Desert of My Real Life











The latest salvo in the “games good for you” vs. “games bad for you” debate has been fired.  For now, it seems that games are good for you.

Researchers at the University of Rochester chose 26 subjects who had never played action-packed first person shooter games like “Call of Duty” and “Unreal Tournament.”  Over a period of months, 13 subjects played these action games while the other 13 subjects played calmer, strategy-based games like “The Sims 2” (which is probably not really a game but that’s another post).  The researchers then tested the players’ ability to make quick decisions in a variety of situations involving visual and auditory perception.  Those who had played the action games were able to make good decisions based on the information presented 25% more quickly than those who had played the strategy games.  In addition, the action game players improved their skills at playing the games more quickly than the strategy game players. 

The theory behind this study involves the use of probabilistic inference, which is an intuitive form of the more formal tool called Bayesian inference.  Bayesian inference is used in all kinds of artificial intelligence problems to make good decisions based on evidence.  Our brains are constantly taking in visual and auditory information as we move through the world.  Using this information, we make inference based on the probabilities of certain events occurring.  For example, when we drive, we use our perceptions to make decisions such as when to brake or make evasive movements and so on.   That is, we make inferences based on the probabilities that we are constantly calculating based on information presented to us.  People who can do this more quickly and more accurately will make better decisions than people who are slower or less accurate.

This latest study suggests that playing a certain type of video game can train our brains to evaluate information quickly and make accurate judgments about the appropriate action to take in a particular situation.  So it appears that game playing can be beneficial and not just a waste of time.  At least that’s the logic I used to justify playing an hour of Dr. Mario Rx today.



{September 28, 2009}   Human Pain

I’ve been watching Battlestar Galatica on DVD.  One of the roles of science fiction, I think, is to raise controversial issues, to help us understand what it means to be human.  Although the original 1970’s miniseries was cheesy and not very interesting, a few changes to the original idea makes the recent TV show one of the best when it comes to asking difficult questions and making us think about things in a new way.

The basic plot of the show is that humans created machines which then evolved into autonomous, intelligent beings called Cylons.  Humans colonized twelve planets and after years of relative peace, the Cylons attacked the humans, destroying much of the human population of the colonies.  The survivors, including those aboard a number of space ships, are now on the run from the Cylons, struggling to survive a war with a superior enemy.

One of the major changes from the miniseries to the TV show is in the look of the Cylons.  In the miniseries, the Cylons were one of the cheesiest parts of the show, looking like robots made primarily of cardboard.  In the new show, some of the Cylons look like machines but now they are computer-generated and sophisticated.  But the most interesting change comes from the fact that Cylons can look and act just like humans.  They bleed and sweat and some of them are even programmed to think that they are human, leading to what appear to be emotional responses such as love.  Human-looking cylons allow the writers to raise questions about civil rights and justice and faith. 

For example, season one of the show, which aired in 2004 and 2005, raised issues about terrorism and torture and justice at a time when the Abu Ghraib scandal was fresh in the news.  The humans on the ship called Galactica discover a human-looking Cylon in their midst.  Their instinct is to kill the Cylon by putting it into space (because human-looking Cylons breathe oxygen just as humans do) but the Cylon claims that there are several bombs planted throughout the fleet, scheduled to go off in a short amount of time.  Sensing an opportunity to prevent these bomb attacks, the military commander sends the best human pilot, Starbuck, to question the Cylon (ok–so the plots are always completely logical).  The Cylon messes with Starbuck’s head, telling her lies containing just enough truth to make her wonder what’s true and what isn’t.  But he won’t tell her where the bombs are.  Starbuck notices that the Cylon sweats and reasons that if he sweats, he must feel fear and pain.  So she and her colleagues begin to torture the Cylon.

One of the most thought-provoking exchanges during this torture comes when Starbuck tells the Cylon that she recognizes the dilemma he is in.  He wants to be human because being human is better than being a machine.  But while he is being tortured, every instinct must be telling him to turn off his pain software.  But if he turns it off, he won’t be human anymore because the defining characteristic of being human is the capacity to feel pain.   I don’t know if I think that’s true or not but the conversation reminded me of research in machine learning that postulates that in order to really learn about the world, a robot must have a body. 

The importance of embodiment to learning comes from the observation that human knowledge, especially that most basic knowledge that makes up our “common sense”, is gained through via perception, through the interaction of our bodies with the physical world.  Not all AI researchers believe embodiment is necessary for learning.  Cyc is probably the most famous example of an attempt to codify all of human knowledge without the use of embodied machines.  The project was started in 1984 and has yet to be completed because of the difficulty of articulating all human knowledge.  Imagine trying to put all human knowledge into a computer by writing statements such as “Bill Clinton was a President”, “All trees are plants” and “Abraham Lincoln is dead.”  Each night, after spending the day coding statements like this, the researchers run some software (called an inference engine) which allows the computer to infer new statements about the world.  Each mornin, the researchers look at what the computer has inferred.  The inference process is somewhat flawed and the researchers find themselves having to correct some of the computer’s logic, encoding such bizarre facts as “If a person is dead, her left foot is also dead.”  Because of the difficulty of encoding these kind of facts, many researchers now believe that embodiment and direct experience of the world is a more efficient way to teach a machine about common sense knowledge.  So perhaps feeling pain is a necessary requirement for being human.

