Sal Khan, CEO of Khan Academy, delivered an inspiring TED Talk last spring foreseeing a revolutionary transformation in education through the integration of A.I. chatbots. Khan, whose nonprofit organization has provided online lessons to millions, expressed optimism about the potential of A.I. to serve as personalized tutors for every student worldwide.
In his widely-viewed talk, Khan’s predictions garnered support from influential tech figures, such as Sundar Pichai, Google’s CEO, who echoed the sentiment that A.I. tutors could offer powerful assistance to learners globally. Pichai, speaking on a Harvard Business Review podcast, emphasized the goal of providing universal access to A.I. tutoring, reinforcing Google’s commitment with the introduction of the Bard A.I. chatbot and a significant donation to Khan Academy.
Khan’s vision aligns with a longstanding Silicon Valley aspiration for automated teaching platforms that adapt lessons instantly to each student. Advocates argue that such systems could bridge educational gaps by delivering tailored instruction more efficiently than human teachers. While tech initiatives, including laptops for students and learning apps, have been championed in pursuit of this ideal, their impact on academic achievement remains inconclusive.
The emergence of generative A.I. tools like ChatGPT, capable of generating answers and crafting human-like book reports, has reignited enthusiasm for automated instruction. Platforms like Khan Academy and Duolingo have introduced A.I. chatbot tutors based on GPT-4, a large language model developed by OpenAI. Despite the growing interest, skeptics caution that there is insufficient evidence to support the belief that tutoring bots will genuinely enhance education.
Tech executives foresee a future where bot teachers, powered by A.I., can interact with and inspire individual students in a manner akin to beloved human educators. As A.I. technology continues to evolve, the debate over its efficacy in transforming education remains ongoing.
The White House appears to be fully supportive. In a recent executive order addressing artificial intelligence, President Biden instructed the government to harness the transformative potential of A.I. in education by developing resources to assist educators in implementing A.I.-enhanced educational tools, such as personalized tutoring in schools, as outlined in a White House fact sheet.
Despite this endorsement, some education researchers caution against the enthusiasm surrounding A.I.-assisted instruction. They highlight the fact that A.I. chatbots often fabricate information and could impart false knowledge to students. Incorporating A.I. tools extensively into education might inadvertently elevate unreliable sources to positions of authority in the classroom. Critics also argue that A.I. systems can exhibit bias and opacity, hindering teachers and students from understanding how chatbots generate their responses.
Generative A.I. tools, according to Ben Williamson, a chancellor’s fellow at the Centre for Research in Digital Education at the University of Edinburgh, may have detrimental or “degenerative” effects on student learning. Dr. Williamson emphasizes the lack of evidence supporting the claim that A.I. chatbots can deliver the promised benefits.
Another concern is that the hype surrounding unproven A.I. chatbot tutors could divert attention from more traditional, human-centered interventions, such as universal access to preschool, which have proven effective in increasing student graduation rates and college attendance.
Privacy and intellectual property issues also arise, with many large language models being trained on extensive datasets scraped from the internet without compensating creators. This poses a challenge for unionized teachers advocating for fair labor compensation. Additionally, there are worries that A.I. companies may exploit educators’ input and students’ comments for their own business purposes, such as improving their chatbots.
Randi Weingarten, president of the American Federation of Teachers, with over 1.7 million members, indicates that her union is collaborating with Congress on regulations to ensure the fairness and safety of A.I. tools. She emphasizes the need for educators to have a greater say in how technology is implemented in classrooms, aiming to harness the potential of A.I. while guarding against potential risks.
This isn’t the first time educational reformers have championed automated teaching tools. In the 1960s, advocates predicted that mechanical and electronic “teaching machines” programmed to quiz students on subjects like spelling or math would revolutionize education.