Magic School AI includes the first chapter of The Great Gatsby as the exemplar activity within their Text Leveler tool. You can rewrite the text at a first-grade reading level. Many of the high school teachers were shocked that the tech could do this and wanted to know why an edtech company like Magic School would use a text like Gatsby as an example. I think it's clear they don't care what teachers think. Screenshot included in a Google Doc.https://docs.google.com/document/d/19b8_GU9uZOLv6sFNLFSqsAkSReaKb6RbD_v_gUHeSCI/edit?usp=sharing
This screen shot exemplifies and summarizes all the key points in article. For me, this is not dissimilar to graphic novels. While I've supported my studentreading graphic novels that are original content, I strongly discourage them from reading a graphic novel 'version' of a classic book.
To be fair, it’s just a new way of producing what has been done for decades. Abridged publications exist for many, many classics. Teachers have spent hours adapting extracts from complex texts with simplified sentence structures and vocabulary to enable pupils with learning needs and disabilities to at least begin to be able to access the texts. Side-by-side modern & original Shakespeare is a necessity in the classroom. It’s not one OR the other, it’s always both. We use adapted texts to help pupils into the originals.
It’s like you or I tackling Shakespeare, Middle English or Anglo Saxon text - we do not understand them immediately and need some help to get in. Then we can appreciate the beauty etc.
If AI is saving hours upon hours of teachers adapting and “translating” complex language for their pupils so that they can then get on to appreciate the original, that’s a good thing, no?
The days of halcyon immersive reading amongst the tulips just doesn’t exist any more. Humans, not just young people, are increasingly screen-addicted to the endless scroll on demand. We can make reading more attractive by making it accessible, but perhaps more importantly by divorcing it from grades and ultilitarian education systems.
I do hear you on (1) the history of abridged versions (I read an abridged version of Little Women that way, and as Rebekah said, was disappointed when I realized it wasn’t the original 🤣), and (2) having some type of translation to support students reading Shakespeare. I just still have doubts about whether AI translations/adaptations are the best way to do that. I almost think an edition curated with footnotes to help students understand the meaning/context of the language (and therefore be able to decipher it) might be more helpful than using an AI version. I do worry about comprehension development, too— giving students some semblance of the plot is okay, but if they’re reading an AI version, even if the original is there, they’re not really building syntactical or vocabulary knowledge. What do you think? Do you think there are alternative ways to balance students’ needs/ meeting them where they are with the cognitive challenge of deciphering a more advanced texts?
That’s a good point about the flattening of vocab by AI adapted texts - a human teacher would probably take more care over each sentence and make sentence and lexical level decisions for their class as they parse the text to be adapted.
That said, obviously an LLM can do this too, with careful prompting. You can tell it which sentence structures and words to keep. You can also far more quickly create a wider variety of versions of the same text at a range of levels.
The balance to be struck is time vs detail I guess.
Do I spend an hour personally adapting a text at maybe two different levels at most? Or, 10-20 minutes working with an AI to produce 3 or more differentiated levels, plus a vocabulary list, plus translations into other languages for my EAL learners, plus quiz questions to check for understanding, which I then check over and tweak?
Also, one other question, if you will indulge me!! What are the benefits/trade-offs of teaching multiple texts at once? So, let’s say you’re teaching 1984, but it’s too advanced for some in your classroom. Is it ever the case where one would assign 1984 to some students, and then let’s say, a text with similar themes that is easier to read, like The Hunger Games, to others? I remember them doing that in my elementary school, but I’m wondering how widespread text choice/multiple texts are.
Good question! Certainly one could do this with extracts in a one off lesson or two but it would be very tricky to teach different whole novels at the same time to different groups in the class over a longer period without compromising quality & depth. Also, the difference in difficulty levels even with extracts would need to be handled carefully such that groups don’t feel stigmatised for being given an “easier” text.
