Author Archive for Anu

Syntax vs. Semantics: What is thought?

Upon reading Searle’s article, I do feel that artificial intelligence is possible only to a limited extent. I was impressed by some readings on the extent to which robots have become capable, but ultimately I do not believe that artificial beings will ever be able to have real thought.

Searle makes a number of good points about the limits of artificial intelligence and why the Strong AI thesis is wrong. He talks about syntax versus semantics – the fact that syntax is a set of formal rules and semantics involves meaning attached to those rules – and I feel that these are all very valid points. Mechanically and methodically doing a simple process does not alone imply thought; that is to say, obeying simple rules of syntax does not imply thought, and without some other meaning, there is no thought. That is where semantics comes in. Furthermore, I really thought that his mention of the brain as biological hardware included a lot of good points. People do tend to view the mind as “something formal and abstract, not a part of the wet and slimy stuff in our heads,” but in fact the mind is “just as much a biological phenomenon as digestion.” Like the digestive process depends on the structure and organization of digestive organs, the thinking process relies on the complex biological structure of the brain. It cannot be separated from the hardware and made into a program, workable on all computers.

Perhaps what makes this study so controversial is that the Strong AI thesis prods at a widely thought unanswerable, controversial question: Can we create life? If real thought is what distinguishes humans from other life forms, and if we can create thought on a machine, then it is as though we have created some kind of life form, which can independently act and think. I personally don’t feel that we can put into machines true thought processes and complicated emotions.

Our World as a Web

On April 5th I attended Sir Tim Berners-Lee’s lecture on the history and future of the World Wide Web. I very much enjoyed his talk and I felt that in his introduction Professor Arora very soundly described Dr. Berners-Lee’s contribution to the world as comparable to the wheel and other elements of technology that have drastically transformed our world and how we do business and live daily life.

While I felt that the lecture was a bit too technical, I felt that some interesting points were made and I gained quite a bit. Prior to the talk I didn’t know the difference between the Internet and the World Wide Web, so that was an interesting insight as well. In discussing the history of the web and some of the phenomena that have arisen from it, like web pop culture that involves things like the “blogosphere,” Dr. Berners-Lee touched upon some of the features of the internet that make it so unique and so transforming. For the first time, people have unbelievably easy access to huge volumes of information. In his book, The World is Flat, Thomas Friedman also comments on this fact, stating that instant access to knowledge through use of websites and especially internet encyclopedias like Wikipedia has completely changed the way we live. In fact, as I wrote this, I looked up Sir Tim Berners-Lee on Wikipedia to verify that he has a doctorate degree so that I could apply the correct title (in 2004 he was presented with an Honorary Doctor of Science from Lancaster University).

I didn’t completely understand Berners-Lee’s thoughts on the Semantic Web, so I searched for it on Wikipedia and read about it. As I understand it, it is a way to gain more use of computers by making web pages understandable by them; in some tasks it seems this would be useful, in others less so. Still, to standardize information and classification of things and to connect different technologies must reap benefits and result in more efficient use of technology.

Boolean Logic: Innovative AND Useful

Boolean logic is based on the idea that everything can be represented in binary terms – by either a 0 or a 1 – and based on this notation, operations can be carried out on this information. Some Boolean operations, AND, OR, and NOT, are operations I have used when working with data sets and Venn diagrams: AND is an intersection of two sets, OR is a union of two sets, and NOT is a complement. These can be expressed using truth tables, like those presented in lecture or in the readings, and they can be translated into circuits to compute functions. Another Boolean function that can lead to interesting results and offer ways to manipulate data is the XOR function. With simple rules for dealing with binary numbers, complex problems can be solved. I found it very interesting that the ADD function for N-bit numbers described in “An Invitation to Computer Science” is comprised of N 1-bit ADD functions. The way that addition is translated into Boolean circuits is a very logical process and this logic itself can offer solutions to problems in computer science and electronics. Boolean logic can be seen in many electronic applications – for example, according to a passage in “Flesh and Machines,” vending machines operate using Boolean logic. They exist in one of two states: either the full amount of change for a soda has been deposited (paid), or it has not been met (unpaid).

