Last November, when ChatGPT was released, many schools felt as if they had been hit by an asteroid.
In the middle of the school year, and without warning, teachers were forced to confront seemingly bizarre new technology that allowed students to write college-level essays, solve challenging problem sets and take standardized tests.
Some schools responded – unwisely, they claimed at the time – by banning ChatGPT and tools like it. But this ban didn’t work, in part because students could simply use the tools on their phones and home computers. And as the year went on, many schools quietly restricted the use of generative AI — as the category that includes ChatGPT, Bing, Bard, and other tools is called — to retreat ban them.
Prior to this school year, I spoke with several K-12 teachers, principals, and university faculty about their thoughts on AI Now. There is a lot of confusion and panic, but also a fair amount of curiosity and excitement. Teachers basically want to know: How do we actually use these things to help students learnInstead of just trying to catch them cheating?
I’m a technology columnist, not a teacher, and I don’t have all the answers, especially when it comes to the long-term effects of AI on education. But I can give some basic short-term advice to schools trying to figure out how to engage with generative AI this fall.
First, I encourage teachers—especially those who teach in high schools and colleges—to assume that 100 percent of their students use ChatGPT and other generative AI tools on every assignment, and in every subject, unless they are physically supervised within the school building.
In most schools, this will not be entirely true. Some students will not use AI because they have ethical concerns about it, because it is not useful for their specific tasks, because they lack access to tools, or because they fear getting caught.
But the assumption that everyone uses AI outside of the classroom may be closer to the truth than many educators realize. (“You have no idea how much we use ChatGPT,” read a Recent article Written by a Columbia University student in the Chronicle of Higher Education.) It’s a useful shorthand for educators trying to figure out how to adapt their teaching methods. Why should you assign a home test, or an essay on “Jane Eyre”, if everyone in the class–except perhaps the most die-hard rule-follower–is going to use AI to finish it? Why not move on to Monitored Tests, Blue Book Articles, and in-class group work, if you know that ChatGPT is as ubiquitous as Instagram and Snapchat among your students?
Second, schools must stop relying on AI detection software to catch students who cheat. There are dozens of such tools on the market now, all of them claiming to detect typing generated using artificial intelligence, and None of them work reliably well. They generate a lot False positivesIt can be easily fooled with techniques such as paraphrasing. Do not believe me? Ask OpenAI, the maker of ChatGPT, who I turned off the AI write detector This year due to “lower accuracy”.
It is possible that in the future AI companies may be able to label the output of their models to make them easier to detect — a practice known as “watermarking” — or better AI detection tools may emerge. But for now, most AI text should be considered undetectable, and schools should spend their time (and technology budgets) elsewhere.
My third piece of advice — and the one that might make me the angrier emails from teachers — is that teachers should focus less on warning students about the shortcomings of generative AI than on figuring out what the technology does well.
In the past year, many schools have tried to scare students off using AI by telling them that tools like ChatGPT are unreliable, prone to meaningless answers and generic-sounding prose. While these criticisms apply to early AI chatbots, they are less true of today’s upgraded models, and smart students are figuring out how to get better results by giving models more complex prompts.
As a result, students in many schools are racing ahead of their teachers when it comes to understanding what generative AI can do, if used properly. And warnings about flawed AI systems issued last year may ring hollow this year, now that GPT-4 is able to obtain Harvard passing scores.
Alex Cotran, CEO of the AI Education Project, a nonprofit that helps schools adopt AI, told me that educators need to spend some time using generative AI themselves to appreciate its usefulness — and how quickly it can improve.
“For most people, ChatGPT is still just a party hoax,” he said. “If you don’t really appreciate the depth of this tool, you won’t take all the other steps required.”
There are resources for educators who want to develop AI quickly. Mr. Katran’s organization has a number of organizations that focus on artificial intelligence Lesson plans Available to teachers, as with International Society for Technology in Education. Some teachers have also begun compiling recommendations for their peers, such as A website Prepared by faculty at Gettysburg College that provides practical advice on generative AI for professors.
However, in my experience, there is no substitute for practical experience. So I would advise teachers to start experimenting with ChatGPT and other generative AI tools themselves, with the goal of mastering the technology like many of their students already do.
My final piece of advice for schools blown away by generative AI is this: Treat this year – the first full school year in the post-ChatGPT era – as a learning experience, and don’t expect to get everything right.
There are many ways that AI can reshape the classroom. Ethan Molick, a professor at the Wharton School of the University of Pennsylvania, believes the technology will lead to more educators adopting a “flipped classroom” — where students learn material outside of class, and practice in class — which has the advantage of being more cheat-proof with AI. Other teachers I spoke with said they were experimenting with turning generative AI into a collaborator in the classroom, or a way for students to practice their skills at home with the help of an AI teacher.
Some of these experiments will not work. some will. that’s ok. We are all still adjusting to this strange new technology in our midst, and occasional stumbles are to be expected.
But students need guidance when it comes to generative AI, and schools that treat it as a passing fad — or an enemy to be overcome — will miss an opportunity to help them.
“A lot of things are going to break,” said Mr. Molik. “So we have to decide what we’re going to do, instead of fighting against AI.”