My research is in the area of reinforcement learning, where an agent is learning to solve a sequential decision making task. In it, the agent takes an action from state, which leads it to some new state and provides it some reward. The outcome of the actions can be noisy or random. For example, when you’re driving your car and you make a decision to take a particular turn, there is some noise in the outcome (you can not predict deterministically what will happen). Instead, there may be construction, traffic, pedestrians, etc which cause your trip down this road to take longer or shorter than you would expect. You can’t predict exactly the number of seconds it will take. However, this randomness, at its core, is really the result of partial observability. Partial observability is where you can only observe some part of the world to make your prediction. In this example, you only have access to what you see at the intersection and the info you get from the internet, radio, tv, and your phone. With this information, you cannot predict the time the road will take precisely and instead there is some noise to it. However, if you knew what every other driver was thinking and the route and timings they would take, along with what the weather would do, what the exact construction plans were, etc, then you could conceivably make this prediction accurately (or a computer could). So you can view this decision making task as having noise or being partially observable. This goes on at all levels, but at some point you would need knowledge of every atom and molecule in the system (which can’t be observed), so its easier to approximate things as being noisy. But as we get better and better at observing such things, we can make more deterministic predictions that consider all the factors that we previously ignored as “noise.”

A lot of interviews I’ve seen with sports stars recently has made me think of this. I’ve seen quite a few instances (like this one: http://msn.foxsports.com/nfl/story/buffalo-bills-stevie-johnson-blames-god) of some seemingly random event happening, and the athlete’s response is to credit it to God. Instead of believing in randomness, they believe the world is partially observable and God is the hidden observation that can not be seen. Random events aren’t random, but instead are things that God did, and of course we couldn’t observe God’s decision making process. Anyway, I thought this connection was interesting, I’m not sure if its better to think there’s some force deciding all outcomes or to think that some things do just happen randomly. But it’s an interesting question: Is God simply an non-observed feature of the world state?

There was an interesting blog post on the NY Times today about an experiment that showed that the decisions of monkeys could be predicted before the monkeys knew they had made the decision. The part I found more interesting connected with my previous post on free will and how the ability to predict our decisions does not have to mean that we don’t have free will:

Let me explain what I mean by way of an example. Imagine we suspend a steel ball from a magnet directly above a vertical steel plate, such that when I turn off the magnet, the ball hits the edge of the plate and falls to either one side or the other.

Very few people, having accepted the premises of this experiment, would conclude from its outcome that the ball in question was exhibiting free will. Whether the ball falls on one side or the other of the steel plate, we can all comfortably agree, is completely determined by the physical forces acting on the ball, which are simply too complex and minute for us to monitor. And yet we have no problem assuming the opposite to be true of the application of the monkey experiment to theoretical humans: namely, that because their actions are predictable they can be assumed to lack free will. In other words, we have no reason to assume that either predictability or lack of predictability has anything to say about free will. The fact that we do make this association has more to do with the model of the world that we subtly import into such thought experiments than with the experiments themselves.

The rest of the blog post was a bit crazy and I think extrapolated a bit much from these experiments, but I think the excerpt above makes clear the argument that we can still have free will even if it turns out we can predict all our decisions by looking at all the neurons, chemicals, signals, inputs, etc to the brain.

Random chance. Somehow the big bang happened, the universe was created, stars formed, our star formed, planets formed around the sun, the Earth formed, it was the right distance from the sun, it revolved and rotated at the right speeds, it had a moon and tides, life appeared, it evolved, we arrived, and no man-made or natural disasters have wiped us out (like the dinosaurs). Eventually your grandparents were born, your parents, and you. That’s it. It’s entirely random dumb luck that we (and you) are here. There’s no special meaning, no higher purpose that we’re here for. We just happen to have occurred, we are here for maybe 100 years and then we die.

We die. Think about that for a minute. You will at some point cease to exist. It’s hard to imagine. You try to think about what it will feel like to die… but you can’t. It won’t feel like anything. You won’t exist. You’ll be gone. There is no magic, no afterlife, your consciousness isn’t going to live on or be reincarnated. When life ends, that’s it. Your existence is really really short, make the most of it!