The same episode that contains this interesting conversation about the nature of humanity also contains a conversation about the purpose and effectiveness of torture.  After many hours of torturing the Cylon, Starbuck and her colleagues are visited by the President of the colonies who asks Starbuck whether she knows where the bombs are yet.  When Starbuck says no, the President asks why she has been torturing this man for eighteen hours, what makes her think she will get him to talk.  Starbuck replies that the Cylon is not a man which she seems to think justifies the torture.  The President orders that the torture be stopped since it has clearly not been effective.  The President later shows that this is not a sentimental choice, one that has been made because she is soft on the Cylons.  After getting the information she needs from the Cylon, she orders that he be placed in the airlock and sucked out into space so that he will no longer pose a threat.  The implication is that she ordered that the torture be stopped so that the humans would remain human, that the torture was damaging to the torturers and their humanity.

Themes of faith and love and treatment of outsiders and many other of the most interesting, controversial debates in our society run throughout this series.  I agree with Diane Winston, who said on Speaking of Faith that shows like Battlestar Galactica represent the great literature of our time, that people will come back to shows like this over and over, just as they read great books over and over.



My area of research when I was in computer science was artificial intelligence.  AI is a broad field with many subfields, each of which has many applications.  Within AI, I was particularly interested in pattern recognition via machine learning techniques. When I left computer science, I turned my research attention to the topic of this blog and began to focus more and more on the impact of technology on society and media technology issues.  So I was quite interested this morning when my favorite National Public Radio show, On the Media, broadcast a story that shows the connection between these two research interests.

Pattern recognition sounds like an esoteric subfield of AI.  But in today’s computer-focused society, there are many useful applications of pattern recognition.  For example, I worked on two problems in microbiology while I was a graduate student.  My master’s work involved looking for patterns in strands of DNA of an organism called Onchocherca volvulus which causes river blindness.  We were trying to determine whether we could determine the evolution history and path of the organism to help with understanding the epidemiology of the disease.  For my PhD, I worked on the famous “protein folding problem“, trying to predict the 3-dimensional structure of a strand of protein by looking at just the sequence of amino acids that make up the protein.  The theory is that if we can predict the 3-D structure, we can predict the function of the protein as well and the implications of that are far-reaching.  As I said, there are many practical applications of pattern recognition by computers.

On today’s edition of On the Media, there was a story that reminded me of the fact that pattern recogniton by computers is everywhere in our society.  The story was about a contest by NetFlix, the DVD rental site.  NetFlix allows subscribers to rate movies via a star system, where one star means “hated it” and five stars means “loved it”.  Based on the ratings that a particular subscriber has given a set of movies, NetFlix attempts to recommend other movies that the subscriber will enjoy.  NetFlix’s business model depends on these recommendations since a larger percentage of their movie rentals come from subscribers listening to these recommendations.  Without the recommendations, subscribers would likely run out of movies that they know they want to see and then would eventually give up their subscriptions.  But predicting what movies a person will like is a very difficult problem.  

NetFlix does a pretty good job with their movie recommendation system, Cinematch, but if they can make better predictions, they’re likely to hang on to more subscribers.  So they created a contest, offering a million dollars to anyone who can develop an algorithm that does 10% better in its predictions than Cinematch.  Apparently, a number of groups immediately were able to develop algorithms that were 5% more accurate than Cinematch.  Even getting to 8% more accuracy didn’t take that long.  But a number of intriguing issues made reaching the 10% mark difficult.  One of the most interesting is known as the “Napoleon Dynamite problem.”  Napoleon Dynamite is a quirky, independent movie that came out in 2004.  It seems that it is quite difficult to accurately predict whether a particular subscriber will like or dislike this movie.  In fact, two people whose likes and dislikes are quite similar can disagree drastically about Napoleon Dynamite.  So getting to the 10% mark will probably require a solution to the “Napoleon Dynamite problem.”

The contest closed a couple of days ago, although no winner has yet been announced.  NetFlix says that they received 44,014 entries from 5169 teams in 186 countries.  One of the requirements of the contest is that the winners must disclose their techniques to the world.  Although getting more accurate movie recommendations is not a  life or death problem, the solution to it is likely to provide insight into how to accomplish other pattern recognition tasks.  And that’s good news for all of us.



et cetera