A new tool for the epistemological analysis of “alternative paradigms” in Artificial Intelligence

Date/Time
Wednesday
18 Apr 2012
12:00 pm - 1:30 pm

Location
470 Stephens Hall

Event Type
Brownbag

Veronica Sanz
STSC Visiting Scholar

Artificial Intelligence (AI) is the discipline within Computer Science that has played a key role in the development of its epistemic and ontological basis. As a project not (mainly) devoted to design concrete products for the market but conceived as an experimental field, AI systems “materialize” the ontological and epistemic assumptions underlying their models, theories, and design methodologies.
Mostly philosophers but also some STS and feminist scholars have paid attention to Artificial Intelligence as an object of investigation and criticism, particularly to the so-called Symbolic approach in AI, the main paradigm for many decades. In particular, Feminist Studies of AI have claimed that Symbolic AI, as all technological practices, involves gender assumptions in the use of the traditional epistemological model of ‘S knows that p’.

However, in the last 20 years some AI practitioners have challenged the traditional paradigm developing “alternative paradigms” as the case of Situated Robotics or Affective Computing. Such is also the case of Soft Computing (SC). Defined by Lofti Zadeh in the early 1990s, SC refers to a collection of computational techniques that try to model and reproduce types of knowledge and behavior for which more precise tools of the past manifested incapable to solve.

The aim of my research is to investigate whether the “gendered character” of Soft Computing may be different than the one of Symbolic AI. By applying the feminist critiques to the ‘S knows that p’ of Symbolic AI to the case of Soft Computing I conclude that, in this case, we require a slightly different tool of analysis which includes one more level of reflection. I name this new tool the ‘D designs ‘S knows that p’’ model, which moves a step forward the ‘S knows that p’ model.

This event is sponsored by CSTMS.
Additional sponsorship comes from:  Berkeley Program in Science and Technology Studies