What is the meaning of life? There isn’t one. Life just happens. Is that depressing? On the contrary, I think life is meaningless until you see that it has no meaning. Then ‘life is what you make it’. Life is random and brief, and that is what makes it so beautiful and amazing. So many things had to fall the right way for you to exist and it will be gone so fast… So appreciate it! Savor it! Enjoy it! Do something with it! Do you want to be remembered after you’re gone? Build relationships! Help others! Contribute to society!

This is a pretty interesting article on children learning self-control by playing make believe: Can the Right Kinds of Play Teach Self-Control?. Play can be useful! I wonder if I can put this idea into an artificial agent somehow.

It seems like people are really stuck to this idea that we’re unique. Perhaps they lose something if they feel we aren’t special. Back in the day, it was that we were at the center of the universe, and anyone who said otherwise was a heretic. Now we’re in the midst of some people saying we have to be unique from the animals, we couldn’t have evolved from them. And then there’s intelligence. We have to be the only ones who are intelligent right? According to Searle, even if we have a machine that behaves and acts indistinguishably from a human, it can’t be intelligent - it doesn’t have our secret brain stuff. Well I’ve got news for everyone. We’re are unique. Each one of us is different in subtle or not so subtle ways. But we’re not special, other creatures can be intelligent, other planetary bodies can have things orbiting around them, other planets can have life. Accept it.

Here’s an interesting article on overconfidence and the banking collapse: Cocksure. It’s interesting how overconfidence can ruin you and lead you to bad decisions, yet in some areas, overconfidence is what’s required, when confidence is vital. Here’s an excerpt:

This is what social scientists mean when they say that human overconfidence can be an adaptive trait. “In conflicts involving mutual assessment, an exaggerated assessment of the probability of winning increases the probability of winning,” Richard Wrangham, a biological anthropologist at Harvard, writes. “Selection therefore favors this form of overconfidence.” Winners know how to bluff. And who bluffs the best? The person who, instead of pretending to be stronger than he is, actually believes himself to be stronger than he is. According to Wrangham, self-deception reduces the chances of “behavioral leakage”; that is, of “inadvertently revealing the truth through an inappropriate behavior.” This much is in keeping with what some psychologists have been telling us for years—that it can be useful to be especially optimistic about how attractive our spouse is, or how marketable our new idea is. In the words of the social psychologist Roy Baumeister, humans have an “optimal margin of illusion.”

There’s an interesting article in the latest edge on free will and consciousness. Basically confirming what I already believed after reading lots of Dennett and others’ books on the mind. Most of our decision making occurs in the subconscious. Our conscious self either only does the difficult decisions, or is only told after the fact so that we feel that ‘we’ made the decision. This doesn’t mean we don’t have free will: our subconscious decisions are still made based on who we are, our genes, our lives, how we grew up, etc. And Bargh raises another interesting point, do we really know who we are if our decisions are made by our subconscious and not our conscious selves as we believe? Here are some highlights from the article:

All organisms are purposive and have reasons for what they do. We certainly have that of course. So it’s not that will doesn’t exist; it’s that the free part is problematic — a lot of people see free will and say, “Well, you’re showing there’s no free will; therefore, people have no intentions or will.” No.

There is will, and will can be shaped by a host of factors: your genetic background, your early experience with your home and your family, your caretakers, you playmates, cultural influences bombarding us through the media and through socializing with your peers (and, thus what they like and what they think and what they believe from their parents). All this is being soaked up like a sponge by little kids.

…

But there is another question that is more pragmatic and I think it’s a wonderful question, “If all these things are going on without my knowledge, then I don’t really know why I’m doing what I’m doing, and I don’t really know myself that well apparently. So how can I make the right decisions or make the right choices for myself when all these biases are throwing my decisions all over the place?”

As a grad student studying artificial intelligence, it might be useful to know what exactly intelligence is.  No one can really define it.  We know that we have it.  We know that tables and chairs do not.  Everything in between is up for grabs.  I’m going to discuss what some historical views on intelligence are and what I think it might be.