While it is evident that Boolean logic is useful in many ways, it has its limitations. It cannot be used for solving problems which cannot be represented in binary terms. Where there are many gradations or states, rather than just two, Boolean logic falls short. Furthermore, using Boolean logic there are proofs of the Existence of God and similar questions thought unanswerable by most, and I feel that in these cases Boolean logic alone does not explore all possibilities and cannot capture the full meaning of such heavy questions.

The Scribbler Robot - A Technical Thinker

The Scribbler robot exhibits artificial intelligence by reacting to stimuli and the environment: using its sensors, it can detect light, sound, and obstacles. It reacts to these detections according to instructions written in pseudocode - “if” and loop statements that specify what to do under what conditions.

Observing the Scribbler, it is evident that it acts based on Brooks’s innovations in the 1980’s that allowed robots to function by directly linking perception to action and eliminating the process of reasoning, which requires a lot of computation and storage of information (Brooks 42). I found it interesting that in the development of robots computing was compared to the neural functions of insects and through these comparisons was simplifed enough to be efficient in robots. In comparing the evolution of robots to the evolution of living beings, the robot’s functions were first simplified and then expanded to become more complex (Brooks 40). Indeed, it does seem as though the creature reacts to the environment as an insect or an animal would. It is astounding how this simple approach to robotics produces such a realistic model of a living being.

In writing the pseudocode necessary for the Scribbler to function, I found that writing pseudocode involves careful analysis of all the decisions, operations, and considerations that we take into account on some level in performing any action. For example, in programming the robot to draw a square, the proper angle had to be given first. Somehow, for us, this is a more intuitive action. If I were asked as a child to draw a square, I would be able to, likely without knowing that such an angle is ninety degrees, simply from being aware on some level of what a square was. Similarly, if I were asked afterwards to draw a star, I would know how to adjust my angles intuitively, but I would certainly not compute the appropriate angle. With a robot, each single step of the process, no matter how intuitive it may seem to us, is a technical calculation. There is no intuition, no history of knowledge that humans have.

I’ve learned from working with the Scribbler more on how robots work and how their sensors are designed. I feel there is much potential for more complicated robots - computer programming in general has become so advanced; what really prevents these advances from also impacting the world of robotics? It will be interesting to see how robotics continues to advance and how robots are instructed in the future to perform more complex tasks.

It’s 10 am and I am sleepy.

Hello everyone, I’m Anu and I’m a sophomore here at Wilson College from Houston, Texas. I am majoring in economics and getting a certificate in finance. So far the classes I am taking towards my degree have been pretty interesting, so that’s cool. I have a number of interests- I enjoy photography (I got a fantastic new Canon digital camera), movies, shopping, on-campus clubs, certain classes, literature, and so on. My favorite books at the moment are Friedman’s “The World is Flat,” which actually talks a lot about how computing has revolutionized society and business, and a classic that I only now got around to reading - “The Count of Monte Cristo.” As for on-campus organizations, I’ve been really involved with P-SAT (Princeton South Asian Theatrics) which was my first theatre experience and a lot of fun (see P-SAT shows when they come out!) and I also participated in other club events. Recently I’ve joined Colonial Club and it seems like it’s going to be a lot of fun.

I’m taking this course because I feel I am lacking basic knowledge about computers that I should know - how they work, for example… Also it fulfills a much needed ST requirement. I’m concurrently enrolled in COS126 - General Computer Science. I’ve had the tiniest bit of experience with HTML - I made my own webpage several years ago - but I think I’ll learn a lot of applicable programming skills in my class. So my hope is that after this semester I’ll be able to carry on an intelligent conversation about computers and even write my own applications.

Here’s my edit: I have always used Windows and I currently have a Toshiba Satellite series laptop that I adore.