Information Age Education
   Issue Number 34
January, 2010   

This free Information Age Education Newsletter is written by David Moursund and Bob Sylwester, and produced by Ken Loge. For more information, see the end of this newsletter.

“Problems cannot be solved by the same level of thinking that created them.” (Albert Einstein; German-born theoretical physicist and 1921 Nobel Prize winner; 1879–1955.)

The very first issue of this newsletter explored the idea that a human’s brain and a computer’s brain are quite different. They have different capabilities and limitations. One has “human” intelligence, while the other has been designed and programmed to have “artificial intelligence.” See http://i-a-e.org/newsletters/IAE-Newsletter-2008-01.html.

The previous issue of the IAE Newsletter (http://i-a-e.org/newsletters/IAE-Newsletter-2010-33.html) provided a definition of human intelligence that included creativity as one of its components. It argued that an intact human brain is naturally intelligent, curious, and creative.

This issue of the newsletter explores the topic of creativity by artificially intelligent computer systems. Keep in mind the idea that human intelligence and creativity may well be quite a bit different than computer intelligence and creativity—but both are of value to us.


The Turing Test

Albert Einstein is remembered as a very intelligent creative genius who made major contributions in Physics. The Einstein quote given at the beginning of this newsletter suggests that creativity is used in both posing and solving problems, and that these activities require thinking. The previous issue of this newsletter argued that ordinary people routinely carry out these three activities in their oral communications with other people.

Alan Turing is remembered as a very intelligent and creative genius who made major contributions in Computer and Information Science. One of his contributions was a 1950 “think piece” in which he proposed a test that could be used to determine if a computer system was intelligent. See http://en.wikipedia.org/wiki/Turing_test. Stated in simple terms, the Turing Test is whether a computer system can carry on a written conversation with a human well enough so that the human cannot tell whether s/he is communicating with another human or with a computer.

As progress has occurred in developing computer systems that can use speech input and can produce speech output, people have also stated the Turing Test in terms of an oral conversation over a cell phone.

In 1950, Turing predicted that by the year 2000 artificially intelligent computer systems would meet the criteria presented in his paper. Computer hardware technology has progressed much faster than Turing might have predicted. Artificial intelligence has developed as a major area of study and research. However, we still do not have a computer system that meets Turing’s criteria for artificial intelligence.

An annual contest has been held since 1991 to select the best of current computer systems designed to pass Turing’s test. The Loebner Prize (see http://en.wikipedia.org/wiki/Loebner_Prize) is awarded each year to the best performing entry in this contest. The entries are now commonly called chatterbots. Some of the past entries are available for free use on the Web. You may enjoy “chatting’ with a computer. You can access some chatterbots from the Website http://www.simonlaven.com/.


Standard Argument Against Computer Creativity

Many people argue that a computer system cannot be intelligent or creative. A computer is a machine that is designed by and programmed by people. The computer merely does what it is built to do and what its programmers and users tell it to do. There is no intelligence in merely following a detailed set of instructions rapidly, accurately, and in a non-thinking and non-understanding manner.

The argument is, if a computer system eventually passes the Turing Test, it will be because humans have developed sufficiently fast computer systems and sufficiently capable programs so that the “dumb” machine can pass the test.

One type of counter argument is that a human brain is “built” by biological processes and programmed by its ancestral genes, and by informal and formal education and experiences.

To some extent then, creativity simply means that a creative person/machine can imagine or has learned more alternatives than the non-creative, and can make connections between elements that normal folks would consider implausible, if they even considered them. We can improve our informal and formal education systems to help make people more intelligent and creative, and we can improve computer hardware and software systems to make computers more intelligent and creative.


Some Examples of AI Successes

The field of artificial intelligence (AI) has had its ups and downs over the past decades. For example, it was considered a success when a computer program defeated the reigning human world champion chess player in 1997.

It turns out that most humans are reasonably intelligent and creative, and thus can do things that AI researchers find very difficult to replicate. Take speech input for example. Consider a noisy room that has several simultaneous conversations in addition to yours. Add in the possibility that your hearing is not as good as it once was.

