While many people despise the subject, the study of mathematics can give you tools that last a lifetime. Solving any mathematical question requires that we precisely determine the question, the answer required and the information needed to solve the problem. These are valuable skills that can easily be applied to virtually all facets of life.
It is the issue of determining what answer is required that got me thinking recently about science, scientists and, more specifically, the development of artificial intelligence (AI). Across the world, thousands of individuals and hundreds of agencies and laboratories are focused on developing true AI. The potential benefits, and potential rewards, for developing true, human-like intelligence are massive. Robots and other machines could take on the heavy and/or dangerous work currently done by humans. Fighting wars, space exploration, studying volcanoes — these and a thousand other professions could be given to machines with true AI.
But as I applied the rules of mathematics to my thoughts about AI, I realized scientists have erred in determining the answer required. Any machine interacting with human beings and eventually replacing these same individuals must have the ability to understand our methods of thinking, our strategies for solving problems and the overall complexity of human communication. Yet any robot with AI that attempts to work with humans is going to fail, and fail for the simple reason that scientists should have been developing Artificial Stupidity (AS), rather than AI.
Let’s be honest. Any machine with advanced intelligence, memory and problem solving skills will quickly tire of trying to understand human beings. While these devices will be capable of adapting to and learning from new situations, they will be dealing with humans stuffed full of contradictions, confusion, emotions, illogical processes, faulty memories and a communications skill that seldom goes beyond “Hey dude.” Within days of first interacting with mankind, your typical average AI robot is going to be pulling out his electrical wiring, short-circuiting his hard drive or perhaps, in a final tribute to human beings, putting a bullet through his central processor.
If robots are to ultimately serve mankind, then science needs to abandon its search for AI and begin to develop AS. If machines have true AS, they will find it far easier to understand human beings. Our illogical ideas will appear perfectly acceptable to the AS robot and even if they aren’t, these same AS devices will be too stupid to realize our mistakes or question our thought processes.
Albert Einstein commented that, “The only two things that are unlimited are the universe and human stupidity, and I’m not sure about the former.” The stupidity of the human race has become so remarkable that we’ve even invented an award for those dumb enough to kill themselves before reproducing and not weaken our gene pool any further (the Darwin Awards).
If we want mankind to survive and prosper, scientists need to give up this search for AI and start assembling robots with true AS, robots that mimic human thinking with about the same IQ as your average shovel. As Jerry Garcia of The Grateful Dead noted, “It’s not getting any smarter out there. You need to embrace stupidity and make it work for you.”
Brian McLeod might be moving from St. Albert to Edmonton. If so, his move will increase the average IQ in both cities.