The classic definition of intelligence in AI is based on the Turing Test.  This is a test created by Alan Turing to determine if something is intelligent.  You have a person and a computer in separate rooms.  In a third room is a judge.  He chats with both of them over a computer.  If he can’t tell which one is the person, then the computer is intelligent.  Basically the computer is considered to be intelligent if it can imitate a person in conversation.  If it doesn’t have this ability then its not intelligent.  This test can be extended to a more complicated task, say if you can’t tell a person and a computer/robot apart in every day life, then the computer/robot would be intelligent.

Then there are a class of people who believe that having intelligent behavior is not enough.  They say its critical what is going on to make that intelligent behavior.  Usually they say that having computer circuits create such behavior is not intelligent.  Some of them say that only having an exact human brain create the behavior is intelligence.  The classic argument from this side is John Searle’s Chinese Room.  There is a man in a room.  He receives pieces of paper with weird symbols on them.  He has a set of directions of what to do when he receives these pieces of paper, which generally involves him creating new weird symbols and passing them back out.  The symbols are Chinese and he is actually answering someone’s questions with someone in Chinese this way. Searle’s main argument is that the man does not know Chinese even though he can act like he does this way.  The room also does not speak Chinese.  Searle goes on to say that even if the man and directions were replaced by a fully accurate simulation of the human brain it would not be intelligent, you need the magic stuff of the brain to be intelligent.  Another point to make is that this example is utterly impossible, it would not be possible to create a set of directions to converse fluently in Chinese.

Another argument from a more computer science perspective is that of Simon and Newell.  They put forth the ‘Physical Symbol System Hypothesis’, which states that “A physical symbol system has the necessary and sufficient means for general intelligent action.”  They define a physical symbol system as any system which physical patterns (symbols) that are related somehow (in a system).  Basically, anything.  They later go on to say that one of the signs of intelligence is avoiding exponential search (Many problems can be reduced to searching through a space of possible solutions.  If the search space is exponential, then it is impractical or impossible to search it exhaustively.  Avoiding this exponential search requires some intelligence in deciding how to avoid it).  This is a pretty interesting idea.  I often feel that one of the key things to do to be successful is to know how much time to put into something and when its not worthwhile anymore, which seems somewhat related to this idea of avoiding exponential search.

Recently, our reading group read a paper by Ned Block that also discussed this topic.  Block is part of the camp that says that the internal processing that creates the intelligent behavior is critical to whether something is intelligent.  He provides a counter-example to the Turing Test.  A computer is loaded with a database of every conceivable conversation that is less than an hour in length.  Then when taking the Turing Test, it looks up a conversation that matches up to the current sentence, and speaks the next sentence in that conversation.  He says that this machine is clearly not intelligent, and therefore intelligent behavior is not enough.  He does allow that computers could be intelligent, but they would have to be doing something better than this, such as learning or adapting.  Another student in the group brought up a great point that the Turing Test is fine if you say that it has to be a feasible machine that passes for a human.  Both this counterexample and the Chinese room are not feasible agents (the number of hour long conversations is much greater than the number of particles in the universe).  Any feasible agent (reasonable memory and computation power) that passes the Turing Test must be doing something intelligent to be acting that way with limited resources.  This is an interesting definition that lets the Turing Test stand and does put a restriction on the internal processing to be ‘feasible’, which any actually realized agent would be.

So where do I stand?  I’m not sure.  Intelligence may be defined on some scale by behavior.  More complex or more efficient behaviors come from more intelligent beings.  I’m not sure that a definition with a restriction on the internal processing of a being will work.  Intuitively I don’t think an agent that is simply looking up a conversation in its database like above is intelligent.  An agent that is learning or adapting seems to me to be more intelligent.  But depending on what level of abstraction you look at, even we aren’t doing anything that exciting.  Sure, you say that we’re understanding what is said to us, thinking of a response, using our memories, and responding.  Which sounds intelligent. But at its basest level, its just chemicals and ions flowing back and forth in our brain, following the inexorable laws of physics.  Which sounds very unintelligent.  The same thing applies in the case of AI.  You can write some cool learning algorithm that adapts and learns how to behave in some environment over time.  But of course how it learns was written by you and is pre-determined and the entire course of what it would do in that environment could probably be predicted if you knew all the variables.  So I don’t think any definition of intelligence that involves the internal processing going on could ever work, because every agent simply has very boring particles following the laws of physics at the core of their internal processing.

