[This post was written by A. Michael Noll and distributed here with the permission of the author.]

It is nothing new. For decades, we have been fascinated with digital computing machines – thinking machines. For decades we have been using human terms to characterize them – intelligent. Artificial intelligence (AI) is over a half-century old. There is nothing new – just faster and more powerful machines with more complex programs controlling them.

It has always been very difficult for us humans to understand what a computer program does precisely by examining the complex and lengthy code. The code of an AI machine is impossibly complex, with code to even change the code. It would have been impossibly challenging for a human to look at the code I wrote decades ago to create computer art and know precisely what it did in fine detail. The code today is many orders of magnitude more complex.

The only ultimate “protection” from machines that are programmed to alter their programs is to require that a human always supervise any ultimate decision. I suggested this in my first published article in 1961. But then how does the program determine that it is really a human supervising it?

IBM’s Watson AI machine answered queries with astonishing speed – and accuracy. It promised to revolutionize healthcare diagnostics. Machines that played chess routinely beat the best human chess players.

Machines that respond – understand – human speech and “speak” back – appear in a number of science fiction movies from decades ago. Some of these AI machines went berserk, threatening and actually killing humans, such as in “Forbidden Plant” and HAL in “2001: A Space Odyssey.” Some of these fictitious machines promoted good behavior by humans, such as the one in “The Day The Earth Stood Still” – although the robot from the spacecraft killed humans with ease for the greater good.

Many decades of research has been done in such areas as natural language processing, computational linguistics, and speech recognition and synthesis. But they do not have the mysticism of resurrecting the decades-old term artificial intelligence – with all its nebulous implications – to describe it all.

All these claims about artificial intelligence, thinking machines, and comparison to the human brain were being made back in the 1960s. John R. Pierce (the “father” of Telstar) was so upset by what he considered nonsense that he contrasted the artificial intelligence of machines with the natural stupidity of humans.

Why are we fascinated by machines that seem to mimic us? Why do we apply human terms – like intelligence and thinking – to them? Does the AI machine have a spirit – a soul? And what “God” does it worship?

A. Michael Noll, Ph.D. — April 15, 2026 —  © Copyright 2026 AMN

Dr. A. Michael Noll worked in speech research at Bell Telephone Laboratories, Inc. during the 1960s, and programmed some of the earliest generative computer art. He retired from the Annenberg School at the University of Southern California where he was its former dean. He has written many opinion pieces debunking hype.