There’s a quote that is credited to Arthur C. Clarke, the author of 2001: A Space Odyssey, that goes something like:
“Any teacher who can be replaced by a machine should be.”
I have to admit that when I first heard this quote I was completely in agreement with Clarke’s sentiments. I believed that teachers offered so much more than a computer could replicate. In my mind, teaching is fundamentally an enterprise built on forging interpersonal relationships. It requires human interaction. It requires individuals who are able to assess students’ needs and respond with targeted instruction. In Education and Experience, John Dewey captures it best when he writes, teachers must:
“be able to judge what attitudes are actually conducive to continued growth and what are detrimental. He must, in addition, have that sympathetic understanding of individuals as individuals which gives him an idea of what is actually going on in the minds of those who are learning.”
To do this, I believed, teachers needed to interact with the student with whom they work. After teaching online for the last decade, I began to believe that the human interaction required for teaching could be successfully mediated via machines (computers, learning management systems, etc.). But I still doubted that machines could replace teachers completely.
I’m not the only one. A few weeks ago, Education Week published an article detailing the use of artificial intelligence in teaching environments. The site had surveyed K-12 teachers about their beliefs about AI-powered teaching robots in their classrooms. The vast majority, 84 percent, didn’t believe that teaching robots would help improve student learning in their classroom. 90 percent did not think that student learning would improve in classrooms where chronically low-performing human teachers were replaced by artificially intelligent robots. The article then went on to examine different efforts that explored the use of teaching robots in academic settings. Take Jill Watson, the “teaching assistant” who was added into Ashok Goel’s computer science classes at Georgia Tech. While students sought assistance from Ms. Watson throughout the semester and even nominated her for an outstanding TA award, they didn’t realize she was actually an AI-powered chat bot created to help offer assistance in Dr. Goel’s classes.
I’ve been thinking a lot about the use of artificial intelligence in classrooms since I heard an episode of the 99% Invisible podcast last month. Titled “The ELIZA Effect,” the episode detailed the historical evolution of chat bots from the initial ELIZA program at MIT in the 1960’s to sophisticated OpenAI programs today. While the episode examined some of the technical aspects of chat bots, it also discussed the human impacts of their use. Take the initial ELIZA program. It was designed to simply parrot statements back at users in the form of questions, but people reported “having these really long kind of soul-bearing conversations with this program.” Despite its simplicity in communication, people found value with interacting with the program and some suggested that the program offered therapeutic benefits to users. And this troubled Joseph Weizenbaum, the computer scientist who created the ELIZA program. Weizenbaum was horrified that people could have meaningful conversations and interactions with a machine and he spent the remainder of his life advocating against the use of artificial intelligence for human interaction. Discussing Weizenbaum’s apprehensions with AI, Delaney Hall, one of the 99% Invisible producers, writes:
“Weizenbaum objected to the idea that something as subtle, intimate and human as therapy, could be reduced to code. He worried that if machines snuck into the therapist’s office, then where else might they end up? He began to argue that fields requiring human compassion and understanding just shouldn’t be automated.”
Which brings us back to teaching. As a field, teaching requires human compassion, understanding and interaction. And, as Weizenbaum predicted, artificial intelligence and chat bots have ended up in our field, too. But maybe that’s not a bad thing. Used as a complement to human teachers, AI-powered teaching bots might be able to help differentiate instruction or provide real-time feedback to students. Almost every teacher I know laments over the demands of the profession and how there’s just not enough time in the school day to individualize instruction or provide targeted feedback to support every student’s needs. Could artificial intelligence help to reduce those demands? Absolutely. Maybe not today, but definitely someday in the not too distant future. The robots are definitely coming.