The Machine That Changed The World

PBS Video Series
A production of WGBH and the BBC


Topic Summaries & Notes
© 1997-9 by David J. Stucki


Many of the links provided below provide more detailed or in
depth information about the highlighted topic or person. Others may lead
you to only loosely related sites that are worth exploring.

(Thanks to John A. N. (JAN) Lee for providing links to many of the resources listed below.)

Table of Contents

Giant Brains  The Paperback Computer  The Thinking Machine 


Giant Brains, Episode 1

Introduction

Writing was invented 5000 years ago as a means of expanding our ability to communicate. The computer may come to rival the written word as a communication medium. 

Most machines are special purpose, that is they are designed to do only one thing. Computers, on the other hand, are universal machines. They can transform themselves into a variety of other machines through the use of software. As a result the computer is a uniquely versatile machine. 

This universal character of computers allows us to describe their functioning as:  In general, computers perform tasks that in humans requires intelligence to accomplish. Originally, computers were invented to provide a fast and effective means of performing arithmetic.  

The Human Computer

Prior to World War II the word computer referred to a person who computed arithmetic tables, such as for logarithms or numerical constants. For example, William Shanks spent 28 years of his life calculating the value of p, the ratio of the circumference of a circle to its diameter,  to 707 decimal places. A modern desktop computer can solve the same task in 7 seconds or less. 

The human computer relied primarily on tables (prepared by other computers) and often made calculation errors, transcription errors, or propagated errors located in the tables. For example, Shanks made an error in the 528th decimal place and so the last few years of his efforts were wasted. 

The industrial revolution was built on numbers. As technology grew it created an increasing demand for accurate and reliable mathematical tables. 

Charles Babbage was a victorian mathematician who became frustrated by the amount of wasted time caused by computer errors. He independently commissioned two sets of tables in one instance to allow them to be checked against one another for accuracy. (See also The Babbage Pages, and the History of Mathematics archive.)  

The Difference Engine

Babbage conceived of an idea of an engine that would compute by steam or some other means of automation. He designed a machine he called a difference engine, so named because it computed using a technique called the method of differences. Because of the practical usefulness of this machine, the British government granted Babbage the financing he needed to build it. The machine, however, was never completed, partially because he was a poor manager and partially because he came up with a better idea. 

The difference engine only could compute results by the single method of differences, and was not, therefore, very versatile. Babbage envisioned an alternative, general purpose machine that would be programmable. He named this device the analytical engine. As a means of encoding the programs, Babbage borrowed the idea of using a series of punched cards from Jacquard, a french textile maker, who had used them to create looms that could weave complex patterns. 

This was the first machine that had been designed not with a specific purpose or function in mind, rather that was left to the user to specify. The concept of software was born. 

The First Programmer

Augusta Ada Byron, Countess of Lovelace, was the poet Lord Byron's daughter. She became fascinated with Babbage's analytical engine and is largely regarded as the world's first computer programmer (although there are some who disagree). She formed a long term association with Babbage after translating a paper written about his work. She spent a considerable amount of time working on the implications of a programmable machine and what it might be capable of. (See also her biography, or another, or other miscellaneous information about her.) 

A New Era Dawns

During the decade between 1935 and 1945 the definition of the word changed dramatically. A common dictionary from before this period identified a computer as a person who performed calculations. By the end of World War II the word computer referred to a machine that was used for the same purpose. 

Konrad Zuse

Konrad Zuse picked up where Babbage left off, implementing several improvements and innovations along the way. His computer was the first electromagnetic computer, whereas Babbage's had been purely mechanical. He also pioneered the idea of using electric relays (common in telephone systems) as simple components, using binary arithmetic simplify the engineering. This was a considerable improvement over Babbage's decimal machines. By 1939 Zuse was the world's leading computer designer. To conserve resources and money during the war, he used discarded movie film instead of punched cards. 

