Artificial Intelligence
Let me begin with an apology for a certain amount of terminological awkwardness. I have settled on the term "intentionality" and have been pondering the notion of a "governing intentionality" for most of my (largely wasted) adult life. There are a number of near synonyms that I could have chosen -- mission, purpose, end -- but each comes freighted with its own connotations, most of which I want to avoid for reasons that will become apparent.
Rather than provide a formal definition of intentionality, let me just say that I first began pondering intentionality in the context of artificial intelligence. The anecdote I use stems from the mid-eighties when I was in graduate school, specifically when I was in a computational linguistics course. I took on, as a class project, the programming of a computer to produce sentences, and then, just to take it to the next level, gave it a vocabulary and syntax so the cumulative sentences would resemble a poem by the poet John Ashbery. It all sounds much more impressive than it was, but the poems that the machine produced were close enough to being poems that they were able to fool some of my fellow graduate students in the English department -- that is to say, some took them to be poems by a person -- which is to say they passed the infamous "Turing test." In order to read a poem as a poem, one must ascribe authorial intent to the text. One must assume, in other words, that the author intended those particular words and that he meant something by them. Those who were fooled by the poems did just that and some had quite ingenious "readings" that ascribed intent and meaning well beyond anything the computer had in mind.
I bring this forward to talk about the disconnects, the first of which is the disconnect between the intentionality governing production of the poem and the intentionality ascribed to the poem itself. At one remove, I created a computer program. While I was quite proud of the body of the program, it did what most would consider a simple and useless task. It chugged from an initial state, a sort of "go-signal," to a final state, "a printed poem." The program "knew" that it had achieved its final state when it had fulfilled certain criteria specified in advance in the body of the program. I have strong doubts that the computer had "intention," in the way that we would normally think of "intention" as a state of mind, but I cannot be sure. I had no doubt, however, that the computer had "intentionality," that it worked through more or less random iterations until it satisfied the criteria of the final state and a produced what was recognizably a poem. In a certain social context, where there were writers and readers of poems, the readers struggled through the "surrealist" combinations until they too had satisfied the criteria and had produced what might be called a "reading." To use Wittgenstein's term, both the computer and the reader engaged in separate, but connected "language games." The outcome of neither game was entirely predictable, at least not to me, but the game itself and the outcome of the game, the satisfaction of the criteria, was recognizable.
The Hunger Game
Let me put this same discussion in another context, one that most would recognize as wholly physiological -- hunger. Most would consider hunger and the need to satisfy hunger a basic need, not only for humans, but for most living things. For those who are reading this, it sits right down there at the bottom of Maslow's hierarchy of needs and it occupies a good deal of our time and effort throughout any given day. I don't want to make too much or too little of this, but most of us would consider hunger mechanistic -- that is to say, it is built into the mechanism of our physiology. We feel hungry, and when we feel hungry, we go about the business of getting something to eat, something that will satisfy our hunger. To use the language of the previous section, our bodies chugged from an initial state, a sort of "go-signal" of hunger, to a final state, a "satisfied hunger." Here, however, I have no doubts whatsoever that I was aware of my hunger, that I had "intention" to satisfy my hunger, and that I took certain actions that I knew would suffice to do so -- I went to refrigerator in search of left overs that I could heat up in the microwave, found them, and ate them. I went through the motions of what most would recognize as a "hunger game," and if my actions were being watched, some hidden spy camera in the light fixtures, my actions would signal (or to use a more loaded term, would signify) my "intention" to get something to eat. Likewise, though, I have no doubt that I had "intentionality" -- that the physiological "go signal" of my hunger was prior to my "intention" to get something to eat and that the instrumental acts associated with getting something to eat were governed by that physiological state. The final state of "a satisfied hunger" is itself a given -- that is to say, I did not choose to be hungry, and while I could choose to resist the impulse or attempt to ignore the stomach pangs, it is there nevertheless -- but the final state of a satisfied hunger does not determine the acts sufficient to its satisfaction. One can imagine many ways of going about "getting something to eat." Instead of opening the refrigerator, I could have grabbed the car keys and popped on down to Burger King. Nevertheless, implicit to my physiological being are certain criteria that the instrumental acts must meet in order to bring about "a satisfied hunger" -- certain acts do (and many more potential acts do not) contribute to the final state of a satisfied hunger.
