Ross and Robots

By Stephen Robbins
Journal of Consciousness Studies: jcs-online@yahoogroups.com

Andy Ross [1] unashamedly endorses Kurzweil's vision, a vision which in turn unashamedly promotes the most mindless view of the capabilities of AI, conceived as the future replicator of human perception and cognition, ever espoused on the planet. Andy, JCS notes, has four degrees in philosophy. Some cognition and perception would help. Apparently, he has never considered that the hard problem and cognition may be inextricably linked. He states flatly, "... so in principle we shall be able to simulate, understand and reconstitute in hardware the higher cognitive functions of human beings."

Here is the problem with this statement and the rampant misconception it represents: A theory of cognition is ultimately dependent on a theory of perception. Without solving the problem of perception, i.e., the origin of the external image of the world, there is no solid basis for the theory of memory. This is, yes, currently dubbed the hard problem. Our perception (with all its structure of forms and space and its qualia) is simultaneously our experience. If you do not truly know what our experience is, how can you have a (grounded) theory of the storage of this experience, namely, memory. If you do not know how experience is stored, how can you have a theory of cognition, the essence of which is to use stored elements of experience in thought? Such a cognitive theory rests on symbols derived from this experience. With such lack of foundations, you cannot know what symbols actually are.

For this reason, a solution to the hard problem will, with near certainty, imply a new theory of memory and cognition. I have argued explicitly elsewhere [2] that this is so and why, indicating the form of this cognitive science and showing that the nature of the "device" necessary for perception and cognition supports a form of "broad computation" beyond that of machines of the Turing class (yes, just as Penrose argued), and with properties completely different from any robots Ross might envision.

So, when Ross glibly promotes the future results of a research program that simulates in detail the processing of a column of the cortex, thus generalizing in theory to the whole cortex, enabling us, "to build robots the equal of us in terms of raw intellect in the foreseeable future," there is another fallacy (among many). One can study a vacuum tube to the nth degree, this does not mean that this gives any insight into the overall purpose and functioning of the radio in which it is embedded. One requires some theory of the principle of global function of the device containing the part. As the best example I can offer, I've argued that, just as the purpose of the parts of an AC motor are for the generation of a very concrete field of electric force, so the purpose of the brain's apparently computational re-entrant architecture, the feedback loops, etc., is ultimately to support a very concrete reconstructive wave "passing through" an external, holographic matter-field, specifying the external image — ultimately a broader form of computation. But this model requires also a certain model of time. It is this global view on the function and purpose of the brain that is deemed by me necessary for the hard problem, and the point, in the case of Ross, is that local studies of parts of the cortex will nevertheless be restricted in their effect without a larger theory of the whole. Further, this larger model, when arrived at, may well involve (and does as far as I am concerned) attributes of the "device" that is the human being that are far beyond the robotics and AI envisioned by Ross.

Apologies, slightly, for the rant. I am, in my old age, tired of the mindless, unending AI myth, which I think, is far from innocent in its effects on this world.

References
1. Ross, Andy: Will robots see humans as dinosaurs?
2. Robbins, Stephen: Semantics, experience and time
 

Will Robots See Humans As Dinosaurs?

From: Michael E. Zimmerman
To: me@andyross.net
Sent: Monday, May 21, 2007 9:52 PM
Subject: Robots

Just read your "Will robots see humans as dinosaurs?" essay in JCS. This is the best short summary that I've read having to do with a (from the current dinosaur human perspective) dystopian vision of the techno-future. Recently, I finished Kurzweil's The singularity is near, which both inspired and frightened me simultaneously. Since then, I've been poring over dozens of books and essays dealing with cyborgs, nanotechnology, etc.

What's occurring, without any supervision and little public discussion, is mind-blowing.

As I finished your essay, I realized that comparing the totally outmoded-humans as dinosaurs doesn't quite work. We (anyway, lots of people!) respect dinosaurs; we like to read books and see films about them. Perhaps to our cyborg descendents we will seem like seriously deficient Untermenschen, or outright vermin, worthy of extinction/extermination. Such terms, of course, calls to mind mid-20th century extermination efforts, bound up with the kind of eugenics technology available at the time.

Reference
J. Andrew Ross: Will Robots See Humans As Dinosaurs?
Journal of Consciousness Studies 13(12), 97-104 (2006)

Michael E. Zimmerman is Director, Center for Humanities and the Arts,
and Professor of Philosophy at the University of Colorado at Boulder, as
well as author of three books and about a hundred papers on philosophy.


From: Andrew Ross
To: Michael E. Zimmerman
Sent: Wednesday, May 23, 2007 9:37 PM
Subject: Re: Robots

Thanks for your kind comments. In fact my main preoccupation there was how far continuity of culture-wide consciousness could be sustained in such a transition. It looks like a rupture analogous to that from dinosaurs to mammals, where a new way of life leaves precious little space for the old. Dinosaurs live on as birds, of course, but they can hardly be said to make much contribution to our culture, except as chicken nuggets! I used to think the Matrix scenario of humans as mere commodity fodder was too desperate, and the idea of fighting the machines as too romantically speciesist, but there won't be much room for people inside the cyborgs, and clearly only the fittest, in the sense of being adapted to a machine world, will survive.