The Cul-de-Sac of the Computational Metaphor
edge.org
edge.org
So perhaps much of the problem is that we're too often using a metaphor of non-adaptive computation. If instead we analogize to forms like simulated annealing, genetic algorithms, neural networks and so forth, the computational metaphor becomes less cul-de-sac and more throughway.
all are attempts to simulate processes in nature computationally
> If instead we analogize to forms like simulated annealing, genetic algorithms, neural networks and so forth
Perhaps I'm missing something, but I see all of these things as well within the boundaries of computation!
But, from a broader perspective, I don't think the "is this computation" question is a very useful one to pursue. Any true understanding of the conscious mind will have to answer the question of why it is different from the things we understand now: if it is, in fact, computation, then it is different in some way from the computations that we do understand now, and it is the differences that will be key to understanding it. When we understand it, we will have a informed basis for fitting it into our ontology.
Look at this list of participants: Daniel Hillis, Neil Gershenfeld, Frank Wilczek, David Chalmers, Robert Axelrod, Tom Griffiths, Caroline Jones, Peter Galison, Alison Gopnik, George Dyson, Freeman Dyson, Seth Lloyd, Rod Brooks, Stephen Wolfram, Ian McEwan.
The responses themselves often don't live up to expectations, though some do.
See, near the end, the reply by Alison Gopnik.