There are a couple of things I want to point out here. First, the whole business of intentionality, as I am describing it above, can be rather uncanny once we recognize it for what it is. Consider, for example, the sci-fi staple, the almost human android. I am imagining the Star Trek character of Data, but I could be speaking of the more sophisticated android like creatures that populated Battlestar Gallactica. Imagine, in other words, that everyone around you is simply a machine, cleverly designed by an advanced alien race to replicate the actions of human beings in every possible way conceivable. Unlike you, they have no "inner life." They do not have "states of mind," but in the limited case here, only the intermittent "go-signal" that sets in motion a "hunger game," which they play to completion, and then go about other "games." Ultimately, it's not terribly difficult to imagine, in part because we suspect that we are indeed machines at some level, and in part because we do not have the sort of direct access to the inner lives of others that we have to our own. We must "surmise" it from actions that we recognize as instrumental to certain intentionalites. Even if we don't recognize the substance, placing it in the mouth, masticating it, swallowing it, all signal someone who is attempting to reach the final state of a satisfied hunger. We must "surmise," as an interpretive act, that the "eater" felt hungry and has the intention of satisfying that hunger. I will return to this point later, but it is important to keep it in the back of one's mind because it has moral and ethical implications. In certain respects, by asking you to imagine a world in which others are simply machines (and by implication, sub-human) I have also asked you to inhabit the world of a socio-path, one who fails to "surmise" the inner lives of others.
Second, as a sort of first noble truth, I am suggesting that intentionality is suffering. We have labeled it differently -- hunger as need, hunger as desire, et cetera -- but at fundament hunger is a state of discomfiture. Having said that, "suffering" overstates the case, "discomfiture" understates the case, but it serves to make another point. I strongly suspect that I have not felt hunger in quite the way that others around the globe have felt hunger. I am more to the "discomfiture" end of the spectrum than the "suffering" end of the spectrum, and the instrumental acts sufficient to its satisfaction are more readily available to me, in wider variety than for many others. We might split a hair and suggest that others play the "hunger game" while I play something more akin to an "appetite game," and such distinctions may prove to be important, but for the moment I simply want to make the point that the game comes as an a priori and a governing imperative. I must do something, and insofar as it must satisfy certain criteria or conditions set out within my physiology, I must eat. I must, in other words, free myself from hunger, and I am more or less free to engage in instrumental acts pursuant to that end. I say "more or less free" because clearly some things can be eaten, other things cannot, and among those things that "can" be eaten from a purely utilitarian perspective, we may choose not to eat. While I have no particular objections to eating insects, and I recognize that they could be a source of protein, I choose not to do so, in part because my hunger is more "discomfiture" than "suffering," in part because it is simply not what "we" do. If my wife were to observe me scrounging the back yard for bugs to eat, she might "surmise" that I was hungry, but she might also "surmise" that I had gone off he deep end.
The Chess Game
Let me put this initial discussion of intentionality into one last context. Current novels and movies aside, most do not think of hunger as a game -- not, for example, in the way we think of chess as a game. Here again, let me point out the basic structure of what I'm describing as intentionality. In any chess game, there is an initial state of affairs in the arrangement of the pieces on the board, a go signal when one of the two players moves the first pawn, a number of intervening moves each of which is intended to achieve a final state of checkmate. There are, of course, profound differences between the hunger game and the chess game. The one is for the most part biologically constructed, the other is a social construct from top to bottom. I use the term "socially construct" rather guardedly, but purposefully on a couple of levels.
At one level, there is the game itself. While the game has an existence outside and before any particular player, and in that sense is a given, but the game itself is wholly constituted within and by its "rules." This would seem to have rather profound implications for the governing imperative. Because the hunger game is implicit to our biological being, we do not choose either the initiating state or the final state, and because it is implicit to our biological being, because it comes from within our very being, so to speak, it would seem to carry a different weight than the governing imperative of chess. We suffer under a failure to satisfy our hunger in ways that we do not suffer under a similar failure to achieve checkmate, and somehow this difference must be taken into account. Perhaps so, and they are very different games, but for the moment, I am merely suggesting a structural similarity. Because the chess game has been defined within our social environment, we do not choose either the initiating state or the final state, and in that sense the chess game, no less than the hunger game, presents us with an a priori and a governing imperative. Once I have chosen to enter into the game, I must achieve checkmate. I must, in other words, free myself from the "not-yet" (as in "not-yet-checkmate") and I am more or less free to engage in acts pursuant to that end. Here again, I say "more or less" because some acts are efficacious to that end, other acts are not, and among those acts that might potentially be efficacious (if only I could move my knight just a bit differently I would have him) are those that are prohibited under the rules of the engagement.