In summary, I think a definition of intelligence has to be about producing intelligent behavior and not about the internal processing that creates it.  The Turing Test seems pretty reasonable if you restrict it to feasible agents.  And I still think algorithms that let machines learn and adapt are more exciting that simply programming in a static solution to a computer (I prefer to think of things at the learning abstraction level than the level of particles following the laws of physics).

I just read Starship Troopers by Robert Heinlein. Some of the most interesting parts of the books are the descriptions of the main character’s classes in “History and Moral Philosophy.” This takes place in the future and looks back at the “failures” of our society and compares them with their solutions in the future. Some of the most interesting parts are about morality, the survival instinct, and how it relates to democracy.

The teacher in the book, Mr. Dubois, says “Man has no moral instinct. He is not born with moral sense. … We acquire moral sense, when we do, through training, experience, and hard sweat of the mind.” This is an interesting concept. Certainly when you look at animals you don’t think they are born with a moral sense. What’s more interesting is where he says our moral sense comes from. “What is ‘moral sense’? It is an elaboration of the instinct to survive. The instinct to survive is human nature itself, and every aspect of our personalities derives from it.” Where the moral sense comes in, is whose survival is imperative to you: “survival can have stronger imperatives than that of your own personal survival. Survival of your family, for example. Of your children, when you have them. Of your nation, if you struggle that high up the scale.” He claims that their theory can solve any moral problem, on any level: “Self-interest, love of family, duty to country, responsibility to the human race.”

This discussion of morality gets really interesting when he ties it in with their political system. In their system, you must serve your country (in some capacity: military, science, research, engineering, etc) before you are allowed to vote or run for office. But this service is not required, moreover it is done on a volunteer basis and discouraged so that only those who are truly committed accomplish it. Mr. Dubois explains that “Under our system every voter and officeholder is a man who has demonstrated through voluntary and difficult service that he places the welfare of the group ahead of personal advantage. And that is the one practical advantage. He may fail in wisdom, he may lapse in civic virtue. But his average performance is enormously better than any other class of rulers in history.”

I’m not advocating this system of government, but this argument makes a lot of sense to me. I’ve argued in previous blog posts here that one of the biggest problems with our government is that people are too concerned with themselves than the best interests of the country. Politicians are more concerned with raising enough money to keep their jobs than serving the country well. Many voters are more concerned with how the policies affect them than what it means for the country or for future generations. Restricting the decision making to only those who actually put the best interest of the country ahead of their personal interests would make the system more successful. I’m not quite sure how to do that. Even when we have candidates who put the country ahead of themselves, they don’t necessarily get elected because there are enough voters who don’t do so. Certainly some interesting questions to think about.

Here’s a quote that sums up how I wish the government (and its voters) operated: “In every deliberation, we must consider the impact of our decisions on the next seven generations.” — The Iroquois Confederacy

There was an interesting article by Andrian Kreye in the Edge this week on religion and reason.  Specifically it discusses some of the scientific research into faith that has been going on and has been mostly ignored by the “militant atheists” Harris, Dawkins, and Dennett.  Kreye discusses research by Scott Atran, who wondered what benefit religion had that made us pay such a cost in time and effort and lives to try to overpower rational explanations.  His conclusion is that religion must have had some evolutionary benefits such as the closer communities it can build.  Justin Barrett has done research showing that faith may be important developmentally, as small children have unwavering faith in their mother’s infallibility when they are young.

I thought the most interesting part of the article was its conclusion, which said:

“One advantage faith has over atheism is that it offers hope for an afterlife. Thus far, we have found only religious answers to assuage the fear of death. It always comes down to a choice between delusion and reality. Reality just may make you love your life so much more.”

This echoes something I have tried to say in previous blog posts on religion but I don’t think I was ever able to state it as clearly or eloquently as this.  It may be more difficult to accept the shortness of our lives than the afterlife promised by religions but I think understanding reality allows us to appreciate and live our lives that much better than living it for some supposed afterlife.