Chatbots (mostly) won't change education

They're just language calculators.

In the early 1970s, the first handheld calculators became available in the United States. They were initially expensive ($400, unadjusted for inflation), but the prices quickly dropped, and the devices became commonplace. This is the way of consumer technology, the path already taken by DVD players, Blackberries, cell phones, iPods, iPads, iPhones, electric cars, and now, AI chatbots.

In 1972 the New York Times ran a delightful bit of analysis titled “Hand- held Calculators: Tool or Toy?” that suggests people might be doing their taxes in public, next to you, on the subway or something? I don’t know, check it out. It’s weird:

Look again. The person next to you, apparently holding a tiny radio in his hand, may actually be using a calculator. Rather than listening to rock or Bach, he could be figuring the commission on a sale, his income taxes, shopping costs, or, on Madison Avenue, how much a big lunch is going to cost a client.

Later that year the New York Times was writing a consumer guide for the holiday season that quoted a salesperson saying “a lot of parents are buying it as a gift for sons and daughters at school.” And I wonder what those children were using it for? Innocently checking their work, if I had to guess.

The article raises the spectre of educational issues:

Bruce R. Vogeli, a professor of mathematics at Teachers College. Columbia University, who has worked extensively with calculators, said that since schools still stress computation skills, a calculator might not be a wise present for a young pupil.

“Teachers will view them as a crutch, and this is an accurate view if you look at the curriculum,” Dr. Vogeli said, adding that eventually educators would have to revise their thinking to take the new hardware into account. Some day, he predicted, every student will have a calculator built into his desk, so that teachers can spend more time on mathematics concepts and less on computation.

So, has there been less of an emphasis on computation as a result? Let’s look at one last article, this time a pro-calculator opinion piece in Arithmetic Teacher from 1976.

He has all sorts of ideas for using calculators in the classroom, but he does have one word of warning:

I suggest one caution when using the calculator. The temptation is to have students use the calculator to check complex paper-and-pencil computation. Put yourself in the position of the student who has just spent five minutes dividing a three-digit divisor into a five-digit dividend and then checks the answer in a few seconds on the calculator. This student is convinced that paper-and-pencil computation is foolish…

The emphasis there is mine because, as far as I can tell, this skill is not really emphasized in the math curriculum any longer. At least, not with numbers of that size. And, I think, that makes a lot of sense, for precisely the reasons George Immerzeel suggests—a machine can do it better, and learning it has little additional educational value.

With all this in mind, let’s start making some predictions about what A.I. chatbots—which are essentially language calculators—will mean for school and teachers.

Prediction #1: There will be only marginal changes to what we teach

Hand-held calculators changed the curriculum, but only a little bit. I’d expect something like this to happen in the wake of chatbots as well.

The future we can expect is one where Microsoft and Google and everybody else have a magic button you can press that takes whatever you produce and turns it into a coherent email, memo, request, application, cover letter, etc. This is a world where skillful writing of that sort is just less important in the real world than it is today.

Perhaps also, your ability to write essays in school will be less useful as a signal that you’re ready for that kind of work.

That might cause us to change what we teach, a bit. The production of writing at length and in conventional forms may be somewhat less important, in the way that dividing 41899 by 243 is maybe less valuable than it once was. Simply summarizing texts may be parallel to that, I don’t know.

OK, yes, but you’ll still need to be able to read and think coherently. Writing—along with computation thinking—is still important, maybe even essential. It has pedagogical value across the curriculum. Writing is thinking. It’s not going anywhere.

It’s just that math teachers won’t be the only ones fielding “when are we ever going to use this” in class.

Prediction #2: There will be some fun fights

That doesn’t mean some people won’t argue about it.

Does our ability to plug any arithmetic question into a computer mean that calculation as a whole is less valuable? Not really, but some say this. There are even districts or Thought Leaders or administrators or teachers who say it. And some people are going to say this about writing as well.

There will be op-eds. “Why I Stopped Assigning Essays,” then “Writing Is Becoming a Lost Art,” followed by “Chatbot Helped Me Graduate College,” “ChatGPT Wrote This Op-Ed,” and so on and so on, until we’re sick of the whole thing. And even then.

Here’s another little micro-prediction…

Prediction #3: Basically nothing will change for younger students

With young kids if you don’t want them to use a tool you…simply don’t give it to them. That goes for teachers and for parents. And there’s no reason why these students—unless they are suffering from a language-based learning disability that makes it difficult to share their thinking—should have access to a language calculator when they are writing sentences and paragraphs.

That’s pretty much the situation with hand-held calculators in math. Kids only start to touch them around the pre-teen years, and for most it only really gets started with high school math, when arithmetic is no longer the goal.

(By the way, Wolfram Alpha can solve all of high school math through calculus, no problem. It’s also worth thinking about how that has or hasn’t impacted the teaching of math.)

The fun here only really starts as kids get older and teachers begin to assign substantial writing assignments that are supposed to be completed out of class.

Prediction #4: Older kids will have to do more writing in class

I’m not an English teacher, but I’ve always been a bit anxious about out-of-class work. I get why it’s helpful for some kids, but still. You can’t help kids with work that happens out of the classroom. You can’t observe them. You can’t offer feedback. You can’t be sure that the work is their own. That’s always true, and that’s especially true when a computer program can pump out plausible-looking prose.

This isn’t the death of the essay, but it will I think put pressure on teachers to assign work differently. And, more importantly, to assess students differently. The logic of term papers never really made sense to me as a way to assess a person’s learning, and I think chatbots only weaken the case further.

Assessment will have to focus on in-class assignments, which as a math teacher, seems absolutely fine to me. I wouldn’t make 30% of a student’s grade depend on a take-home test. (Though, yeah, some math teachers would.)

The big challenge here is for college professors, especially in online courses. Which leads to my most tenuous prediction.

Prediction #5: Online coursework will become less of a thing

At a time when work from home seems posed to become very much a thing, it feels tenous to predict that school from home will take a hit. But I don’t see how you can assess students in any way at all with these tools available.

Now, if I’m wrong, it’s because as institutions, colleges don’t care about that. They care about customers, and I don’t blame them: the demographics are not on their side, the environment is not friendly, and it’s going to be at least a rough couple of decades for higher education.

But online education where a piece of software could literally pass the course feels like a step too far. Will customers want that? I don’t know. But I bet that they won’t.

So, in short, the picture looks like this:

• Small changes in what is valuable to teach

• Overall stability in what we teach and how we teach it

• The biggest changes are to what we’ll assign as homework, especially for older students

• The biggest shift will be towards briefer learning/assessment work that can be done in the presence of a teacher

And none of that, I think, is necessarily a bad thing.