He also conceived of the idea of building a completely electronic computer out of vacuum tubes, but was denied funding by the German High Command because they thought they would be able to win the war before he could finish building it. The idea of the vacuum tube machine, however, was a breakthrough since it could have speeds up to 1000-2000 times as fast as the mechanical relay machines. 

ENIAC and its Peers

The U.S. government was highly motivated during the war to develop digital, electronic computers to help them calculate artillery guidance trajectories. They had been utilizing the skill of human computers, such as Kay Mauchly from the University of Pennsylvania, who was employed building firing tables. Manually, it took about 4 years for one human computer to complete a single set of tables. 

John Mauchly and J. Presper Eckert, who were at the University of Pennsylvania's Moore School of Electrical Engineering (founded by Benjamin Franklin), developed a plan to build an electronic digital computer using vacuum tube technology. Their design called for 18,000 tubes (which had a failure rate of one every five seconds), which filled a room 50'x30'. The ENIAC (Electronic Numerical Integrator and Calculator), in addition to its complexity, incorporated several innovations. It was built according to a modular design that allowed it to have built-in fault tolerance (redundancies and backups to prevent overall system failure). The ENIAC was a technological triumph, but it was too late for the war effort. 

Kay Mauchley became an ENIAC programmer, which involved learning all the circuit diagrams for the machine, since reprogramming it involved rewiring all its circuitry. There were no user manuals in those days. Because of the work involved in reprogramming the ENIAC it spent much of its time sitting idle. 

America's most distinguished mathematician, John von Neumann, joined the ENIAC project towards it end and authored the report describing a new kind of machine that could store its programs in memory, rather than in the physical configuration of the machine. This stored program computer became the model for every modern computer designed and built since. 

Due to contractual and patents disagreements, Eckert and Mauchly left the University of Pennsylvania to start the first commercial computer company,where they developed the UNIVAC  

In July, 1948, Freddie Williams designed and built the Mark 1, the first working stored program machine at the University of Manchester. 

Maurice Wilkes, designed and built the Edsac computer (Electronic Delay Storage Automatic Calculator) with the idea of presenting a more user friendly interface that would attract the interest of scientists who may not have been competent electrical engineers. 

Predicting the Future

The invention of the computer ushered in the birth of entirely new disciples of science and research. For example, the field of radio-astronomy could not exist without its incredible speed and computational power. 

In 1950, most scientists might have been able to envision the continued innovations and technological breakthroughs that led to the development of the supercomputer, but few of them would have been able to predict the way in which personal computers have become a commodity. In fact, the popularity of the personal computer is very much a surprise given the ways in which the early computers were utilized. 

There were those, however, who suspected that computers were far more significant a technology than simply a fast means of automatic calculation. 

Alan Turing

Alan Turing was a British mathematician who was instrumental in the British military's code-breaking efforts in World War II. He is also recognized by many as the founder of computer science. With the publication of two important papers, one on the limits of computation (what computers can and cannot do) published in 1936 and the other somewhat related paper on whether machines can be built that are capable of thought, published in 1950, he laid the theoretical groundwork for much of the work done in computational theory and artificial intelligence. 

Turing recognized that there is nothing special about arithmetic that allows it to be automated as opposed to logic, code-breaking, or any other task that can be described symbolically in a formal system. He was associated with the Colossus project where he worked at Bletchley Park, the heart of the British code-breaking effort, which helped to win the war. 

Turing's main interest, though, was in designing a computer for non-numerical calculations. He designed a machine called the Automatic Computing Engine (ACE) in 1945 which he intended to use for symbolic computations, but a working model wasn't constructed until 1950. 

In a lecture at his alma mater, Turing described computation as anything that can be described in symbols, and compared the behavior of machines manipulating numbers as not any different from those manipulating characters. He suggested that computers are able to learn by experience, and that as a result they don't really do only what they're told to do. He also explained a kind of game (now called the Turing Test) that would allow a computer to convince a person that it was intelligent.

