It's Logical: Computer Is Not a Brain


July 15, 1992|By GERALD M. EDELMAN

LA JOLLA, CALIFORNIA — La Jolla,California--We may soon see a headline, ''Machine Beats World Champion Chess Player.'' Indeed, it is likely that we will be able to design a computer to carry off this feat. Does such a machine have a mind of its own? Is our brain just a machine in a machine-like world?

Some scientists have suggested, in fact, that the brain is like a computer. Others have concluded that computers can think.

If all this makes some humans uneasy, there is comforting news. Brain science and modern physics both point to the conclusion that neither the universe nor the human brain are deterministic machines. Although the computer is the most significant invention of the 20th century, it is not a thinking object comparable to our own brains.

Instead, computers are powerful logic engines that operate under precise instructions. They do not have bodies, are not conscious and cannot function without an explicit program.

In contrast, brains are not primarily logic engines. They are structures that have evolved to deal with novel events. Given how many things your brain must deal with while you drive home from work, it is not surprising that the brain is the most complicated object in the universe.

If you counted at a rate of one per second all of the connections in the part of your brain responsible for consciousness, you would finish counting 31 million years from now.

The finest wiring of the brain is individual; no two brains are alike. Brains allow their animal owners to categorize and act on events that cannot be foreseen. Computers with this much individual variation could not even run their programs. A brain memory, unlike the memory in a computer, is creative, variable and dependent on context.

Computers, being non-conscious, cannot develop a language with changeable meanings that refer to things or events. Programmers assign the meanings before and after they run their programs. Meanings developed in and by brains, by contrast, have poetic capabilities.

They depend on the bodies of which they are part and on the circumstances these bodies encounter. It is unlikely that a computer-waitress would say to another computer-waitress, ''The ham sandwich just left without paying.''

A rich brain works like a cross between a jungle and a map. The variations in a brain's structure and function help determine which neural circuits are most fit. Just as individual animals are favored in Darwin's theory of natural selection, so individual groups of neural cells in a brain are enhanced in their connectivity if their activity results in rewarding behavior.

Maps between these selected cell groups lead to rich responses, transforming past signals in novel ways. Solid evidence supporting these ideas is now beginning to accumulate.

These observations have deep implications for human concerns, particularly if each brain is necessarily individual and if each of its responses is part of a historical reaction to novelty.

If these notions turn out to be true, then every act of perception is in part an act of creation. Every act of memory is in part an act of imagination. These are not acts to which computers can aspire.

Although neuroscientists like myself depend heavily on the power of computers to help us model the brain, the computers themselves are poor models for what is going on in our heads and bodies. Used well, they can help us find out about human nature and its bases. That is a sufficiently valuable role without exaggerating their capabilities.

Together with the avalanche of discoveries in neuroscience, the proper use of computers will provide an exciting view of what it truly means to be human.

The Enlightenment of the 17th and 18th centuries had a hard time reconciling human freedom with a machines-like model of the world. Now we are recognizing that, far from operating like machines, individuals have identities that are profoundly unique. revealing this, brain science is laying the ground for a new Enlightenment.

President Bush has declared the 1990s to be the Decade of the Brain. Properly supported, it may be expected to yield discoveries that justify a century of freedom.

Gerald M. Edelman, M.D., a Nobel Prize winner, is chair of the neurobiology department at Scripps Research Institute. This article is adapted from his new book, "Bright Air, Brilliant Fire."

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.