Rodney Brooks & the Bioethics of Animated Machines

www.flickr.comphotos13963375N003533146556

A review of Creatures of Artifice: Rodney Brooks and the Bioethics of Animated Machines, by Nicholas S. Anderson.

Nicholas Anderson’s dissertation begins with an introduction to the puzzles of conceptual dichotomies such as human-animal, machine-animal, human-machine, life-nonlife occasioned by the rise of robots, particularly the biologically inspired robots pioneered by Rodney Brooks (1954- ), Australian computer scientist and former Panasonic Professor of Robotics at the Massachusetts Institute of Technology.

As the introduction rightly notes, before Brooks, the Artificial Intelligence (AI)/robotics community primarily took a top-down approach to AI, and so came up with chess playing machines and the like, but signally failed at apparently simple tasks (e.g. picking out one object from a set, walking down a flight of stairs) that human toddlers have mastered.

Brooks believed the reason for the failures was the approach: instead of top-down programming, he advocated building robots based on biological designs, and letting them engage in bottom-up “machine learning” in order to gradually learn how to perform such tasks. His approach has been wildly successful, with the Roomba vacuum cleaner merely the most commercial of his many breakthrough robots. And his machines are now routinely anthropomorphized, or perhaps a better term would be “canispomorphized,” as the Roomba is compared to a “slightly guilty puppy” (p. 2).  The result: our old concepts are being strained and revised, and old verities are liable to fall by the wayside, as technical advances make us reassess what it means to be human, or to be a machine.

Brooks’ popular text Flesh and Machines: How Robots Will Change Us (New York: Vintage, 2002) thus serves as the springboard for the dissertation, in the context of Brooks’ self-described research as “The Quest for an Artificial Creature” (p. 3) and the resultant potential transmutation of humanity into a “Lamarckian” species in which we become (partially?) the product of our own technology (p. 5). (Even for Brooks, we presumably will remain, for the foreseeable future, at least partially a product of our evolution, rather than wholly self-invented.) Whether it is described by “playing God” or “defying nature” or by some other loaded terminology, discussions of this sort inspire much discussion in bioethics, and Anderson’s dissertation enters into this conversation. Anderson puts it as follows: “What is at stake for me is the tentative cultivation of a bioethics that is responsive to redefinitions of life by technological means, one that stretches beyond the anthropocentric model, especially ascendant in the medical sciences, which is based on a discourse of human rights specific to modern Western culture” (p. 6).

In the chapter entitled “Curious Lives” Anderson assays the challenges Brooks’ approach poses to the AI community, particularly in his attempts to build a “sense of life, almost” (p. 28).  The robot he seeks to create “should do something in the world; it should have some purpose in being” (p. 34).  The puzzle that Anderson begins to unravel is what these purposes should be – and who or what should be choosing them. This leads to a discussion of Artificial Life as a synthetic biology; not the usual scientific approach of learning through dissection, but the attempt to “put living things together” (p. 36). Discussing social constructivist approaches indebted to Latour and Pickering as well as Foucault and Keller, Anderson then investigates how these developments will “synthesize new narratives about what it means to share a life with increasingly lively technologies” (p. 39), and takes up whether they commit the representational fallacy and Pygmalionism.

The next chapter, “Ways of Living,” discusses the Errol Morris documentary Fast, Cheap, and Out of Control, which features Brooks alongside three other “obsessed eccentrics” and in which Brooks explains one of the reasons why he calls his robots “artificial creatures”: “I like to think of them as prototypes towards entities that exist in the world and live in the world in the same way that animals live in the world.” (p. 49) Anderson then asks: “What is in the nature of his artificial creatures that approaches that of more natural ones? The most obvious answer has to do with autonomy and control” (p. 50). Pursuing  the idea that Brooks’ creatures are the result of the artificial selection due to the exigencies of institutional funding cycles rather than mere biological survival and reproduction, Anderson then investigates Pickering’s notion of an “evolutionary approach to design” (p. 52) and interprets Brooks’ insistence on “Intelligence without Representation” (p. 53) in his bottom up approach as  a way of life for his creatures in which doing is privileged over representation, and definitional issues (such as the boundary between life and nonlife) become less important. Following Latour and Andy Clark, Anderson endorses a “distributed model of intelligent life” (p. 67) whereby our subjectivity may be a by-product of our environment, and idiocy becomes crucial to the evolutionary design of humanity: “We use intelligence to structure the environment so that we can succeed with less intelligence. Our brains make the world smart so that we can be dumb in peace!” (Andy Clark. Being There: Putting Brain, Body, and World Together Again. Cambridge, MA: MIT Press, 1997, p. 180).

