The Computational Universe

I completely agree with Brooks…but unfortunately, only with the title statement of this blog entry. Basically, after reading Searle, and Brooks, and listening to class discussion I can only agree with the fact that what all of this discussion about the consciousness of machines boils down to is one’s core set of beliefs. This seems like a cop out answer — as if somehow, logically I cannot make a good case for believing computers will or will not ever be conscious. Maybe this is true, but there is no shame in it…in fact, it is what Brooks concludes, even after his diatribe on the ability of machines to be in some way conscious. What I can say, is that after our class discussion I do believe that Searle was mistaken on a major technical point about the ability of machines to pass the Turing Test and still not be conscious. I think on some level, our physical bodies carry out the same ‘mindless’ procedures as do machines. ‘Programmed’ reactions our body has since birth, it’s all true, machines do in fact resemble humans in the way they operate on a basic level. And if a machine were to become complex enough that it could interact with humans on the same deeply emotional levels that we do, there is no doubt that we would have to consider its understanding of the world, and its consciousness. The problem is, that I fundamentally do not believe that man will ever be able to know himself so well as to program a machine to act in the same ways and at such a high level as to essentially replicate himself. There are properties of our emotions–our souls–that are simply unknowable and impossible to simply simulate, and to save the class from reading about my religious beliefs I will write only that this understanding I have of the world is fundamental to what I believe religiously. I will admit, however, that I can see the other side of the coin. It is hard not to be enticed to believe that we could indeed create a machine one day that could pass the Turing test, what with all of the progress we have already made in terms of hardware capability and with such deep understandings of human interaction that we have (what went into making Cog was impressive). I suppose my idea that humans are special comes from something more than just consciousness. Our uniqueness as humans to me is something that surpasses the Turing test and its idea of consciousness. While Brooks makes a case for sympathizing with animals, my religious beliefs do not point to any fact that animals have souls as human beings do. For this reason, and my core sets of beliefs, I cannot ever believe that machines could be as special as we humans.

Orchestra n.

1. Music.
1. A large group of musicians who play together on various instruments, usually including strings, woodwinds, brass instruments, and percussion instruments.
2. The instruments played by such a group.

The Princeton Laptop Orchestra could be considered an orchestra if you inserted “computer simulations of…” after “usually including”. And this of course is a comment which may arouse some debate. Is PLOrk really just a cheap imitation of traditional instruments? Well, I would argue yes and no. The problem is, for at least me the listener, that when I was listening to PLOrk “play”, there were several things running through my mind. First off, some of the most beautiful portions of the concert were when I closed my eyes and thought real woodwind/percussion etc. were on stage playing. I thought “wow, what a great sounding simulation, it’s almost real.” Secondly, when I heard the somewhat gratuitous “computery” sounds thrown in here and there I found it somewhat derailing in my train of listening. Though Professor Lansky is a composer who I have always admired and whose work I enjoy greatly, I found that his composition generated the previous response more than the others. I guess I personally do not think that the sound of regular instruments meshed with awkward computer-like sounds really works well. If computer generated music is to become its own genre, I would much prefer the sound of computer generated noise to evolve into its own beauty without the help of simulated “real” instruments. Not only does it cheapen the genre itself, but it is pointless to watch students who look like they could be surfing the internet for all I know generate the simulation of a flute with the click of a mouse. More inspiring to me were compositions such as The Swamp (the second piece) which really made unique and spectacular use of computer generated noise. It was rhythmic, alarming, well done. I assumed that the mounting rhythm had to be worked through the mouse clicking, and this made it truly a unification of sounds among those “playing” their computers; a kind of synergy. This to me is more impressive than well timed entrances in a piece with instrument simulation. Keeping in mind that this genre is evolving, I might conclude that they put on a fantastic performance. But as a whole, PLOrk would do better to separate itself from the definition of a traditional “Orchestra” and the idea that it is just a computer simulated orchestra, and strive to come into its own in a completely new and unique way.