As a proficient reader at 5, I never understood the need for simplifying texts until I was tutoring. Today, the side by side modern and original basically is doing what a private tutor will do, explain the meaning in today's terms. That said, the original is still read, to expand a student's education. However, as you point out, the foundational issue is removing reading from grades, and returning it, especially in the younger years, to the accessible of reading to foster a love for reading. I have one student who loves words, is incredibly witty and can create metaphors instantly. Yet, despises reading because of school.
I couldn't agree more. I can still recall exact passages of books I loved when I was a teenager, the parts that I returned to again and again to read and enjoy. I also recall the disappointment I felt when I found books that I *thought* were the classic texts but had actually been abridged for younger/less advanced readers (to the point Madeleine makes about abridged versions of books). I wanted the real thing!
I appreciate the sentiment and agree that the bot can’t rewrite Shakespeare and call it Shakespeare. But I clicked on the two Atlantic articles for evidence that kids aren’t reading and was disappointed. I’ve been looking for peer-reviewed evidence to provide an anchor in reality. I’ve searched academic databases. I keep reading this broad generalization and want to know some empirical facts. I’m getting very sensitive to generalizations that are rooted in anecdotes and, as near as I can tell, biases. As a former English teacher, I love literature, but I love students more and want to see fewer denigrating generalizations and more clear-headed facts.
I would say that the phrase, "literacy crisis" is overblown, in the sense that reading scores haven't plummeted; they have remained relatively stagnant for the past 30+ years (with the exception of COVID times, where scores dipped). What has changed, though, is the number of kids reading for fun, especially outside of school, and this, coupled with increased screen time, is what I see as a major issue, affecting attention spans, etc. I have yet to listen to this podcast episode, but it's on my list: https://www.apa.org/news/podcasts/speaking-of-psychology/attention-spans
I don't know that there is peer-reviewed evidence for kids not reading full books in school, but I have heard a lot of anecdotal evidence on that matter. I will see what else I can find!
Magic School AI includes the first chapter of The Great Gatsby as the exemplar activity within their Text Leveler tool. You can rewrite the text at a first-grade reading level. Many of the high school teachers were shocked that the tech could do this and wanted to know why an edtech company like Magic School would use a text like Gatsby as an example. I think it's clear they don't care what teachers think. Screenshot included in a Google Doc.https://docs.google.com/document/d/19b8_GU9uZOLv6sFNLFSqsAkSReaKb6RbD_v_gUHeSCI/edit?usp=sharing
Fascinating that they would use this as an example.
This screen shot exemplifies and summarizes all the key points in article. For me, this is not dissimilar to graphic novels. While I've supported my studentreading graphic novels that are original content, I strongly discourage them from reading a graphic novel 'version' of a classic book.
I saw this when I was researching Ed tech companies for this article… this is wild to me!
To be fair, it’s just a new way of producing what has been done for decades. Abridged publications exist for many, many classics. Teachers have spent hours adapting extracts from complex texts with simplified sentence structures and vocabulary to enable pupils with learning needs and disabilities to at least begin to be able to access the texts. Side-by-side modern & original Shakespeare is a necessity in the classroom. It’s not one OR the other, it’s always both. We use adapted texts to help pupils into the originals.
It’s like you or I tackling Shakespeare, Middle English or Anglo Saxon text - we do not understand them immediately and need some help to get in. Then we can appreciate the beauty etc.
If AI is saving hours upon hours of teachers adapting and “translating” complex language for their pupils so that they can then get on to appreciate the original, that’s a good thing, no?
The days of halcyon immersive reading amongst the tulips just doesn’t exist any more. Humans, not just young people, are increasingly screen-addicted to the endless scroll on demand. We can make reading more attractive by making it accessible, but perhaps more importantly by divorcing it from grades and ultilitarian education systems.
