BRAIN-COMPUTER ANALOGY
| Collection | International Encyclopedia of Systems and Cybernetics |
|---|---|
| Year | 2004 |
| Vol. (num.) | 2(1) |
| ID | ◀ 334 ▶ |
| Object type | Epistemology, ontology or semantics |
It could be said that the superficial analogy between brain and computer which filled the pages of so many popular magazines has considerably obscured the issue of comparison of artificial as compared to natural intelligence.
Only the introduction of some “deus ex machina” would permit us to sustain such superficial analogy as for example in this statement by N.A. COULTER Jr. “Intelligence may be defined as the set of software programs which enable the Supervisor (or Ego) (sic) to acquire, create, and execute application programs” (!) (1976, p.36).
Curiously enough COULTER himself adds: “… the self-determinism of brains is one fundamental difference between brains and computers”… but he immediately relapses again into the idea that these self-determined brains depend on “the consent of its Supervisor — Ego” to be “programmed” Where could such a “Supervisor” be located?
More recently, M. BODEN makes, from a different standpoint, the following criticism of the analogy: “… current neural-network models, for all their likeness to the brain, are significantly unlike brains too. For instance, nearly all involve two-way connections, whereas brain-cells send messages in one direction only. Any one unit is directly connected to only a few others, whereas the lacy branches of a given neuron usually abut to many hundred of cells. Computer models contain no analogue of the neurochemicals that diffuse widely through the brain. Further, neuro-scientists still know very little in detail about what computations are carried-out by brain-cells, and how” (1990, p.131). BODEN admits however that: “Certainly, the brain is a connectionist system” (Ibid).
Somehow, the brain is probably working simultaneously in parallel and sequentially, a “trick” that computers have just started to learn.