Speech #10: John Henry (continued)
Can machines take over from mankind? This has been a popular theme in science fiction; think of the HAL 9000 computer in "2001: a Space Odyssey", "The Stepford Wives" or "Blade Runner". The word "robot" was actually coined in a science fiction story written before I was born. But it goes back even further. There is a story from the Yiddish, mostly used to scare kids, about a "golem", that Shelley borrowed from extensively when writing Frankenstein. Clearly this is a deep-rooted fear, not really in our conscious, so let see how it stands up to the light of day.
Machines can be more powerful than muscles, but more powerful isn't always what is desired. A too powerful diamond cutter makes diamond dust, not a stone for an engagement ring. You need some strength to tune a guitar, but if you twist too hard, you will either move directly from flat to sharp or even break the string. For most tasks, what is required is strength and control or finesse. If our strength is lacking, then there is usually a tool available to multiply it. Tools that apply control are rarer and more expensive, and often not as good as humans. Examples of some of the more successful ones are thermostats, anti-lock brakes, and automatic pilots.
Fine control is necessary, but where and how should it be applied? This is the realm of thought. How good are machines at thinking? Thought has many attributes, but a few are pattern recognition, calculation, and judgement. Camcorders can see and hear just fine, but hook them up to most powerful computers and they are still poor at voice recognition and next to hopeless at figuring out what they are seeing. We are not yet to the John Henry point of machines approaching man's abilities in pattern recognition, except for some especially standardized tasks like checking out of grocery store. Even there, the point was to make the process easier for the barely trained, so they could perform more like the most expert cashier, rather than to try have a machine beat the human. The muskets used in warfare prior to 1750 were actually inferior weapons to the longbows that won the battle of Agincourt in 1415, but those archers were highly trained, and muskets could be used by almost anyone. The improvements that have come since then are due to the pressures of competition, whether for your grocery shopping dollar, or the more deadly kind on the battlefield, not simply to be better than a man.
Calculators perform calculations better than all of us and computers do them blindingly fast. In generalized judgement, computers have much further to go that in pattern recognition. In a very narrow field, if it is governed by a clear set of definable rules, then an immense investment can beat the best human. IBM's "Deep Blue" chess program finally beat world champion Gary Kasparov in a match. But the real world is not often like chess. Humans may not be the best at any one thing, but we are far more adaptable than machines. In one package we are mobile, discerning, knowledgeable, and pro-active. We can get to a problem or opportunity, diagnose the situation, figure out what needs to be done and apply the right tools. Sure, machines will keep getting better, because we make them so. If you suggest an improvement to a process, then that we includes you. If not, then the follow the example of the thirsty cave dweller, and learn how to use the new tools that come your way. Embrace, or least investigate, new technology, not because it is new, but because it can multiply your capabilities. Remember! a journey of a thousand miles begins with a single step. If you find learning how program in Java too daunting, then create your own web page instead. You will learn something along the way, thereby investing in yourself. And perhaps that will help you innovate into your next, more valuable role.