At another level, there is the competition between the players. I could have chosen solitaire as my example, and avoided this discussion for the moment, but the chess game allows me to make a number of additional preliminary points. If we imagine a chess game in progress, the two players are sitting slumped over the board, and one makes a move. Both players are now attempting to reach a state of checkmate, both have entered into a state of "not-yet-checkmate" and the white player's move has now set the context for the black player's move. He makes a move, and his move sets the context for the white player's next move, and so on. I am suggesting that there is an evolving contingency within which each of the subsequent moves take place. The white player is free to make his moves, the black player is free to make her moves, but each move changes the contingency within which the next move takes place -- changes, that is, the potential efficacy of any individual move. The while player may have had a strategy, but the black player's move just threw it out the window, and a move that is possible -- that is permitted under the rules -- no longer has pragmatic efficacy. The black player is, in effect, still free to make the move, but insofar as the white player's last move precludes its effectiveness, insofar as it no longer contributes to his freedom from "not-yet checkmate," one might say there is a difference between a theoretical freedom to act and a pragmatic freedom to act. I am simply making the common sense observation that the imperative remains the same, but what is possible is not always practical, and it is good to keep the distinction in mind.
The Institutional
Ok, at one level there is the game itself, and it is socially constituted within the mutually understood rules of the game, and then there is the playing of the game. Some will recognize this as de Saussure's structuralist distinction between the synchronic and the diachronic -- between the grammar of a language and the actual production of contingent utterances within that language -- between the rules of chess and the actual production of a contingent game within those rules -- and so on. I am less interested in the temporal argument per se (though clearly, intentionality, as I have described it, exists within the temporal frame of a "now" fraught with the "not-yet") somewhat more interested in what might be called the "institutional" argument. It is the admittedly structuralist distinction between an institution, which we generally understand to have a synchronic existence and the individuals who participate within the institution, which we generally understand to have a contingent diachronic existence. In general parlance, we hear this referred to as "being part of something larger than ourselves," and I am suggesting that the a priori and governing imperative -- that intentionality -- brings into being an institution.
The choice of language is not altogether gratuitous. Within political and moral philosophy, it is rather commonplace to imagine points of origin, so I will engage a version of that particular thought experiment and allow the uncanny to peek through momentarily again. Imagine, if you will, that technology has progressed to the point that we can replicate exactly existing human beings, with one exception -- they have no memory, no contingent experience, prior to the moment of being "activated." Imagine further that a group of human beings have been placed on a deserted island, which we will call Eden to capture its paradisiacal nature. At the moment of their activation, they have he inherent biologically constituted imperatives, but no others, nothing that would resemble a "socially constituted imperative." They are merely an aggregate of "individuals," in what might be called a "state of nature," and I will refer to this aggregate as the "polis" for lack of a better term. It is clear enough that the polis lacks a collective purpose, and whether the polis devolves into a Hobbesian war of all on all, or evolves into a utilitarian commune, depends upon any number of contingent factors, but is also clear enough that those in the polis will create institutions which do have a collective purpose. If one imagines a Hobbesian war of all on all, it will not be long before one has something that resembles a police force. If one imagines a utilitarian commune, it will not be long before one has something that resembles a redistributive economy, moving goods (e.g. food) from those with a surfeit to those in need. The institution as such may have an historical point of origin, but once created, once given an a priori and governing intentionality, the institution takes on a life of its own -- to use Wittgenstein's phrase, it becomes a "form of life."
I am also suggesting that some of the same distinctions pertain as we speak. The individual, as an individual, is brought into being by intentionality. He or she or it must engage instrumentally with the world to satisfy what has been given as an intentionality. I am, as it were, subject to the subjectivity of my given imperatives, you are subject to the subjectivity of your imperatives, et cetera. We can look at the city of New York or Los Angeles and see nothing but an aggregate of individuals, each engaged instrumentally with the world to satisfy their individual imperatives. The institution, as an institution, is brought into being by intentionality. It must engage instrumentally with the world to satisfy what has been given as an intentionality. We can look at the City of New York or Los Angeles and see a polis proper, with institutions, each engaged instrumentally with the world to satisfy their individual imperatives. Insofar as individuals participate within institutions, within socially constituted intentionalities, I am subject to the subjectivity of those institutions within which I participate, you are subject to the subjectivity of those institutions within which you participate, et cetera. Here again my language is not gratuitous. I use the phrase "brought into being" and "subject to the subjectivity" rather deliberately, and I return to them in my next set of posts, but let me close out this post with an observation. Just as I don't know if my poem writing computer program had an intention to write poems, a consciousness or a subjectivity within the wiring of the machine, I do not know if an institution has an intention to satisfy its intentionality, a consciousness or subjectivity that transcends the consciousness or subjectivity of the individuals that participate within it.
No comments:
Post a Comment