The use of artificial intelligence (AI) in education is currently one of the most hotly-debated topics in academia. What should be allowed? What should be used, by both teachers and students? Should it be a part of the curriculum? What would be the cost of using it? Of ignoring it? What would getting an education, or at least receiving a diploma, mean under such a change?
Emotions run high, it's too new and capital-poisoned for good data, and informed discussion is absent, caught in a Bermuda Triangle between self-aggrandizing AI companies and their fanboys, the predictable "No AI is good AI" absolutist reaction, and black-pilled pseudo-Cassandras disappearing into a sweet-smelling mist. Is there any reasonable voice out there?
Brave New Words: How AI Will Revolutionize Education (and Why That's a Good Thing) (2024) looks like any other book you might see in the "Current Events" section, but I was drawn to it because it was written by Salman Khan. Khan is the eponymous founder of Khan Academy, one of the greatest gifts to the world of K-12 education. Even almost two decades later, the charming original videos are golden examples of pedagogy, and the playlist of scratchy YouTube clips has become a full-fledged company educating millions of students around the world. They seem like one of the few genuine positive forces in tech. I had imagined that they, if anybody, could skirt the sleazier side of the EdTech world by virtue of starting as a guy just videochatting with his younger relatives instead of a scheme hatched by gathering of suits (or more likely, Patagonia fleeces) to "disrupt education." Maybe the guy who jump-started the whole online education movement would have something interesting to say about AI and education?
Unfortunately, I couldn't have been more wrong. Unless you are interested in learning about Khan's biographical details, which you can gather from just the Introduction and Conclusion, you shouldn't bother. In a better world, this would not have made it onto the shelves. But when trillions are being spent to establish an AI monopoly, the co-opting of prestige and debasement of value was almost inevitable.
A charitable reader would characterize Brave New Words as a book-length ad for Khan Academy's AI chatbot Khanmigo (conmigo, "with me"), originally built in 2023 based on early access to OpenAI's GPT-4, the first major advancement on 2020's GPT-3, which brought Large Language Models into the mainstream with ChatGPT in late 2022. But it is not the technical details that are concerning, it is the underlying philosophy, or rather the absence of thought behind it, that makes Khan seem more like any other YouTube viral sensation than the guy helping stadiums full of children around the world understand how to add fractions. I couldn't care less about what Mr. Beast has to say about education, and regrettably Sal Khan is on the fast track to join him.
The promise of AI for students is the always available, infinitely patient teacher. No educator, no matter how technophobic, is denying that such a thing would be a fantastic tool for learning. In Brave New Words, Khan traces the inspiration for his ideas on education back to his favorites from science fiction: A Young Lady's Illustrated Primer from The Diamond Age, the personal AI tutor Jane from Ender's Game, robots from Asimov's story "The Fun They Had" and the Foundation series. A not uncommon origin story for the sort of teenager who ends up at MIT. Khan has arguably done more than anyone else alive to turn such dreams into reality.
However, the main roadblocks in education are structural and systemic, not merely technical. The issue is not access to the teacher that never sleeps, it's society's response to it. If the effect of AI on education was that students engaged with the material more, teachers would not have an issue with it. But it is rarely used that way. Instead, users usually minimize that engagement, outsourcing their own thinking and learning to the machine. Fluency in using AI is touted as the newest essential skill for the workforce, but without developing the expertise needed for effective use, what is actually being accomplished? This is all passed over in silence.
Khan at least acknowledges the obvious backlash from the front-line teachers whose jobs are on the chopping block, but instead of addressing their very real concerns, he just keeps repeating "Isn't this awesome? Wouldn't this help you teach better? Would you like me break it down again?" for about 200 pages. At this point, you're probably better off talking to Khanmigo; at least you can ask it to deliver its answers in the style of Donald Duck and have a bit of a laugh as both you and your students head to the unemployment office.
Maybe it was naive to expect the leader of one revolution to have something interesting to say about the next. I guess fame, wealth, and success can insulate you from the realities in the trenches even for something as down-to-earth as explaining long division, even with access to resources that would been unfathomable at the turn of the century. If you're looking for some insight on the interaction of AI with education, go elsewhere. You're not going to find anything here that you won't find on a press release for the latest AI model.
In the conclusion of Brave New Words, Khan confesses that he had thought that he would be an AI researcher growing up, before heading down a more practical path. As such, I can understand why Khan speaks with the zeal of a new convert. All the good he's done is not negated even by a series of colossally bad takes and a stunning lack of foresight. But inexperience comes with a lack of discernment, and combined with childlike enthusiasm, you are perfect prey for scammers and grifters. Fear the G(r)eeks, even when bearing gifts.