I do hear you on (1) the history of abridged versions (I read an abridged version of Little Women that way, and as Rebekah said, was disappointed when I realized it wasn’t the original 🤣), and (2) having some type of translation to support students reading Shakespeare. I just still have doubts about whether AI translations/adaptations are the best way to do that. I almost think an edition curated with footnotes to help students understand the meaning/context of the language (and therefore be able to decipher it) might be more helpful than using an AI version. I do worry about comprehension development, too— giving students some semblance of the plot is okay, but if they’re reading an AI version, even if the original is there, they’re not really building syntactical or vocabulary knowledge. What do you think? Do you think there are alternative ways to balance students’ needs/ meeting them where they are with the cognitive challenge of deciphering a more advanced texts?
That’s a good point about the flattening of vocab by AI adapted texts - a human teacher would probably take more care over each sentence and make sentence and lexical level decisions for their class as they parse the text to be adapted.
That said, obviously an LLM can do this too, with careful prompting. You can tell it which sentence structures and words to keep. You can also far more quickly create a wider variety of versions of the same text at a range of levels.
The balance to be struck is time vs detail I guess.
Do I spend an hour personally adapting a text at maybe two different levels at most? Or, 10-20 minutes working with an AI to produce 3 or more differentiated levels, plus a vocabulary list, plus translations into other languages for my EAL learners, plus quiz questions to check for understanding, which I then check over and tweak?
Also, one other question, if you will indulge me!! What are the benefits/trade-offs of teaching multiple texts at once? So, let’s say you’re teaching 1984, but it’s too advanced for some in your classroom. Is it ever the case where one would assign 1984 to some students, and then let’s say, a text with similar themes that is easier to read, like The Hunger Games, to others? I remember them doing that in my elementary school, but I’m wondering how widespread text choice/multiple texts are.
Good question! Certainly one could do this with extracts in a one off lesson or two but it would be very tricky to teach different whole novels at the same time to different groups in the class over a longer period without compromising quality & depth. Also, the difference in difficulty levels even with extracts would need to be handled carefully such that groups don’t feel stigmatised for being given an “easier” text.
As a proficient reader at 5, I never understood the need for simplifying texts until I was tutoring. Today, the side by side modern and original basically is doing what a private tutor will do, explain the meaning in today's terms. That said, the original is still read, to expand a student's education. However, as you point out, the foundational issue is removing reading from grades, and returning it, especially in the younger years, to the accessible of reading to foster a love for reading. I have one student who loves words, is incredibly witty and can create metaphors instantly. Yet, despises reading because of school.
I couldn't agree more. I can still recall exact passages of books I loved when I was a teenager, the parts that I returned to again and again to read and enjoy. I also recall the disappointment I felt when I found books that I *thought* were the classic texts but had actually been abridged for younger/less advanced readers (to the point Madeleine makes about abridged versions of books). I wanted the real thing!
I appreciate the sentiment and agree that the bot can’t rewrite Shakespeare and call it Shakespeare. But I clicked on the two Atlantic articles for evidence that kids aren’t reading and was disappointed. I’ve been looking for peer-reviewed evidence to provide an anchor in reality. I’ve searched academic databases. I keep reading this broad generalization and want to know some empirical facts. I’m getting very sensitive to generalizations that are rooted in anecdotes and, as near as I can tell, biases. As a former English teacher, I love literature, but I love students more and want to see fewer denigrating generalizations and more clear-headed facts.
I have some stats I can send you. Or actually, let me see if I can drop the links here!
Check out the NAEP data, which does show a long-term decline in reading for pleasure: https://www.nationsreportcard.gov/highlights/ltt/2023/
I would say that the phrase, "literacy crisis" is overblown, in the sense that reading scores haven't plummeted; they have remained relatively stagnant for the past 30+ years (with the exception of COVID times, where scores dipped). What has changed, though, is the number of kids reading for fun, especially outside of school, and this, coupled with increased screen time, is what I see as a major issue, affecting attention spans, etc. I have yet to listen to this podcast episode, but it's on my list: https://www.apa.org/news/podcasts/speaking-of-psychology/attention-spans
I don't know that there is peer-reviewed evidence for kids not reading full books in school, but I have heard a lot of anecdotal evidence on that matter. I will see what else I can find!