Moravec III:
Complexity, Simulation, Explanation, Justification, Robopets, Respect, and Otherness (!).

As we wrap up our discussions of Moravec and the Robopet exercise, it's worthwhile to call attention to several concepts the clarity of which might be questioned. Among these are:

• Complexity: What is complexity and why is it important to Moravec's discussion? If there are so many ways to characterize and measure complexity, might we conclude that the concept itself will eventually be replaced by several others?
• Explanation: On Wednesday, someone insisted that we could never evaluate a robot's "intelligence" until we had a better understanding of what our own intelligence amounts to. What would an explanation of intelligence look like? What are your criteria for a "good" explanation, either in science or in everyday life?

  Aside:In Sunday's NYTM,the cognitive scientist Steven Pinker is asked whether all of human nature, "..even culture itself, [is] reducible to evolutionary biology?" He begins his reply: "I prefer the word unification to reduction." What is/are the explanatory goal(s) and distinction(s) he has in mind here?

• Simulation: According to a number of contemporary cultural theorists, one of the things that characterizes our "post-modern" age is the rise of simulation, the blurring of distinctions between the simulated and the "real", and the appearance of imitations that try to improve upon the real or simply lack an "original" altogether (e.g. Celebration, FL). But for our purposes, and in the concrete case(s) of AI, we might ask whether our ability to accurately simulate a behavior presumes a "good" explanation of that behavior, and whether the successful simulation of a human quality or behavior in a machine (e.g. one that passes the Turing Test) is enough to justify the claim that that machine ability/behavior is "real".
• Justification: The Robopet exercise was meant to inspire, among other things, some thinking about ethical judgements. The difference between our discussion of these judgements in class and the way we might talk about them in normal conversation, is that we'll be as interested in the beliefs underlying these judgements as the judgements themselves. The biggest issue in Moral Philosophy, after all, isn't the question of right and wrong but the question of how we can/should justify our moral beliefs and claims.
• Respect: The Robopet exercise also raises explicitly the question of respect, which is perhaps a more complicated term than it first appears to be. Although the roots of the word imply deference and heirarchy, a more recent view argues that symmetry and reciprocity in relationships is at the core of respect.
• Otherness: Questions about ourselves, our responsibilities, and the limits of our freedom, originate most often in our encounters with others. There is a large literature on "Otherness" and the thoughts and behaviors encounters with "The Other" can and should inspire (e.g. is my encounter with another person inherently a struggle to avoid becoming an "object" for her/him, or is the encounter inherently ethical, inciting responsibility first and thought second -- "Before it is gaze, the other is a face," writes Finkielkraut in The Wisdom of Love). Whatever their arguments, most analyses of "otherness" assume the basic humanity of the the Other (or at least that the Other is a living being, in the case of animals), but the Robopet exercise asks us to consider our encounter with the Other when the Other is a machine. What, if anything, does that change in our sense of what is involved in encountering the Other?

Moravec's Mind Children, chapters 3-6

We'll finally get to the chapter summaries of Moravec, and ask the presenters to incorporate their Robopet remarks into their presentations.

Tom Dean has been kind enough to join us again today, and his own response to the Robopet exercise is extremely interesting.

A(nother) Brief Introduction to Arendt's The Human Condition

Arendt's book is one of the most difficult we will read this semester, and perhaps we can share strategies for reading the text successfully on Wednesday. In the meantime, here are some questions that might help you identify and make sense of her arguments:

• In the Prologue (pp. 4-5), Arendt discusses the idea/myth that automation has freed (or can free) humans from labor's "toil and trouble", and ends with the remark that ".. we are confronted with .. the prospect of a society of laborers without labor .." In light of her (later) distinction between work and labor, how would you explain this remark to someone who hadn't had the pleasure (or whatever) of reading pp. 1-50?
• In Chapter 1, Arendt distinguishes between "human nature" and "the human condition". What is the importance of the distinction and what is purpose of thinking about the latter?
• In Chapter 2 (p. 38), Arendt writes "We no longer think primarily of deprivation when we use the word 'privacy,' and this is partly due to the enormous enrichment of the private sphere through modern individualism." How do the classical (e.g. Ancient Greek) notions of the "private" and the "public" differ from those Arendt thinks characterize the modern age? How does your own sense of the relationship between the public, the private, the social and the political compare with Arendt's and/or with the Ancient Greeks?

For Next Time:: 1) Send your written Robopet remarks to the CHV-L list, and make sure to read all the responses carefully (you might use this as a first opportunity to think about topics for your first essay); and 2) Read the first 50 pages of The Human Condition if you've not already done so, and prepare at least one question about the reading.

Back to the Syllabus