The Machine That Changed The World Table of Contents  
Annotated Link Index   C SC 100 Page  

Comments, feedback, or inquiries, please contact DStucki@otterbein.edu



The Paperback Computer, Episode 3

Introduction

Scribes of the Middle Ages pioneered a medium that became the foundation for modern culture. Before the invention of the printing press access to books was confined to the few. Because of the labor-intensive process of copying books, they were priceless, often a single book being equivalent in value to an entire farm. 

Even after the arrival of the printing press, books were generally large and expensive. In the 16th century, libraries began to make the stacks accessible to the public, but only by chaining the books to the shelves to prevent theft. In order to become central to modern culture, books would have to lose their chains by becoming smaller and cheaper. Today the paperback book is mass produced and has become an ubiquitous part of society. 

In a similar historical progression, the early computers were controlled by an elite priesthood. They were enormous machines that were astronomically expensive, making them inaccessible to the masses. Just like the scribes, a rigorous education was a prerequisite to uncovering the "mysteries" of computation. In order to threaten the book as a rival medium of culture, the computer would also have to lose its "chains". 

Early Programming

Programming in the '50s and '60s was a tedious enterprise that involved encoding the program in a series of punched cards (often hundreds or thousands of them). This era is described as "bleak", and "dehumanizing". Programs were prepared on punched cards and then delivered to the computer to be run non-interactively in what became known as "batch processing". To dramatize the stress and frustration of interacting with the inflexible and temperamental machines, and the calamity of dropping a stack of sorted punched cards, some students at MIT produced a short film depicting a student ending his life in despair. 

The first significant example of interactive computing was Ivan Sutherland's 1963 Sketchpad program, for which he received an ACM award 30 years later (he also is one of the recipients of the prestigious Turing Award). Sketchpad allowed the user to draw geometric figures on the screen with a pen, and then manipulated the images. 

Doug Englebart also contributed to the transition to interactive graphical user interfaces (GUIs), with the invention of the mouse. Mitch Kapor, founder of Lotus Development Corp., says Englebart "...is the single most important person in the history of computing," despite the fact the Englebart is almost unheard of and is often overlooked in getting credit for inventing the mouse (Apple and/or Xerox-PARC are often mistakenly given credit). 

Xerox-PARC

In the late '60s Xerox Corporation recognized that the computer had the potential to create a paperless office. Not wanting to become obsolete, they set up a research center in Palo Alto, California (Xerox-PARC) in 1970 and charged it with the mission of becoming the "architect of the information age." PARC was funded for a ten year project whose goal was to bring the computer to the individual. 

The people working at Xerox-PARC realized that they were so used to the arcane computer interfaces that were then state of the art, that they would have difficulty understanding the mentality of ordinary people and ways in which they would find it easy to interact with a computer. To aid them in resolving this conundrum, they studied the ways in which children learn and interact with their environments. Borrowing ideas from the developmental psychologist Jean Piaget, they developed an interface where the focus was using the senses of touch & feel and vision to navigate, rather than on some kind of symbolic reasoning. This was the graphical user interface (GUI) in embryonic form. (For a list of where various aspects of modern GUIs were originally developed see the Unofficial History.) 

Much in the same way that video games create artificial realities (we all know that Donkey Kong is not real), the people at PARC created an artificial simulation of an office as an integral part of the user interface. Representing graphically on the screen the kinds of objects you would normally find on someone's desktop. 

The resulting system was called the Alto Office System (click here for a picture). It's cost of $45,000 per unit is a good indicator of why it was not marketed by Xerox and never became a consumer product. 

The Microprocessor Age

In the same year that Xerox-PARC was getting off the ground, a fledgling company by the name of Intel began work on the design of an integrated circuit chip that would become the 4004 microprocessor, the first microprocessor ever developed (picture; enlarged view). The 4-bit 4004 ran at 108 kHz and contained 2300 transistors. It could execute approximately 60,000 instructions per second. By comparison, Intel's Pentium Pro microprocessor, runs at 133 MHz, contains 5.5 million transistors, and executes 300 million instructions per second (MIPS)..The invention of a microprocessor that could be mass produced cheaply opened the floodgates of interest in computers. People who had previously been unable to afford the new technology were now forming hobbyist groups, creating startup companies, and holding tradeshows and conventions. 