People are able to deal with this oral communication problem by drawing on their knowledge of the topic under consideration, their knowledge of the person they are conversing with, the filtering capabilities of their ear/brain sound processing system, their ability to read gestures and facial expressions, and so on. They process the input they receive in a sense-making manner—in essence, filling in the gaps in a manner that produces a received message that they understand and that makes sense to them, and filtering out extraneous information.

Researchers in AI have produced voice input systems that can convert spoken voice into written text. Under very good conditions (not like the noisy room described above), an accuracy of above 95% may be achieved. Some of the speech input systems are self-taught and can continue to learn through experience. That is, computer programs have been developed that learn to translate speech into text through listening to a very large amount of spoken text, comparing its own written decoding with the actual written text, and changing its own programming to produce a more accurate decoding.

Somewhat similar uses of machine learning have been quite successful in developing computer systems that can analyze and detect patterns in photographs, x-rays, brain scans, and output from an optical or electron microscope. As a simple example, many people now make use of computer software that has face recognition capabilities. You provide the software a digitized photograph that contains Aunt Millie, and then the computer system processes all of your digitized photographs to find other pictures that contain Aunt Millie. See, http://www.technologyreview.com/computing/22234/?a=f.

When you use a Web search engine such as Google, you are making use of a quite sophisticated AI system. The system accepts your written input describing the information you are looking for, and it produces as output responses that it “thinks” might fit your needs. The Wolfram Alpha search engine is able to respond to requests for factual information and is gradually growing better at this task. See http://www.wolframalpha.com/. The “growing better” comes from a combination of both humans and the computer system itself working individually and together to improve the system.

There are many examples of computer systems producing “original” works of art, music, poetry, stories, math proofs, and so on. See, for example, Douglass Hofstadter at http://en.wikipedia.org/wiki/Douglas_Hofstadter. His 1979 book Gödel, Escher, Bach: an Eternal Golden Braid, was awarded the 1980 Pulitzer Prize for general non-fiction.

You might enjoy reading about an autonomous science lab that uses computers, robots, and other lab equipment to carry out experiments and analyze the results. In keeping with the chatterbot terminology mentioned earlier in this article, such systems are called Lab-Bots. (See http://www.scientificamerican.com/article.cfm?id=robots-adam-and-eve-ai.)

Final Remarks

A computer system is “merely” a tool. This brain has capabilities and limitations, just as a human brain has capabilities and limitations. A human brain is created through biological processes and programmed through informal and formal education and experiences. A computer brain is created through manufacturing processes that now use highly computerized machines. It is programmed by human programmers, but may also be programmed to learn on its own.

From an educational point of view, we now have computer systems with a certain type of intelligence and a certain type of creativity that humans find useful. It is now common for humans to pose problems, ask questions, and create tasks, and then make use of computer systems to help deal with these situations.

The capabilities—including intelligence and creativity—of computer systems will continue to grow through the work of many thousands of computer designers, engineers, and programmers. A good education system prepares students to make effective use of their own intelligence and creativity, and the steadily growing intelligence and creativity of computer systems.

So perhaps the issue posed by this article is related to the kind of issue that genetics historically posed—whether nature or nurture is more important in determining biological behavior and destiny. Scientists now consider it a non-issue, since both nature and nurture are required to meet the challenges we confront. We can’t function without both.

Contemporary humans have thus similarly added computers as an external dry brain to our internal wet brain, and it’s now difficult to imagine successful life without both of them. We can thus take credit for the intelligence and creativity that added computerized processing systems to our biological processing systems—and the resulting computerized information storage and retrieval concept can take credit for incorporating a huge reference library into something small enough to carry in our pocket.


About Information Age Education, Inc.

Information Age Education is a non-profit organization dedicated to improving education for learners of all ages throughout the world. IAE is a project of the Science Factory, a 501(c)(3) science and technology museum located in Eugene, Oregon. Current IAE activities include a Wiki with address http://IAE-pedia.org, a Website containing free books and articles at http://I-A-E.org, and the free newsletter you are now reading.

To subscribe to this twice-a-month free newsletter and to see back issues, go to http://i-a-e.org/iae-newsletter.html. To change your address or cancel your subscription, click on the “Manage your Subscription” link at the bottom of this e-mail message.