While the claims of early visionaries that AI would soon surpass human intelligence (expressed as the Turing Test) now seem grandiose, and the field itself has gone through several boom-and-bust cycles, a half-century of AI research nonetheless has had many practical applications in all areas of modern life, education included. In education, some of these applications include voice recognition, handwriting recognition, intelligent tutoring systems, writing assessment and computer supported collaborative learning using simulated students. While AI in education is not likely to replace classroom teachers, it will inevitably refine their role. When accepted by educators, AI can help make the American classrooms of the future more dynamic, learner-focused and creatively stimulating.
Technology in Education
The term artificial intelligence (AI) wasn't coined until 1956, when a small group of computer science researchers gathered at a conference on the campus of Dartmouth College in New Hampshire to discuss how they could use computer technology to change the world for the better by making machines that think. One of those researchers, John McCarthy (2007), defined artificial intelligence this way:
It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biological. (McCarthy, 2007)
With the expectation that machines would quickly out-think humans, these pioneering AI researchers attracted both attention and large sums of money. When they failed to deliver the promised results, both their fame and their funding took a hit. This cycle of promise and failure has repeated itself several times since the 1950s, and there are even seasonal metaphors attached to each extreme point of the cycle: AI summer and AI winter.
Despite past disappointments, AI researchers remain undeterred in pursuing their goal of creating a "human" machine. Their goal is to produce a machine that can pass the Turing Test, a test first articulated by Alan Turing in his 1950 paper "Computing Machinery and Intelligence." The Turing Test involves a robot and a human having a conversation, with both looking human; the robot passes the test if the human cannot detect that the robot isn't another human. So far no computer has passed the test.
The history of AI shows that researchers have made progress in the narrower goal of creating "intelligent computer programs," even if the larger goal of creating human-like machines has thus far proven elusive. Such computer programs include as IBM's Deep Blue, the computer chess program that defeated former world chess champion Gary Kasparov in the 1990s.
AI Roots in Current Applications
Many ideas that germinated in the labs of AI researchers have now seen the light of day (Kurzweil, 2006), even if AI itself doesn't get the credit:
Ironically, in some ways, AI was a victim of its own success. Whenever an apparently mundane problem was solved, such as building a system that could land an aircraft unattended, or read handwritten postcodes to speed mail sorting, the problem was deemed not to have been AI in the first place The effect of repeatedly moving the goal-posts in this way was that AI came to refer to bluesky research that was still years away from commercialization ("AI by another name," 2002, para. 6).
Other applications of AI that most of us use everyday include Bayesian filters that scan incoming e-mail to identify and delete spam before it reaches our inbox. There's even a whole industry dedicated to using AI to make better video games, and many popular titles now make use of at least some AI technology. As Rodney Brooks, director of the MIT AI Lab, puts it,
A.I. is everywhere around you every second of the day. People just don't notice it. You've got A.I. systems in cars, tuning the parameters of the fuel injection systems. When you land in an airplane, your gate gets chosen by an A.I. scheduling system. Every time you use a piece of Microsoft software, you've got an A.I. system trying to figure out what you're doing, like writing a letter, and it does a pretty damned good job. Every time you see a movie with computer-generated characters, they're all little A.I. characters behaving as a group. Every time you play a video game, you're playing against an A.I. (cited in Kurzweil, 2006).
Many applications of AI are finding their way into K-12 classrooms across the United States. These include:
* Voice recognition,
* Handwriting recognition,
* Intelligent tutoring systems,
* Writing assessment and
* Computer supported collaborative learning.
Voice recognition software has become an important tool for special education students with developmental disabilities which prohibit them from using standard computing methods. The software works by allowing a student to speak commands for a computer to perform -- anything from opening a file to typing a paper -- rather than the student having to type on a keyboard or use a mouse. Voice recognition software operates on any modern computer by using a microphone as the input device.
In order to train the computer to recognize the student's commands, the student must use discrete speech, which, unlike conversational speech, consists mainly of individual words and phrases. Using discrete speech, the student speaks words or phrases that correspond to a specific action to be performed by the computer. The software attempts to compensate for "mishearings" and mistakes by using a list of similar words or asking users to spell out words.
The cost of voice recognition software has plummeted since the late 1980s, when it cost approximately $9,000 ("Voice recognition," 2003). In 2013 commercial voice recognition software packages cost less than $100 per user. Various open source voice recognition software projects, which can run on multiple computer platforms, also are available. In 2004, IBM open sourced the code for their speech recognition software, making it available to anyone desiring to produce their own free speech recognition products (Lohr, 2004).
The bridge between handwriting recognition and AI is the software that the interprets handwriting put down on an electronic tablet with a special stylus or pen. As the student writes on the tablet that is attached to a computer, the computer reads the handwriting and transcribes the words into text like what you're now reading on the screen. From a pedagogical standpoint, there are several reasons why teachers would want to use such handwriting recognition software. Some research seems to indicate that teaching children handwriting at an early age helps them to better form letters, better understand the concept behind a given word and become better spellers (Read, MacFarlane, & Horton, 2005).
If this research on the link between handwriting and educational progress stands up, handwriting analysis software can provide another tool for teachers to use in helping students see how their handwriting translates into printed text. As the handwriting software provides feedback, students can identify and then learn from their mistakes. This can help both their handwriting and spelling skills, to say nothing of other intangible benefits such as teaching the value of perseverance.
Intelligent Tutoring Systems
Intelligent tutoring systems (ITS) have evolved out of older systems from the 1980s known as computer-based training (CBT) and computer-aided instruction (CAI). These programs enabled students to practice answering questions in various subjects, and they could even serve up a rudimentary sequence of questions based upon student answers, but they suffered from an inability to provide individualized attention. Each ITS consists of five basic components:
* Student model: Stores student information
* Pedagogical model: Tutoring techniques/methodology
* Domain knowledge: Repository of information to be taught and learned
* Communications module: Determines how the information is presented
* Expert module: Identifies areas where the student needs more help (Beck, Stern, & Haugsjaa, 1996).
Students use ITS to solve problems chosen by the system. Once the student provides his or her answer, the ITS compares the answer to its domain knowledge and then makes an assessment. This assessment will then let the computer adjust, based on the given pedagogical model, the next problem for the student to solve. In the end, the student is given problems and feedback that will...
(The entire section is 3912 words.)