In January of 1975 the Altair 8800 PC kit was unleashed on the market for less than $500. The first commercially available personal computer, it was an instant hit with hobbyists and computer enthusiasts. It was based on Intel's 8080 microprocessor, had 256 bytes of basic memory and was programmed by flipping toggle switches on its front panel. The Altair now has a dedicated virtual museum. 

Also in 1975, IBM released their attempt at a personal computer. The IBM 5100 cost anywhere from $5000 to $20,000, had only a 5" screen, and weighed 30-50 lbs. Compared to the Altair and other later competitors, the 5100 was a dismal failure. 

Apple Computer, Inc.

In the mid-70's, Steve Jobs and Steve Wozniak, who had met working at Hewlett-Packard and later were both members of the "Homebrew Computer Club," decided to build and market a personal computer. Wozniak was the hardware genius of the pair, while Jobs had visions of marketing and was a better manager than Wozniak. Thus, Apple Computer was born. The Apple II computer is said by some to have been the first affordable, useful computer. It was released in April of 1977 and was an instant success. 

In 1979 Jobs visited the Xerox-PARC facility and was amazed at the innovations they were making in user interfaces. Jobs said that when he returned to apple he knew that "[Apple's] future has just changed. This is where we have to go." He began work on a new computer called the Macintosh. It was the direct successor to the Lisa computer, was the first commercially successful personal computer to have a graphical user interface (GUI), and was released in January of 1984. Apple produced a television commercial (directed by Ridley Scott) to announce the Macintosh, that was only ever aired during the 1984 NFL Superbowl, which compared IBM to George Orwell's Big Brother in the novel 1984. 

In the meantime, IBM had released its personal computer (PC) in 1981. It had a disk operating system (DOS) and was comparatively hard to use. Steve Jobs observed that "the paradox was that to make a computer easier to use you need a more powerful computer to begin with." (For a complete history of Apple Computer, click here.) 

The Software Explosion

Once personal computers became widely available in the 1980s, the demand for software that would make them useful created a hotbed of opportunity for an entire new industry. Tens of thousands of software applications, development tools, and utilities were released in the first decade. 

Some of the major players that emerged were Lotus Development Corp., founded by Mitch Kapor, whose initial product was the first commercially successful spreadsheet application, 1-2-3, and Bill Gates' Microsoft Corporation, whose operating systems MS-DOS and Windows have become the de facto standard for Intel based PCs. Microsoft now employs more than 10,000 people and has annual revenues in the billions of dollars. 

One expert observed that the fact that we call them computers is an historical accident. What they really are is general purpose machines that can be turned to any purpose, given specific, unambiguous instructions for how to accomplish the task. It is this very process, however, that turns out to be rather difficult. Writing code correctly the first time is very hard and almost never happens. 

The factors that contribute to the difficulty of programming are 
  1. building in the abstract: we are manipulating ideas and concepts, not concrete, steel, and glass.
  2. complexity: many applications are more complex that even a Boeing 747.
  3. non-locality of error: an error in one part of a program may cause a performance error hundreds of lines of code away.

The New Generation

The personal computer revolution has produced, over the last 15 years, a whole new generation of computer literate children. Students have been introduced to computers at earlier stages of education every year, so that now many are familiar with basic operations of a PC by the time they leave kindergarten. Computers are also providing a mechanism for children with disabilities to gain access to mainstream education. 

In fact, what the current generation has discovered is that the computer is not a new kind of machine at all, but rather a new medium for communication, exchange of ideas, and the transmission of culture. They may not be as important yet as books, but they are still in chains. In order to lose these chains, they must continue to become cheaper and universally accessible, just as the paperback book has become. When we live in a world in which computers are taken completely for granted, then they will be fundamental to our culture. 

