Do you remember the No Fear Shakespeare books?
They would have Shakespeare’s language on the left, and modern English on the right. A reader could easily go back and forth, reading Shakespeare’s play and referring to the other page when they were stumped. They could use the modern English version to supplement their understanding of Renaissance English.
But did they?
If you were like me, your eyes seldom left the right page. You’d read Hamlet or Macbeth in modern English. Once in a while, you’d glance at the left page to see what you were missing. Turns out, not much. Just some fancy, barely recognizable words. Probably not important.
And how could I resist reading the modern English version when it’s right there, staring me in the face? Time and time again, as a young reader, I would choose the modern English over the Renaissance English. It wasn’t even that Shakespeare was too difficult for me.
Faced with a challenging version and a relatively easy version–juxtaposed for my convenience– I couldn’t help but choose the latter.
Most humans will choose the path of least resistance.
But here’s a question that would occur to me later: if I only read the modern English version of Hamlet, did I actually read Hamlet?
The Button
Ethan Mollick calls it simply, “The Button.”
Open Google Docs or Gmail. You’ll see a “Help Me Write” button.
Go into Grammarly. You’ll see a “Generative AI” button. Click on that, and you’ll open up a chat window where you can (1) generate almost anything you want and (2) hit a button to insert it right onto your document.
Wham, bam, thank you Sam! Writing has never been easier.
The Pedagogy Behind Khanmigo
Now, reading hasn’t caught up with that…yet. But it’s getting there.
I mean, Sal Khan presented this reveal of the Khanmigo bot teaching children how to read almost a year ago:
Seems cool, right?
But let’s hit pause. Here’s the moment he highlights. A young reader has trouble understanding why Jay Gatsby keeps looking at the green light.
Thank God, she tells herself. I have Khanmigo. So, I can just ask Jay Gatsby.
What’s the pedagogy here? It boils down to a few major principles:
Principle #1: You can ask a character for a definitive reading.
Principle #2: Chatting with a character supports the reading process.
Both are fundamentally flawed. Of course, Jay Gatsby wouldn’t answer this way. If you asked me why I light to look at the sunset and I went on a long diatribe about how it represents the end of my life, you’d roll your eyes and walk away. Wouldn’t you?
Characters don’t see the symbolism of their actions. That’s part of the point. We’re aware of things that they are not.
(Sidenote: I hate symbolic readings of colors in novels. This one is no exception.)
It appears to run a subpar interpretation (which is easily found on Sparknotes or Study.com.
Wait, actually. Hold my beer.
I just got back from Sparknotes. Here’s what that site said about Gatsby and his green light:
In its largest sense, then, the green light represents the American Dream. The American Dream is the idea that someone from a lower-class background can work hard and move up the social ladder because American society has historically had more class mobility than other countries. The novel explores whether the promise of the American Dream is actually true. On the surface, Gatsby appears to have achieved the American Dream, because he has managed to move from a lower-class background into the highest echelons of New York society, entirely through his own self-invention. In reality, though, Gatsby illustrates the hollowness of the American Dream, because even once he has accomplished this goal, he still is unable to attain Daisy, who represents a traditional elite background. Tom consistently mocks Gatsby for his humble beginnings, calling him a “common swindler who’d have to steal the ring he put on (Daisy’s) finger.” This not only implies that the American Dream is ultimately unfulfilling, but it also suggests that despite the illusion of social mobility, people from the lower classes will never be fully accepted by those who were born into wealth.
Look familiar?
What we have here is an AI program that passes off an easily available (and need I say again, subpar) interpretation of Gatsby’s green light as the authoritative interpretation of Gatsby himself.
I dare say that Khanmigo’s answer isn't intelligent (artificial or otherwise) at all. It's a conventional answer that's easily findable in an Internet search.
Of course, a struggling student will take that as authoritative. They might even do it at the disservice of their own idea (which may have been just hiding in the background, waiting to get out).
“The Button” will be huge for education, and not entirely in a good way. Our students, I suspect, will constantly be surrounded by chatbots that—sycophantic as they are—are there ready to help in whatever way they can.
In some ways, that's great. Students always have a support system.
In other ways, the prevalence of AI support (which apparently needs to be worked into every little thing) could be catastrophic for reading.
Sometimes, we need to just be alone with our thoughts.
What’s My Point?
Although I leaned into a video released more than a year ago, the errors behind Khanmigo’s reading tutor tell us a lot about AI programs and reading.
They make the same error again and again.
They suggest that reading is about content: it’s about identifying the plot, understanding something about the characters, etc.
They’re squarely on the right side of No Fear Shakespeare.
When you’re reading a tough novel and that chat window shows up on the right side asking if it can help, what would you do?
I mean, I’d click on that window again and again, rather than sitting there to struggle with the reading.
If it’s there, I’ll click on it.
At some point, someone will take it away from me. At some point, someone will take the training wheels off.
What do I do then?
https://terryu.substack.com/p/teaching-high-school-students-about
I’d ask you to dig a little deeper into the problem. Motivation to read can’t be required. Putting readers in a system where they are required to jump from Hamlet to Jay Gatsby, make annotations for evidence to use in support of a claim, giving them the blessing of a grading system that holds their life in its hands—and then bringing in AI… What do you expect? By the way, they are taught to read literature as though it were a newspaper article or a biology textbook. The problem isn’t AI, Sparknotes on steroids. The problem is instructional culture. AI is an innocent bystander. I’ve talked to kids who tell me it helps them decipher the indecipherable. It helps them dish up what the teachers value.