In the next chapter, “Steering to Other Ends; or, Modern Encounters,” Anderson discusses Brooks’ theme of the “merger” of bodies and machines, saying this goes “beyond cyborgs.” Nevertheless, Anderson contends that Brooks’ thinking still retains an individualistic anthropocentrism that seeks mastery over the material world, with the organic body arguably the privileged target of that mastery. Brooks’ arguments suggest that, even if humans and machines are to merge, “as long as engineers can be more clever than evolution, perhaps we may yet still be Man” (p. 75).

Anderson’s next two chapters (More Humanist Than Human” and “Anaesthetic Animals and Animated Machines: That Lifelike Feeling”) investigate “the ontological, ethical, and political implications of this high-tech remediation of the modern liberal humanist subject” (p. 75) that he finds expressed in Brooks’ thinking, using Martin Heidegger’s and Bernard Stiegler’s philosophies of technology to critique the received promises made by Artificial Intelligence, Artificial Life, and biotechnology more generally.  The critique enables a jumping off point for discussion of cybernetic organisms (in Cybernetic Organisms; or, The Tortoise and the Pussycat), including discussion (in Cybernetics: or, Control and Communication in the Animal and The Machine. 2nd ed. New York: The MIT Press, 1961) of Norbert Wiener’s cutting a cat’s quadriceps muscle and attaching it to a lever in order to record its contractions, contrasting this (p. 186) with Donna Haraway’s depiction of lab animals as workers who are “significantly unfree partners” (p. 72 in When Species Meet. Minneapolis: University of Minnesota Press, 2008).

Anderson argues that our inability to recognize ourselves as fragile machines has left us with a problem of adaptive feedback; cybernetics offers novel possibilities for bioethics based on  communicability and “ecological collectivity, rather than upon the rights of human individuals alone” (p. 189).

Anderson’s next chapter (“Unhomely at Home: Dwelling with Domestic Robots”) discusses the implications of two domestic robots, iRobot’s Roomba vacuum and the heroic Chibi-Robo (from Nintendo’s offbeat video game series).  In Flesh and Machines, Brooks discusses moving from mere Roombas to “pucksters”: “small, hockey-puck-sized robots with small legs that they use to slowly, slowly drag themselves around” (p. 119). Anderson avers that Brooks seeks “a household ‘ecology of robots’ … that, in turn, emerge into productive behaviour” (p. 196). But Anderson finds Brooks’ vision would accomplish far more profound changes, stating that:

Despite the marketing claims, Roomba and its cousins are not so much about increasing the leisure time of individuals by having more machines to do the work of everyday living. The labour they perform is much more productive than that. They reorganize the form-of-life, the bios, in the home: remapping the domains of labour, reorganizing the speeds and potentials of the household’s day, adjusting the affective climate within its four walls (p. 199).

Anderson suggests that such changes will alter the domestic economy, as economic changes inevitably ensue from those in the domestic ecology. He also examines the relation to the concept of the uncanny, using a Freudian and Heideggerian analysis to link it to “a strategic animism, by which the dynamic agency of the nonhuman becomes an acceptable object for ethical and political thought” (p. 202).

In his conclusion (“Ethical Robotics?”), Anderson discusses Brooks’ take in Flesh and Machines on the prospect of robot slaves, in which Brooks argues roboticists can create ethical ‘slaves’, just like our current refrigerators, vacuum cleaners, and automobiles, but with more autonomy. But Brooks warns that we may develop an emotional attachment to more intelligent robots with emotions and this attachment may cause us to feel morally responsible for their well-being, akin to the moral responsibility we feel for our children.

Anderson is deeply troubled by what Brooks says, especially by the “humanist ethics” Brooks assumes, even as he contemplates extending ethical responsibility to the non-human. This problem extends to rights claims; Brooks appears to believe that only when robots come to be sufficiently like humans in intelligence and emotions would we have a moral responsibility toward them; that is, only then would robots have rights. But Anderson finds the anthropocentric history of rights theories vexed by its relationship to the field of animal rights. Anderson’s critique of Brooks challenges us to consider the implications of what Brooks calls our sense of “tribal specialness” and “overanthropomophization” when it comes to our relationship with nonhumans.

Keith Abney
California Polytechnic State University – San Luis Obispo
kabney@calpoly.edu

Sources

Work by authors including Rodney Brooks, Jacques Derrida, René Descartes, Michel Foucault, Sigmund Freud, Donna Haraway, Martin Heidegger, Evelyn Fox Keller, and Bruno Latour.

Dissertation Information

Ryerson University. 2011. 236 pp. Primary Advisor: Stuart Murray.

 

Image: “Roomba Time-Lapse,” the path taken by a Roomba robotic vacuum cleaner as it cleans a room. Photograph by Chris Bartle.