Virtual Interfaces

The real advantages that computers will contribute towards this end are the abilities and services that they provide that books are incapable of doing. One example of a computer technology that may provide an aspect of this new medium is virtual reality (for more information click here). Although computers are more interactive than books, they are still externalized, two-dimensional (2D) media. The idea behind virtual reality is to place the user in an artificial 3D world in a manner much more realistic to the senses. With specialized headsets equipped with stereoscopic video, sensory gloves that can track movement and gestures, and treadmills that allow locomotion, a virtual reality system allows the user to fully participate in the illusion. 

The transition from 2D to 3D is a significant change in how we interact with communication media. The 2D paradigm has existed for centuries. It remains to the creative research efforts of this generation to determine how a new paradigm might change the way in which we think, interact, and participate in a dynamic culture.


The Machine That Changed The World Table of Contents  
Annotated Link Index   C SC 100 Page  

Comments, feedback, or inquiries, please contact DStucki@otterbein.edu


 

The Thinking Machine, Episode 4

Introduction

How are people different from computers? Can computers think? The answers to these questions, although seemingly obvious to children, have taxed the minds of some of the greatest thinkers of the 20th century. 

Although computers were originally designed to serve as calculating machines, many of the earliest computer scientists and engineers thought of computers as thinking machines. As far back as the 1950's there was work on developing computer programs that could play checkers, or chess, or solve calculus problems. 

Thinking, however, is a mysterious activity. Scholars have often described thinking not as a process of the brain, but one of the mind. Pioneers of artificial intelligence (AI), following this philosophy, had little interest in how the brain was constructed. They saw the problem of artificial intelligence as that of simulating the mind, an endeavor entirely different from simulating the brain. 

A strong analogy was made, identifying the brain as the hardware (or wetware) upon which the software, the symbolic processor of the mind, was executed. This metaphor allowed researchers to take the position that the solution to the problem didn't necessarily have to mimic nature's solution, just as in the example of manned flight not involving airplanes that flap their wings to remain aloft. We could bypass the biology and evolution and directly engineer intelligence. 

The Golden Years

In 1958 Marvin Minsky and John McCarthy formed an AI department at MIT.One of their first students, Jim Slagle, developed a program that could solve college level calculus problems. The program consisted of less than several hundred rules governing symbolic integrals, yet it was able to get an 'A' on the MIT calculus exam. 

Hubert Dreyfus, a philosopher and AI critic, recalls students coming to him talking about philosophy having spent centuries on the issue of mind and making little progress, while in only a very short span of time they had practically finished understanding all of the aspects of cognition through the model of computation. He comments that although initially he had no idea what they were talking about, he soon became interested in investigating their claims for himself. 

In general, the decade of the 50's was one of optimism. The technological breakthrough of the computer was still so new that there were no visible limits to what could be accomplished with computation. Oliver Selfridge, a respected scientist, in an interview said that he thought that machines would be able to think, at least that they would be able to do things that we would describe as thinking when done by people. This optimism persisted, for the most part, until a scathing report was published by Sir James Lighthill in 1973 criticizing much of AI research. 

One of the early successes in AI was in the area of microworlds (also called toy problems). For example, computer systems that were designed to manipulate a collection of blocks on a table by means of a robotic arm. Or Edinburgh University's Freddy system that was able to recognize objects placed on a round board. These early systems faced serious problems that would continue to haunt AI for years. The block's world system lacked the common sense to build a stack from the bottom up. It had no knowledge of gravity. Freddy required over ten minutes to recognize a cup: a task a child could accomplish instantly.

Many researchers began to gravitate away from perceptual problems into a more abstract domain. The computational issues in areas such as vision were too immense to work with. An example of this is the Stanford Cart project from the early 70s, which attempted to navigate across a field of obstacles. It required a fifteen minute pause between each meter traversed. 

