How we are not computers, and why it matters.
As we become ever-more aware of the shortcomings of the mechanical and Information Processing worldviews, and of the power of artificial intelligence networks, we need an intellectual framework for understanding, interpreting and strategically defending our own place in a universe of increasingly intelligent systems. I have one to offer.
“Might it be time to move beyond the computer as a metaphor for how the brain works? A group of scientists now say yes.
Unfortunately, I didn’t find the alternative all that understandable from this”
I think I can identify the alternative Rosenblatt seeks. And it matters. So here goes.
Criticizing the computer metaphor that dominates current thinking about intelligence, Epstein asserts (and I agree). ”We are organisms, not computers. Get over it. Let’s get on with the business of trying to understand ourselves, but without being encumbered by unnecessary intellectual baggage.” He goes on to suggest (and again I agree) that the correct alternative to the information processing metaphor is not to be found in deistic metaphors of a supernatural creators, Cartesian hydraulic metaphors, or Hobbesian mechanical metaphors.
But for three reasons, I do not agree when Epstein calls for a “metaphor-free” framework for intelligent human behavior. First, because the map will never be the territory, we will never have a metaphor-free framework for understanding. (Get over it.) Secondly (to paraphrase Churchill on buildings) “we shape our metaphors and then they shape us”. Our future is at stake.
And third, the “right” metaphor for our era is staring us in the face. I’ll introduce it by picking up on Epstein’s recommendation that we think of intelligent behavior, not as information processing and representation , but rather as “a direct interaction between organisms and their world.” This observation invites the “right” metaphor.
Biological evolution, like intelligent behavior, also results from “direct interaction between organisms and their world.” Species, like intelligent animals, also change adaptively through interaction with the world. So here’s the metaphor: Species are intelligent, and the processes responsible for species intelligence provide useful insights into the processes responsible for animal and human intelligence.
So let me see if if I can tempt you to pick up the trail with me (see “Are Species Intelligent”, 1992 and my articles on William James) by commenting on the three processes Epstein uses to illustrate how intelligent humans are changed by experience.
Of special note [he writes] are experiences of three types: (1) we observe what is happening around us (other people behaving, sounds of music, instructions directed at us, words on pages, images on screens); (2) we are exposed to the pairing of unimportant stimuli (such as sirens) with important stimuli (such as the appearance of police cars); (3) we are punished or rewarded for behaving in certain ways.
Let’s take these in reverse order. (3) Darwin’s theory of natural selection is the theory of how species are driven by “punishment and reward” into more adaptive states. If we avoid the temptation to say this is “merely” a matter of differential reproduction, selection and mutation, we are in a good position to recognize the richness of biological evolution both as a phenomenon, and as a model of intelligence and social change. (Dan Dennett brilliantly championed this idea in “Darwin’s Dangerous Idea” and in a recent recent talk here.)
(2) Associative learning of paired stimuli is not usually described as a mechanism of evolution, but it occurs every time a species evolves an adaptive response to a stimulus (such as shortening day-length) which is incidentally associated with a survival-relevant circumstance (the impending winter which makes preparation for hibernation adaptive.).
And what about (1) observational learning? Well, perhaps it is a form of intelligence not be found in species. or perhaps our understanding of biological evolution is too reductive to allow us to see the evolutionary analog: consider Horizontal Gene Transfer, in which a species instantly acquires adaptations that took the originating species eons to achieve through direct interaction with its environment.
These metaphors — intelligence is like evolution; species are like intelligent organisms — are useful because they help us think about processes of adaptation at a level of abstraction that liberates us from the specificities of neural and computer intelligences, because they help us understand and develop those intelligences, and they are useful because they help us see ourselves as parts and participants in a larger evolving system that is more wonderful and potentially more simpatico than one might otherwise think. Which is why this really matters today.
We are at a crossroads of human evolution.
The conceptual models that created our techno-social world have brought us to a moment when the very character of humanity and the future is at stake. Those models have undeniable power. They now threaten to turn us into minor players and mere cogs in an inhuman socio-economic information processing machine driven by the power and affordances of mechanistic robotic techno-economic theories.
It need not be so. If we can see ourselves — accurately — as parts and participants of an organic rather than machine-like world, we can better shape a future that is responsive to the needs and potential of purposive intelligent compassionate humans.