Envatment – Brains and Vats

Perhaps you have heard of the brain-in-a-vat thought experiment. Such fanciful “intuition pumps” are supposed to tell us something about the mind, knowledge, and reality, but how much can we learn from such a mad science fairytale? According to Evan Thompson and Diego Cosmelli, not much. Thompson and Cosmelli address the thought experiment head-on, and in a lively manner, tackle just what it would mean to actually be a “brain in a vat”. They come up with a surprising answer:

Any truly functional “vat” would need to be a surrogate body subject to control by the brain. By “body” we mean a selfregulating system comprising its own internal, homeodynamic processes and capable of sensorimotor coupling with the outside world. In short, the so-called vat would be no vat at all, but rather some kind of autonomous embodied agent.

This supposition has an important implication. It implies that our default assumption should be that the biological requirements for subjective experience are not particular brain regions or areas as such, but rather some crucial set of integrated neural-somatic systems capable of autonomous functioning. This assumption is one of the core working assumptions of the enactive approach.

If you are interested in learning more about the enactive approach, check out Embodiment and Philosophy of Mind by Andy Clark.

Perception versus Representation

The world is it’s own best model…why? We can put the answer in another slogan that [AI researcher] Brooks would probably like: Perception is cheap, representation expensive. Such a slogan might surprise many AI workers, who are acutely aware of how difficult pattern recognition can be. But the point is that good enough perception is cheaper than good enough representation-where that means “good enough” to avoid serious errors. The trouble with representation is that, to be good enough, it must be relatively complete and relatively up to date, both of which are costly in a dynamic environment. Perception, by contrast, can remain happily ad hoc, dealing with concrete questions only as they arise. To take a homely example, it would be silly, for most purposes, to try and keep track of what shelf everything in the refrigerator is currently on, if and when you want something, just look.

This quotation is from chapter nine of John Haugeland’s essays in the metaphysics of the mind, Having Thought. In the chapter, Haugeland argues against the “interrelationist” account of the mind and the body, which accepts the premise:

that the mental, or at any rate the cognitive”, has some essential feature, such as intentionality or normativity, and then argue that this feature is impossible except through participation in some supra-individual network of relations.
…
interrelationist arguments are holistic in the specific sense that they take cognitive phenomena to be members of some class phenomena, each of which has its relevant character only by virtue of its determinate relations to the others-that relevant character being, in effect, nothing other than its “place” in the larger pattern or whole.

As a competing theoretical framework, Haugeland offer’s what he calls the “intimacy of the mind’s embodiment and embeddedness in the world…[with the] term “intimacy” suggesting more than just necessary interrelation or interdependence but a kind of commingling or integralness.”

This embodiment relates to my earlier entry on the MIT Cog project, which is following the general thesis outlined by Haugeland, which is that intelligence does not depend on a “furniture of information” or “complex symbol structure that are, in many respects, just like the contents of the traditional Cartesian mind.”, but rather, on the “concrete details of the agent’s embodiment and worldly situation”.

This embodied and embedded approach to perception has probably been most controversially put forward by psychologist James J. Gibson. Haugeland quotes him as follows:

the words animal and environment make an inseparable pair. Each term implies the other. No animal could exist without an environment surrounding it. Equally, although not so obvious, an environment implies an animal(or at least an organism) to be surrounded. This means that the surface of the earth, millions of years before life developed on it, was not an environment, properly speaking.

Thus, in this sense, Haugeland says that we can understand animals(including humans) as “perceivers” if we consider than “inseparably related to an environment, which is itself understood in terms appropriate to that animal”.

So for humans, in order to understand perception, we must understand our complexly relative relationship to the environment, which necessarily implies our dynamic socialization schemas. Furthermore, Gibson claims that in order to understand this perception of the environment, we must take into account the affordances of the environment, or what it offers to the animal. Thus, “the central question for the theory of affordances is not whether they exist and are real but whether information is available in ambient light for perceiving them”.

This perception of affordances puts the first quotation of this post in context. The world is it’s own best model and our perception of it is in relative(embodied and embedded) terms with the environment.

Minds and machines

In this post, I want to briefly overview MIT’s exciting Cog project

Simply stated, Cog is “a set of sensors and actuators which tries to approximate the sensory and motor dynamics of a human body.” So what is the point of trying to replicate the “sensory and motor dynamics” of humans? Basically, the Cog researchers are trying to create an Artificial Intelligence.

In order to understand why an AI seemingly must have a humanoid body in order to be intelligent, one must have a basic understanding of embodiment theory.

The main thesis behind embodiment theory can be found in Shaun Gallagher’s How the Body Shapes the Mind. In this seminal work, Gallagher precisely defines the vocabulary necessary to talk about the thesis stated in the title: how the body shapes and influences the mind. Another overview of the embodiment thesis can be found here, by important embodiment researcher Andy Clark.

So, what does all this have to do with Cog and artificial intelligence? The MIT webpage has a nice overview and states “If we are to build a robot with human like intelligence then it must have a human like body in order to be able to develop similar sorts of representations.” Thus, the morphological(form) as well as the functional characteristics of our body-brain system play a critical role in shaping the dynamics of intelligent human interaction with the environment. The Cog project is not trying to “simulate” human intelligence on a symbolic level,which has been the traditional approach of Good Old-Fashioned Artificial Intelligence(GOFAI) but rather, is attempting to get human-level cognition to emerge from an intermodal and dynamic interaction with the environment.

Justification for the importance of Cog’s humanoid facial features is the fact that social interaction is perhaps the most important facet of human-environment-reciprocity that makes human intelligence uniquely human relative to the other great apes. It is the early prenatal and postnatal social-learning and development that gives rises to important conceptual constructs such as relativity(self/other, inside/outside, etc). If you are interested a neurological discussion of how such concepts arise from our embodiment, see my paper Mirror Neurons and the self

If this brief discussion of Cog as piqued your interest, you will probably be interested in some of MIT’s video overviews.

Lastly, I will end this post with another quote from the MIT page:

In any case….it turns out to be easier to build real robots than to simulate complex intereactions with the world, including perception and motor control. Leaving those things out would deprive us of key insights into the nature of human intelligence.