The Disembodied Mind

In 1950 Alan Turing wrote a paper describing what is now known as the Turing Test. The idea was that if a computer could be programmed to engage in conversation in such a way that it could fool someone into thinking it was human, then it must be intelligent. Early systems inspired by the Turing Test were much too brittle to be considered intelligent. Take, for example, Joseph Weizenbaum's program Eliza. Eliza used simple text processing and keyword matching techniques to simulate the kind of discourse common in psychotherapy. Because Eliza did not analyze more that the superficial surface structure of the sentences, it was easily unmasked as nothing more than a parlour trick. 

Without experiential common sense or general background knowledge of the world, a computer will always struggle to discern the appropriate context that it should bring to bear on understanding a given situation. Doug Lenat provides an example of ambiguity that must be resolved by general background knowledge in a story in which a girl sees a bicycle through a store window. We are told that she wants it. The question becomes how do we know that it refers to the bicycle and not the window or the store? 

Without the ability to apply contextual or general knowledge to a situation a computer will never be able to fill in the detail required to completely, or even adequately, understand a body of text. This became painfully obvious in the area of natural language translation. Although early predictions indicated that technology was on the verge of providing Russian-English translation systems that would accomodate the entire technical written output of the Soviet-Union in a few hours of processing a week, the actual results were much more disappointing. (One story told that a system that was developed for translation in both directions was tested by translating English->Russian->English and comparing the result with the original. The quote, "The spirit is willing, but the flesh is weak," came back as, "The wine is good, but the meat is spoiled.") 

One of the roles that common knowledge plays in helping us to understand natural language is to resolve its ubiquitous ambiguity. 

By the end of the 60s, it was becoming clear that the problems AI had tackled were much more complex than had at first been thought. Marvin Minsky observed that the things we have traditionally associated with intelligence, things we find to be difficult, are relatively easy to program computers to do, whereas those things that are so easy for us that we often take them for granted are exactly the tasks that are the most difficult to program computers to do. With the publication of Perceptrons by Minsky and Papert in 1969, What Computers Can't Do by Hubert Dreyfus in 1972, and the Lighthill report in 1973, the future of AI looked bleak and much of the funding of research was curtailed or killed. 

Expert Systems

Researchers did not give up, however. One of the significant advances during the 70s was the development by Terry Winograd of a system called SHRDLU. SHRDLU was a microworld system that was successful at using natural language to converse and follow instruction with respect to a block's world environment. Because it could only operate within a limited domain, people outside of AI were not impressed, mainly due to the brittleness of systems that operate in a narrow domain. Edward Feigenbaum, on the other hand, saw an opportunity to produce systems commercially that could serve as experts in selected domains: expert systems. He developed an expert system, DENDRAL, that was able to perform structural analysis of organic compounds based on mass spectrometry data. The commercial applications of expert systems helped to shore up AI research through the dark years of the 70s. 

Expert systems are comprised of the codified knowledge of human experts, who were shocked and dismayed to learn that they had devoted their professional lives to tasks that turned out to be nothing more than the application of a few hundred rules. 

Although expert systems provided a renewed opportunity for research, many critics also found a new opportunity for condemning AI. Hubert Dreyfus points out that although an expert system for blood disease diagnosis may be able to outperform some human professionals, it doesn't even understand what a germ is. Doug Lenat (not a critic, but one of the foremost proponents of AI) describes a system that grants auto loans to 19 year olds who claim to have had 20 years experience at their jobs and a system that diagnoses an old car to have measles. The message is that systems that have deep but narrow knowledge are extremely brittle, and can't really be considered to be intelligent. Similarities can be drawn between such systems and human idiot savantism. For example, some individuals can tell the day of the week for an arbitrary date, past or future, but are unable to do simple arithmetic. 

Commonsense Knowledge

Work also began in the 70s to address the brittleness AI systems. One approach was to build a collection of frames or scripts that captured the stereotypical features of a scenario or event. Then once a story was identified with the appropriate script or frame, the system would be able to fill in any details missing from the narrative. This approach worked fine until there were details missing that did not appear to be appropriate to any script or frame. Thus was identified the commonsense (or general background) knowledge problem. 

