ChatGPT could be an effective and affordable tutor

(The Conversation is an independent and not-for-profit source for news, analysis, and commentary from academic experts.)

(THE TALK) Imagine a tutor who never tires, has access to massive amounts of data, and is free for everyone. In 1966, Stanford philosophy professor Patrick Suppes did just that when he made this prediction: One day, computer technology would advance so that “millions of schoolchildren” would have access to a personal tutor. He said the conditions were like young Prince Alexander the Great being tutored by Aristotle.

Now ChatGPT, a new chatbot with artificial intelligence and advanced conversational skills, could have the ability to become such a tutor. ChatGPT has accumulated massive amounts of data on a variety of topics and can pass graduate school exams. As a researcher studying how computers can be used to help people learn, I think ChatGPT can be used to help students excel academically. However, as it stands, ChatGPT shows an inability to focus on any specific task, let alone tutoring.

Philosophers, engineers, and artificial intelligence scholars envisioned using the computer as an “intelligent tutor” long before the Internet became a global commercial network in the 1990s. I believe the lessons from developing these early tutoring systems can provide insight into how students and educators can best use ChatGPT as tutors in the future.

Computers as tutors

Suppes – the Stanford philosophy professor – was a pioneer in a field called “computer-aided education”. He developed some of the earliest educational software. This software provided one-to-one computer tuition and resulted in students scoring better on test scores than those who did not use the program. I worked for Suppes from 2004 to 2012 developing software and other online programs.

Since then, experiments to build “smart tutors” to help students have fueled advances in artificial intelligence, social networking, and computing hardware. And today, ChatGPT’s ability to write essays, answer philosophical questions, and solve computer coding problems could finally achieve Suppes’ goal of truly personalized tutoring via computer.

Early versions of personalized learning

READ :  What is Chinchilla AI: Chatbot Language Model Rival from Deepmind to GPT-3

In 1972, a new personalized learning system called PLATO, for Programmed Logic for Automated Teaching Operations, debuted. It was the first widespread personalized learning system of its kind.

Developed by Don Bitzer, an electrical engineering professor at the University of Illinois, PLATO allowed up to 1,000 students to simultaneously log on to a mainframe computer. Each student could take various online courses in foreign languages, music, mathematics and many other subjects, receiving feedback on their work from the computer.

PLATO enabled students to reach the same level of performance as in face-to-face classes in less time. And most students preferred this type of teaching to sitting in a big lecture. However, the system was too expensive for many colleges and universities to use. Each computer terminal was marketed for over $8,000 — about $58,000 today — and schools were billed additional fees each time a student used the system. Nonetheless, PLATO’s success with students inspired a number of companies to develop software that provided a similar type of tutoring, including the College Curriculum Corporation, co-founded by Suppes.

Popular PC brands like Apple and Commodore touted the availability of educational software as a reason for families to invest in a home computer.

By 1985, researchers at Carnegie Mellon University were designing software using advances in artificial intelligence and cognitive psychology. They claimed that current technology had advanced to a level that made it possible to design computer systems to be as effective as human tutors. Although over 10,000 educational software was available at the time, much of it was of relatively low quality and did not provide real tutoring.

Although the more advanced designs of educational software developed at Carnegie Mellon allowed students to learn significantly more than students in traditional classrooms, they were not widely adopted in schools.

In the 1980s and 1990s, a school needed a significant number of expensive, powerful computer workstations so that students could use an intelligent tutor. Today’s computers are much more powerful and much cheaper.

And early intelligent tutors were used primarily in math, science, and engineering courses, limiting their appeal. The students could not ask the software programs a question. However, students could get feedback on answers to quizzes and tests.

READ :  Meet Ai-Da, the First Robot to Speak Before U.K. Parliament | Smart News

In 2001, the next generation of intelligent tutoring systems was able to hold conversations with students in written English. These systems, early chatbots, used advances in natural language processing to engage in dialogues with students about a range of topics. Each system was specially designed for a subject, e.g. B. Physics developed. Adding another subject such as English would require its own specialized tutoring system.

The next few years brought three developments that enabled the dream of computer-aided tutoring to take a leap forward. One was the availability of broadband internet, which was faster than the dial-up connections of the 1990s. The second was the lower cost of more powerful computers. The third was investment by tech companies and government funding for university research in artificial intelligence, leading to the development of what we now call chatbots.

The first chatbots

Up until 2007, early AI chatbots offered tutoring by guiding students through their answers to questions. Research has shown that these chatbots supported similar learning outcomes as human tutors. Despite this, the use of chatbot technology to provide tutoring has been primarily experimental.

In the 2010s, social networking and advances in artificial intelligence dramatically changed the way people interact with technology. Chatbots are integrated into many areas of our lives and are no longer experimental. People are now talking on their phones, with their cars and with their household appliances. Many people have virtual assistants at home. You can ask Alexa, Siri or Google for directions or answers to simple homework questions.

Today, ChatGPT has advanced capabilities that allow for more complex conversations and can provide one-on-one tutoring, much like Suppes predicted in 1966.

I asked ChatGPPT why it would be a good tutor and it replied “I am a good tutor for a student because I am able to understand and respond to a wide range of questions and prompts and I can provide detailed and precise information on this gives a wide range of topics. In addition, I am able to communicate this information in a clear and easy-to-understand manner, making it accessible to students of all levels of understanding. Being a machine learning model also allows me to provide a large amount of information quickly and efficiently, which can be especially helpful for students trying to learn a lot of new material in a short amount of time.”

READ :  Artificial Intelligence in the Global Cancer Diagnostics Market

In contrast, Siri said, “Hmm… I don’t have an answer for that. Is there anything else I can help you with?”

Recent research into the use of chatbots in college courses shows that the way chatbots are used is important in encouraging students to learn more deeply. For example, AI chatbots enabled significant improvements in learning in a graduate course on cloud computing at Carnegie Mellon. These learning gains occurred when these chatbots asked students to build on an existing argument or provide more information about a claim they were making. In this case, the chatbot asked the student a question and not the other way around.

Many educators are concerned that students will learn less with ChatGPT because it can be used to cheat on assignments and papers. Others are concerned about ChatGPT giving wrong answers or spreading misinformation.

However, the history and research of intelligent tutors shows that leveraging the right design to harness the power of chatbots like ChatGPT can bring deeper, personalized learning to almost anyone. For example, when people use ChatGPT to ask students questions that prompt them to revise or explain their work, students will achieve better learning outcomes. Since ChatGPT has access to far more knowledge than Aristotle ever had, it has great potential to tutor students to learn more than they would otherwise.

This article was republished by The Conversation under a Creative Commons license. Read the original article here: