-human mind, I encourage you to study the reflections of Henri, conveniently consolidated into a playlist here.


2 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. I think the area of 'phenomenology & artificial intelligence' is a really thought provoking one. Although phenomenology stems from a philisophical enterprise, the research/literature on phenomenology and its applicability to human- robotic interaction is captivating.

    "Unless a particular behavior is meaningful to a robot and not just the programmer, the robot will never function in a truly autonomous fashion nor be able to make
    decisions through mechanisms that even remotely resemble the mechanisms by which we make decisions"

    Here is the link to a 2011 paper on phenomenology and robotics from which the above quote comes from.

    http://www.ipsi.utoronto.ca/sdis/Robot-Hegel-final-proof.pdf

    ReplyDelete