Nov 23, 2011Science and Technology
Finding dinosaurs, criminals, and playing jeopardy; the uses of Artificial Neural Networks

I’m sure I wasn’t the only first grader to fantasize about being a paleontologist (although I called it dinosaur hunting).  As a child, I was fascinated with this field of scientific research and to be completely honest, I still am; however, when I learned the amount of dedication and patience it took to pursue such a career, I realized it wasn’t for me.  I definitely have dedication to spare, but patience, on the other hand, is another story.


Fossil hunting is notoriously tedious and usually involves meticulous scouring of aerial imagery in the search of regions rich in sedimentary rock.  This is then followed by a meticulous ground survey, and then a meticulous search for any elusive fossils that may or may not actually be present.  While relying on hard-work, experience, and science-acumen, a healthy dose of luck also plays an important role in the search for fossils. 


According to the press release, Human, artificial intelligence join forces to pinpoint fossil locations, a researcher developed and tested a new artificial neural network system, which by inputting important data points, successfully pinpointed a high percentage of known and hopefully unknown fossil sites.  The artificial neural network won’t ease all the hard work involved in fossil-hunting. It will, however, ease the first steps required in finding an optimal dig site.


What exactly is an artificial neural network (ANN), besides, of course, being a clear indication that computers have come a long way since I was sitting in my room playing text-based video games on my big brother’s commodore 64?  ANN is a computer/mathematical model that uses the basic principles of a neuronal network (our brain) in order to make rapid and complex calculations.


The use of ANN has made some major splashes in the media.  How many remember Deep Blue?  Deep blue, an ANN computer turned chess player was the brainchild of IBM computers.  In 1997 Deep Blue became a grandmaster chess player after beating former reigning champ Garry Kasparov in 1997.  In 2011, the IBM ANN computer, Watson, won Jeopardy and beat two of the best humans to ever play the game. 


More than playing games, ANN systems are often used when standard means of calculations are simply insufficient. ANN systems are used to successfully analyze massive amounts of science data in a relatively short amount of time and make more accurate conclusions through improved statistical analyses.  ANN systems are presently being tested on their ability to streamline drug discovery methods, improve diagnostic tests, and digitize overall healthcare.   


This last summer, the Santa Cruz police department tested out a new ANN system.  By analyzing a massive amount of crime data in tandem with Google maps, the ANN system successfully predicted crime, leading to the arrest of two would-be car thieves.   


Perhaps this is my paranoid nature coming out, but the use of an ANN to predict crime is a very slippery slope.  First comes prediction, but in the hands of an overzealous person, then what’s next?  Thought police? One can imagine science-fiction scenarios like the movie "Minority Report" with Tom Cruise, where criminals are arrested for crimes they have not yet committed, based on computerized predictions that they would commit them.


ANN systems are also a type of artificial intelligence.  They are based on our brains, and are obviously smarter than us; after all they did beat our best jeopardy players.  So it begs the question, what hope do we have?  An ANN system is an extremely oversimplified version of our brain, and while they may be excellent computers, they are not replacements for the human mind.  So, if your fear is that an ANN system, will, in the nearby future, cause machines to up-rise against humans, don’t worry.  Unless our livelihood is somehow attached to a life and death game of chess or jeopardy, I think we are okay.


Related Articles: Artificial Neuronal Networks

Be the first to comment.