Logic formalizes or puts rules around our reasoning as humans. In the pursuit of ‘truth’, logic helps to organize our thought processes. Boolean logic in particular is extremely adroit at categorizing truths and falsehoods (not truths). Using the basic operators Not, And, Or, Xor, (as well as others which combine these, such as Nor) we may evaluate given statements. We evaluate statements based upon whether they fit the parameters of the operators rendering an expression either true or false. They are then given a value of 1 if the statement is true, or 0 if the statement is false. In written form Boolean logic uses truth tables, circuits, or boolean expressions. Boolean expressions use the symbols * for And, + for Or, a line over the variable for Not, (and others which we have not learned?). We may then evaluate the expressions if we know which variables are true and which are false. A truth table will manipulate the value of the variables so that all possible combinations of true and false for each variable has been covered. It will then evaluate the statements based on these combinations to reveal which combinations render the expression true and which render it false. Circuits are another way to visualize Boolean logic using differently shaped ‘gates’ for each of the operators. We can combine gates to create more complicated expressions.

Boolean logic is handy for our purposes because we as humans do not easily organize our own reasoning. Because of the simplicity of the rules in Boolean logic, we are able to program computers to compute truth tables or evaluate expressions for us. This proves extremely useful because truth tables can become massive as we increase the number of variables. Logic problems that might take a human days to figure out can be figured out by a computer using Boolean logic much faster.

The problem with Boolean logic is that it must assume that if we say a statement is true that it is true. Thus, the parameters of a statement may not actually be true negating the entire expression. Boolean logic has no way of evaluating a statement like ‘The book War and Peace is good’ as either true or false. It is still up to the human to determine the value of this statement. Other problems include fuzzy logic, which represents multifaceted evaluations of a statement. The example used in our reading was ‘the iron is not hot’. This does not mean that the iron is cold, but it could mean that it just isn’t burning, or that it is warm.

Robots are capable of doing a lot currently. Brooks outlines the research he had been doing on robots which might perform functions similar to animals, insects, or even human beings. He was able to create a very advanced robot, Bit (the baby), which involved dozens of sensors and outputs which were very similar to that of a real human baby, though based upon action-reaction type methods of processing (not on intelligence. As he mentions when he creates Genghis, this is something that slows down reaction time). However, currently, he has limited the production of such robots to that which today’s market would deem useful. So, I guess our question should be less directed toward what robots are actually capable of, and more about what they are capable of doing for us. The robots and intelligent entities (such as Babel fish and the MS Word paperclip) which we have today are directed mostly at fulfilling what humans would want to have done for them. Beyond this, robots could not be immediate moneymakers, but research is still being done on how lifelike we can make robots. Currently robots such as AIBO, Furby, and even Bit can only take the information given to them by their sensors and react accordingly to how they were programmed to react. However, that is the extent of their intelligence. These robots have no explicit understanding of the world around them, nor is it clear that they will in the next 10-15 years. Programming a robot to learn and discern even simple things like the difference between a table and a stool is difficult without trying to program every piece of knowledge into the computer that we have inherently as human beings. The ability of these intelligent entities is more focused around storage of information which we as humans are not easily capable of doing, as well as performing menial physical tasks for us. These are areas in which robots trump humans, but the capabilities of understanding the real world have yet to be developed.

Hi Everyone!

I’m a politics major and to be sure, I really enjoy researching and studying the field. I get a lot of questions, however, when I mention that I’ll be working for Microsoft over the summer. A lot of people ask “How in the world did you think to work at Microsoft when your major is politics?” Well I’ve always had an affinity for gadgets and technology and though I like the politics field, I don’t think I’m interested in taking my academic interest into the real world (too cut throat!). Although I’m not very “tech savvy” in terms of understanding what goes on behind the scenes of those interesting gadgets, I do work well with the end product. Of course, I won’t be programming or be involved with any of the technical product managing out in Redmond, what I will be doing is marketing for the firm. But I realized that I need to have a better base knowledge of the inner workings of the technology I would be promoting if I were serious about working in the field. And so this is one of the major reasons I took this class. So far I am loving the topics we will be covering, especially those about Google and the principles of programming. I think I will find those very useful when I’m out in Redmond this summer! One of the things I did last summer was to promote a the Tablet PC technology, and actually launched a campaign for Princeton which I think will end up with an adoption of a tablet to the SCI computer arsenal. Right now, they are selling through Lenovo (IBM) for $1000 off of the list price to Princeton students (sorry for the plug!). Just post a comment or write me if you want to know more

 edit update 1:33 Feb. 12 ~ The computer I most runs on Windows (of COURSE! What kind of intern would I be?) I own the Dell offered by SCI as well as a tablet which I use for class notes.