How do we accomodate discourse that assumes knowledge that is common to all people of all cultures, but that is so obvious to us that we never would have thought to tell the computer? Things like: children are never born before their parents, if you are running a marathon you are awake, or if you fly to Denver the other passengers on your flight also go to Denver. Marvin Minsky has estimated that there are approximately 10,000,000 items or units of general background knowledge, although Dreyfus claims that the number of items is practically unlimited.  

General background knowledge became the nemesis of AI as we entered the 80s, a severe obstacle to progress. In response, the focus of interest shifted from knowledge engineering to machine learning.  

CYC

Doug Lenat began CYC in 1984, a ten year project, with the goal of codifying all of the general background knowledge necessary for computer systems to intelligently engage in natural language communication. The idea behind CYC was that it would catalog information like an encyclopedia, but exactly that information that you would never find in any normal encyclopedia because it is so obvious and widely known. 

One of the principles upon which his project was based is that learning only occurs at the fringe of what you already know, and that the more you know the more and more quickly you can learn. His contention was that AI systems were generally too impoverished for any kind of sophisticated performance. 

There were, however, other approaches to machine learning that did not seem to be concerned as much with background knowledge. One of these approaches was based on the idea that common sense relies primarily on something no computer possesses: a human body. In other words, common sense is not based on facts, but on experience and skills. 

Connectionism

From the early 50s there have been ongoing attempts to build artificial brains. The computational models used are called neural networks, and the researchers who work with them are connectionists. A neural network learns by adjusting the strengths of connections between neurons, weighted patterns of neural discharges. An early neural network, the perceptron, was popular in the 60s. One application that was attempted with perceptrons was distinguishing male from female faces. Interest in neural network research dropped significantly in the 70s, after Minsky and Papert published their analysis of the weaknesses of the perceptron. 

Connectionism experienced a resurgence in the 80s with the development of more powerful neural networks based on the backpropagation learning algorithm, which overcame many of the weaknesses of the perceptron. One of the most successful projects was Dean Pomerleau's NAVLAB project at Carnegie Mellon University. In the summer of '94 a neural network navigated a vehicle "No Hands Across America" from Pittsburgh to San Diego. 

Neural nets, however, have problems of their own. There is no way to "look inside" a neural net to see what it has learned. Sometimes what they actually learn is not what you think they have (or what you would want them to). Hubert Dreyfus tells the story of a net that mistakenly discriminated photos based on which were taken on sunny or cloudy days, rather than the intended difference of the presence or absense of army tanks in the pictures. 

Another successful (and perhaps the most famous) neural net, Sejnowski's NetTALK, uses backpropagation to learn to convert text input into pronounced English audio signals. Initially the net makes a random attempt at pronunciation, but its output is compared to the correct output, and the difference (error) is fed back through the net (backpropagated) to effect a change in the net's connection pattern. NetTALK understands nothing of the text it pronounces! Researchers have determined that the brain is not a single, general purpose machine, but rather an integration of many specific modules: vision, language, etc. 

Future Prospects

Artificial Intelligence has developed many technologies that have been developed into commercial applications. Among these are chess playing programs, expert systems, airline scheduling systems, hospital robots, readers for the blind, and software for limited translation of Japanese into English. All of these, however, would fail the Turing Test. They don't achieve the quest for a general purpose intelligence. 

So is there any hope for the future? Some have suggested that the last best hope is the continuing work of the CYC project team. Doug Lenat claims that his prospects for the success of CYC are greater than two thirds (six times what he would have predicted in 1984). Hubert Dreyfus speculates that if CYC does not fulfil its promises, and never is able to understand the kind of stories understood by four year olds, that AI will be dead for however many hundreds of years it takes for us to figure out what intelligence really is. 


The Machine That Changed The World Table of Contents  
Annotated Link Index   C SC 100 Page  

Comments, feedback, or inquiries, please contact DStucki@otterbein.edu