Prompt:

Much is made of students using A.I. to “cheat” in their classes. However, some argue that A.I. could become a powerful tool to further education and make it accessible to the masses. Should we use A.I. tools like ChatGPT in education, or should we ban them? Why or why not? If you use the article below to support your answer, please make sure to cite it using MLA format.

Article:

A.I. Could Actually Be a Boon to Education

Peter Coy

May 3, 2023, NY Times Newsletter

Sal Khan is an uber-nerd, and that’s why people love him. In early 2020, to demonstrate that Khan Academy, his platform for free online education, was a worthy cause, he didn’t produce a brochure with smiling children and glowing testimonials. He made one of his teaching videos, with the usual black screen, colored pens and handwritten equations. The phrase “standard deviation” came up a lot. He multiplied the number of highly active users of his program by their average academic improvement, then linked that to the resulting increase in their projected lifetime earnings, compared it to operating costs and, presto, showed a benefit-cost ratio of 480 to 1, or around 240 to 1 if calculating only the benefits to students in high-need schools.

Those are crazy high benefit-cost ratios, considering that a lot of nonprofits would be happy with a ratio of 10 to 1. I’m not vouching for the calculation, though it seems reasonable at a glance. I’m citing it because Khan has a new plan to make the ratio, whatever it is, even higher. It’s all about harnessing artificial intelligence. I watched a TED Talk he gave on April 18 about his A.I. plans, and I interviewed him last week to get more information.

More on A.I. in a minute. First, I want to say that I admire Khan because he does beautifully what I try to do myself, which is to make complicated ideas understandable. I’ve watched dozens of his videos over the years to learn or relearn topics including statistics, calculus and science. He has a knack for walking through a problem patiently and developing an intuitive explanation that clicks. The payoff from watching one of his videos is that “aha” moment when you suddenly get the concept.

“My intellectual selfishness is that I really like making videos,” Khan told me. “It’s just fun. It’s beautiful.” Though Khan Academy has grown immensely since he founded it in 2008 — it has more than 150 million registered users and is available in more than 50 languages — he said he still tries to spend a third of his time on content creation. (Though he’s no longer the only person making videos.)

“My trick is to know the material really well,” Khan said. “I spend a lot of time pondering. I’m a professional ponderer. If I ponder something enough it will click. Then I have the confidence to just press record and start talking.” He added: “I’ve gotten into arguments with some team members. They’ll say, ‘You covered only six of the objectives. Not eight.’ But if the person gets the essence of it, they’re off to the races.”

Khan told me that he got an email last June from Sam Altman and Greg Brockman of OpenAI, the research and development company behind ChatGPT. They wanted to work with him to show that their A.I. engine could be a force for good. They invited him to take an early look at GPT-4, the powerful language model that wasn’t released to the public until this March. He went for a demo in July and was blown away. GPT-4 was far stronger than GPT-3.5, which itself was a knockout. GPT-3.5 powered the version of ChatGPT that caused a sensation when it was released in November.

“The technology we were seeing was mind-blowing,” Khan told me. “It was hard for me to keep it secret.”

Bill Gates had challenged OpenAI to build an artificial intelligence system that could pass the Advanced Placement biology test. So the OpenAI team used Khan Academy’s course materials to train GPT-4 in biology. By September, the A.I. was answering nearly every A.P. biology exam question correctly and explaining its answers. “It was a this-changes-everything type of moment,” Khan said. “This felt like it had a ton of potential.”

At the time and in the following months, when ChatGPT came out, most of the buzz around A.I. and education was negative. Teachers and parents worried that kids would use it to do their homework for them, including personalized essays. Khan saw it

Khan and his team used GPT-4 as the engine behind software called Khanmigo (“Khan” plus “amigo” — a little goofy).

Khanmigo isn’t supposed to give away answers. Like a good flesh-and-blood tutor, it engages students in a Socratic dialogue to guide them. It’s good at figuring out what they aren’t getting. Plus, Khan said, “it lets students do things you could never do before, like talk to literary or historical characters.” He chuckled about one girl who had a conversation with an A.I. incarnation of Jay Gatsby. The experience was so real to her that when she finished she apologized for taking so much of the character’s time. For safety, all interactions with students are recorded and shared with parents and teachers.

Khanmigo is being tested in the public schools of Newark, N.J., as well as in School City of Hobart in Indiana, a charter school in Phoenix, the Mountain View, Calif., campus of Khan Lab School and Khan World School, which offers classes online.

There have been some growing pains. For example, GPT-4 doesn’t understand what it spits out and is not naturally good at math. It can generate wrong answers, which it gives with supreme confidence. Khan said that he and Kristen DiCerbo, the chief learning officer of Khan Academy, and others spent two weeks giving Khanmigo guidance on the ideal way to respond as a math tutor. He said the tool is much better at math now.

Another concern: When I tried Khanmigo on finance, I found that when I played dumb and asked for hints, it pretty much gave me the answers. “That’s good feedback,” Khan told me. “We’re trying to thread the needle between giving people too much and not enough.”

By highlighting the potential for artificial intelligence to do good, as I have in a few newsletters, I don’t mean to dismiss the risks. I’m as scared as anyone when I read warnings from people like Geoffrey Hinton, who was an A.I. pioneer at Google, or Yuval Noah Harari, who wrote in The Economist last week: “We have just encountered an alien intelligence, here on Earth. We don’t know much about it, except that it might destroy our civilization.”

But I hope it’s possible to get the good from A.I. while seeking to throttle back the bad. Khan, for one, has been using Khanmigo as his own intellectual amigo. Recently, he told me, he used it to satisfy his curiosity about an aspect of supernovas. He said it’s better at helping his kids with their programming homework than he is, even though he has a master’s degree in computer science from the Massachusetts Institute of Technology. “It is eerily good,” he said.

Khan Academy has been a game changer for education. Khanmigo, Khan told me, “is a game changer for Khan Academy.”