Thinking has always been an ethereal thing. It is the most private of human activities, and while the expression “I know what you’re thinking” is part of the lingua culturae, it is a bold-faced lie. Notwithstanding crypto-keys and blockchains, the only truly protected storage place in the universe – at least for now – is the thought-swarm deep within our respective skulls.
Brains themselves are fascinating stuff, and no one is quite sure what’s in there. This is important, since a vessel of chemicals and tissue generating a cluster of electrical activity is one thing, but that which we call consciousness is quite another. Physicist/Neuroscientist Paul Nunez wrestled with this, building a foundation of scientific facts with which to construct some reasonable theories. Science also warns us, he notes, that some things are fundamentally unknowable.
Philosophers, who spend considerable time thinking about thinking, have had a lot to say about the subject. One of the most influential philosophical works, Discourse on Method (1637) by René Descartes, not only gave us the Cartesian coordinates that we scientists have come to love, but also introduced the discipline of Methodic Doubt. Doubting things methodically led René to the conclusion that he could not doubt his own existence since after all he was the one doing the doubting. The coup de grâce for self-doubters was the legendary “Cogito, Ergo Sum”, aka “I Think Therefore I Am.”
Psychologists make a living trying to figure out what the human mind is up to, and how that relates to the stuff we experience in the material world. They know for a fact that if a pink elephant were to walk into a room with a test subject, that person would be thinking about a pink elephant. They are also pretty sure that telling that subject not to think about a pink elephant will achieve the same result. Thanks to the philosophers, anyone who doubts this can at least rest assured that they themselves exist.
It is against this backdrop that scientists are also trying to figure out if machines can think like humans. Long before IBM’s Watson started beating humans at chess and Jeopardy, people were predicting that machines would eventually think. Alan Turing saw this coming back in 1950, before computers were a thing, and developed the simple test which bears his name. Avoiding the somewhat murky details of human thought, Alan decided that if a researcher could not tell the difference in responses between a machine and a human, then the machine will have passed his test. After nearly 7 decades of thinking about it, the Turing test may still be the best we can do.
A recent entry in this thinking competition is Duplex, a conversational AI that made its debut at Google’s yearly developer conference. The demonstration involved the very Turing-like task of calling and making a reservation at a restaurant where they didn’t know they were talking with a computer.
After the usual back and forth about times, dates and number of people dining, the restaurant staff settled on the details. Given the issues with pronunciations, response times and cadence that plague most digital assistants, this is quite remarkable. Google’s CEO Sundai Pichai noted that the breakthrough moment for him was when, during the conversation, the computer said “Umm.” Although that is admittedly very human-like, my personal epiphany was the realization that a machine might soon be able to call and negotiate with my cable provider.
Machines will certainly continue to challenge the Turing test, and it will get more difficult to distinguish them from humans in many settings. Centuries before Alan Turing was born, Descartes recognized that “automata” might someday show human-like behavior, but predicted they would never respond any better than “…even the lowest type of man…” Automata have come a long way since then, but the idea that IBM’s Watson might someday comprehend “Cogito, ergo sum” remains unthinkable.
“I don't really worry about machines thinking like people; I worry about people thinking like machines.” – Tim Cook, Apple CEO
Author Profile - Paul W. Smith, a Founder and Director of Engineering with INVENtPM LLC, has more than 35 years of experience in research and advanced product development.
Prior to founding INVENtPM, Dr. Smith spent 10 years with Seagate Technology in Longmont, Colorado. At Seagate, he was primarily responsible for evaluating new data storage technologies under development throughout the company, and utilizing six-sigma processes to stage them for implementation in early engineering models. He is a former Adjunct Professor of Mechanical Engineering at the Colorado School of Mines, and currently manages the website “Technology for the Journey”.
Paul holds a doctorate in Applied Mechanics from the California Institute of Technology, as well as Bachelor’s and Master’s Degrees in Mechanical Engineering from the University of California, Santa Barbara.