The New Great Depression: A Brief History of Artificial Intelligence and its Role in the Collapse of the Economy
As humanity has progressed through time, we have always been looking for ways to make our lives easier. It started with sharpening a rock to hunt and has progressed to using nuclear decay to generate electrical power. Though these devices were built to make a task at hand easier, there will always be people who use these tools to do great things. The pyramids, for example, couldn’t have been constructed without the lever or the ramp. The Colosseum: impossible without the invention of the pulley. Inventions like these are what makes humanity move forward as a species, but they always must leave something in the dust. In the late 19th century, the invention of the steam engine sparked the industrial revolution, a manufacturing innovation that changed the world forever. The machine that the steam engine replaced was organic: it was the horse. When horses were replaced by the steam and internal combustion engines, they quickly dropped out of public eye and their uses today, as we know, are purely ceremonial. The steam engine marked humanity’s ability to create mechanical muscle, but 100 years later, engineers are working to create mechanical minds that can outmatch a human in processing power. Will our fate be that of the horse’s? Will we become obsolete?
In 1958, Jack Kilby, a researcher at Texas Instruments, created the world’s first integrated circuit (The Chip That Jack Built, 1). An integrated circuit (IC) is an electrical circuit that is housed on a single substrate (board). This drastically reduced the space required for electronic devices. One of the things that made Kilby’s discovery so revolutionary was the fact that he used a transistor in his design. A transistor is a type of electrical switch that can be turned on and off at thousands of times a second. These two groundbreaking inventions were developed in parallel for the next few years. In 1965, the co-founder of Intel and Fairchild Semiconductors, Gordon Moore, made a prediction about the future of electronics that continues to hold true to this day (Mack, 1). He stated that every two years, the number of transistors in a given area of an IC will double. For the next 4 decades, his postulation has been nearly spot on with the cutting edge of industry. Until electrical engineers reach limits that have been recently discovered by quantum physicists, this exponential increase in processing power will continue for the foreseeable future. But what does that mean for us? Computing power is directly related to the number of transistors in a chip. As such, computing power has increased drastically since Kilby’s day. This computing power has been put to use by programmers to do all sorts of thing that save humans time and money. When the first practical programmable computers were introduced, they drastically increased the throughput of accountants and scientists who now didn’t have to wait for their calculations to come back from actual “Human computers”.
Nowadays, we rely heavily on computers for anything from getting the day’s weather report to playing video games with people from all over the world. Though consumer uses like these are incredibly common, much of the research that pushes the envelope of society is done in the dark recesses of neural networks and artificial intelligence. Artificial intelligence is a relatively new field. It was first explored in depth in sci-fi stories like 2001: A Space Odyssey, Blade Runner, and I, Robot. These stories captivated the minds and hearts of many people, particularly children. Little did they know that as early as the early 1960’s, these robots were being worked on by scientists and engineers to become a reality. Projects like AARON (Prosthetic Knowledge, 1), Nouvelle AI, and ALVINN may not ever be known by the populace for their role in the development of Robotics or AI, but there are some popular robots that are widely known by the public. Watson, IBM’s supercomputer that recently beat the two best players in the history of Jeopardy, or Deep Blue, the supercomputer that beat Garry Kasparov,
the world’s best chess player, in 1997. These isolated examples themselves do not spell death to all humans, but they do usher in a sobering reality that humans can be beaten in games we traditionally thought ourselves the best at. The ability of mechanical minds to replace other human tasks has only increased over time. When the first industrial robots were introduced in the late 1960’s, they quickly took the manufacturing world by storm. That was then and this is now. What can practical robots do today?
In the United States, the transportation industry provides employment to roughly 4.4 million Americans. It is the third-largest blue-collar industry in the US. Replacing these jobs with robotic drivers would be very difficult. It would require many different areas of math and computer science to do so. From PID loops and Kalman filters to linear transducers and quadrature encoders, replacing a human driver is incredibly complex. The decisions that a driver must make are crucial to the safety of himself, the people around him, and the shipment he is carrying. Enter the Google Car. Google has been working on an autonomous vehicle project for the past 5 years, and their progress has been breathtaking (Wood, 1).
Recently, their fleet of self-driving cars was allowed to drive autonomously on California Highways and has had an incredible record of 0 at-fault collisions in over a million miles of travel (Wood, 2). This represents an unprecedented step forward for the world of automation. If we can build systems that can react to the complex behavior of human drivers, directing these cars from point A to point B is almost trivial. GPS systems have been guiding us humans around for years, and what’s to say they can’t guide robot drivers around as well? In the next 20 years, automated drivers are expected to take over their human counterparts (Wood, 1). This presents a problem. If, for example, these jobs were replaced by robots, that means there are 4.4 Million people out of work. Another great industry for automation to move into is the retail sector. Point-of-sale and cashier robots have already been rolled out, and McDonalds has begun replacing some of their workers with self-order kiosks when they go on strike (Rubics, 1). Though the long-term ROI on these machines has yet to be determined, initial reports show they work the same or better than their human counterparts. This industry employs about 15,000,000 people annually. If these two sectors alone were to be automated, there would be an increase in unemployment of 16%. Added to the projected unemployment figure for 2020, this would bring the unemployment rate to over 20%, well within Great Depression figures (U.S. Bureau of Labor Statistics, 1). In just these two industries alone, the effects of automation can be seen to be devastating. But what if you don’t work in a factory? What if you play music for a living?
A commonly held belief is that just like how the steam engine freed up humans to specialize in other things like factory and production jobs, Artificial Intelligence will allow humans to specialize in creative roles like artists and musicians. While this is possible, it would not be feasible with the current economic landscape of the world. The global economy is by and large a capitalist one. For it to function properly, there must be many consumers for each producer to pull from. This is the exact same in music and art circles. There can be only a handful of very talented artists and millions upon millions of devoted followers. If everybody started producing their own music and marketing it to others, the supply of music would far exceed the demand for it and the market would become diluted with low-quality products. (Thompson, 1). To quote CPG Grey, “We cannot survive as a capitalist economy when millions of people are simply unemployable, by no fault of their own.” (Grey, 1)
In conclusion, the problems we face with respect to AI are new ones. Never before has such a drastic technological innovation been poised to create such waves in the world’s economy. When it comes to automation, very few jobs are safe. Though humans have successfully survived many technological developments over the years, we have never dealt with one that could potentially render us unemployable. Through a combination of economic and political reform, our governments must work together to secure the future of the world. We must ensure that we don’t become obsolete.
Works Cited
“AARON — The First Artificial Intelligence Creative Artist.” Prosthetic Knowledge. N.p., n.d. Web. 1 Dec. 2015.
Humans Need Not Apply. Perf. CGP Grey. CGP Grey, 2014. Online Video.
Mack, Chris. “Cover SemiconductorsProcessors The Multiple Lives of Moore’s Law.” IEEE Spectrum. IEEE, 30 Mar. 2015. Web. 2 Dec. 2015.
Rubics, Darius. “New McDonald’s In Phoenix Run Entirely By Robots.” News Examiner. News Examiner, 4 Aug. 2015. Web. 1 Dec. 2015.
Thompson, Derek. “A World Without Work.” The Atlantic. Atlantic Media Company, 22 June 2015. Web. 1 Dec. 2015.
“U.S. Bureau of Labor Statistics.” U.S. Bureau of Labor Statistics. U.S. Bureau of Labor Statistics, n.d. Web. 08 Dec. 2015.
Wood, Andrew. “The Future of Self-driving Cars: CNBC Explains.” CNBC. CNBC, 17 June 2015. Web. 1 Dec. 2015.
