HOUSE_OVERSIGHT_016859.jpg

THE UNITY OF INTELLIGENCE
Frank Wilczek

Frank Wilczek is Herman Feshbach Professor of Physics at MIT, recipient of the 2004 Nobel Prize in physics, and the author of A Beautiful Question: Finding Nature’s Deep Design.

I. A Simple Answer to Contentious Questions:
• Can an artificial intelligence be conscious?
• Can an artificial intelligence be creative?
• Can an artificial intelligence be evil?
Those questions are often posed today, both in popular media and in scientifically informed debates. But the discussions never seem to converge. Here I’ll begin by answering them as follows:
Based on physiological psychology, neurobiology, and physics, it would be very surprising if the answers were not Yes, Yes, and Yes. The reason is simple, yet profound: Evidence from those fields makes it overwhelmingly likely that there is no sharp divide between natural and artificial intelligence.
In his 1994 book of that title, the renowned biologist Francis Crick proposed an “astonishing hypothesis”: that mind emerges from matter. He famously claimed that mind, in all its aspects, is “no more than the behavior of a vast assembly of nerve cells and their associated molecules.”
The “astonishing hypothesis” is in fact the foundation of modern neuroscience. People try to understand how minds work by understanding how brains function; and they try to understand how brains function by studying how information is encoded in electrical and chemical signals, transformed by physical processes, and used to control behavior. In that scientific endeavor, they make no allowance for extraphysical behavior. So far, in thousands of exquisite experiments, that strategy has never failed. It has never proved necessary to allow for the influence of consciousness or creativity unmoored from brain activity to explain any observed fact of psychophysics or neurobiology. No one has ever stumbled upon a power of mind which is separate from conventional physical events in biological organisms. While there are many things we do not understand about brains, and about minds, the “astonishing hypothesis” has held intact.
If we broaden our view beyond neurobiology to consider the whole range of scientific experimentation, the case becomes still more compelling. In modern physics, the foci of interest are often extremely delicate phenomena. To investigate them, experimenters must take many precautions against contamination by “noise.” They often find it necessary to construct elaborate shielding against stray electric and magnetic fields; to compensate for tiny vibrations due to micro-earthquakes or passing cars; to work at extremely low temperatures and in high vacuum, and so forth. But there’s a notable exception: They have never found it necessary to make allowances for what people nearby (or, for that matter, far away) are thinking. No “thought waves,” separate from known physical processes yet capable of influencing physical events, seem to exist.
That conclusion, taken at face value, erases the distinction between natural and artificial intelligence. It implies that if we were to duplicate, or accurately simulate, the physical processes occurring in a brain—as, in principle, we can—and wire up its input

56

HOUSE_OVERSIGHT_016859