"There was something terribly enthralling in the exercise of influence. No other activity was like it. To project one’s soul into some gracious form, and let it tarry there for a moment; to hear one’s own intellectual views echoed back to one with all the added music of passion and youth; to convey one’s temperament into another as though it were a subtle fluid or a strange perfume: there was a real joy in that—perhaps the most satisfying joy left to us in an age so limited and vulgar as our own, an age grossly carnal in its pleasures, and grossly common in its aims."
> AI systems don't recognize distinct texts or trace lines of influence; they process all written material as an undifferentiated sea of language patterns
Is that true?
o1-pro already seems above the average PhD level here
Yes it is true. As put here: "traditional LLM limitations: because they lack the ability to think and decide and verify they are best thought of as a tool for humans to leverage. " https://stratechery.com/2025/ais-uneven-arrival/
Thanks for helping confirm my plan. When teaching analytic geometry this spring, I intend to show the subject derives from work of over two thousand years, by Euclid, Pappus, Descartes, Desargues, Hilbert, and Hessenberg. The students will already have spent time reading the first of these.
By your description of teaching the Odyssey though, I don't believe I was taught it, though I read it in classes in high school and college; we just talked about it.
You say, "The Efficient Workforce Development camp's call for market alignment and employment metrics becomes less compelling when AI can help students acquire technical skills more efficiently outside traditional university structures."
This suggests that AI alone might satisfy the metrics better than the humans.
Terms like "machine learning" suggest that our students themselves are machines, only not as good as the *real* computers.
If computers can learn, then we go to college for something else.
A friend had a clever question for parents: did they want their children to get sex education in school, or sex training?
OK, I looked up the question; it is attributed to somebody called Dennis Rubin in the linked paper Sasha Borovik, who says, "if a certain mathematical skill can be taught by a computer, this is the best proof that this skill is economically redundant ... to have a happy and satisfying professional career, one has to be future-proof by
being able to re-learn the craft, to change his/her way of thinking. How can this skill of changing one’s way of thinking be acquired and nurtured? At school level—mostly by learning mathematics."
Me, I hope students will learn *that* they can know things
I don't think how something has influenced us is a big determining factor. People get into reading and pick up the western canon without having been influenced by the western canon to read itself. People still enjoy it because it stretches literature beyond its horizons.
I strongly disagree with this. There isn't one canon and ai can't ever write literary works (at least in the sense where it pushes literature beyond its treaded boundaries).
We're at a point where need to participate in meaning and a great conversation throughout history. I'm interested in a Christian canon. We don't need to, and shouldn't, reduce the concept of literary canons to some new technology that may already be peeking. Even if ai could write literary works, it's more important what it adds to the great conversation.
Oscar Wilde, The Picture of Dorian Gray:
"There was something terribly enthralling in the exercise of influence. No other activity was like it. To project one’s soul into some gracious form, and let it tarry there for a moment; to hear one’s own intellectual views echoed back to one with all the added music of passion and youth; to convey one’s temperament into another as though it were a subtle fluid or a strange perfume: there was a real joy in that—perhaps the most satisfying joy left to us in an age so limited and vulgar as our own, an age grossly carnal in its pleasures, and grossly common in its aims."
> AI systems don't recognize distinct texts or trace lines of influence; they process all written material as an undifferentiated sea of language patterns
Is that true?
o1-pro already seems above the average PhD level here
Yes it is true. As put here: "traditional LLM limitations: because they lack the ability to think and decide and verify they are best thought of as a tool for humans to leverage. " https://stratechery.com/2025/ais-uneven-arrival/
Yeah that's functionally how they work. It's not a bug it is the feature.
Thanks for helping confirm my plan. When teaching analytic geometry this spring, I intend to show the subject derives from work of over two thousand years, by Euclid, Pappus, Descartes, Desargues, Hilbert, and Hessenberg. The students will already have spent time reading the first of these.
By your description of teaching the Odyssey though, I don't believe I was taught it, though I read it in classes in high school and college; we just talked about it.
You say, "The Efficient Workforce Development camp's call for market alignment and employment metrics becomes less compelling when AI can help students acquire technical skills more efficiently outside traditional university structures."
This suggests that AI alone might satisfy the metrics better than the humans.
Terms like "machine learning" suggest that our students themselves are machines, only not as good as the *real* computers.
If computers can learn, then we go to college for something else.
A friend had a clever question for parents: did they want their children to get sex education in school, or sex training?
This is the conversation about AI and education I want to have! I'll be laughing all week at your last question...
OK, I looked up the question; it is attributed to somebody called Dennis Rubin in the linked paper Sasha Borovik, who says, "if a certain mathematical skill can be taught by a computer, this is the best proof that this skill is economically redundant ... to have a happy and satisfying professional career, one has to be future-proof by
being able to re-learn the craft, to change his/her way of thinking. How can this skill of changing one’s way of thinking be acquired and nurtured? At school level—mostly by learning mathematics."
Me, I hope students will learn *that* they can know things
https://arxiv.org/abs/1407.1954
I’m glad to see this ! Great Books! https://open.substack.com/pub/naomik/p/which-publisher-has-the-best-classics?r=2wtp2&utm_medium=ios
I don't think how something has influenced us is a big determining factor. People get into reading and pick up the western canon without having been influenced by the western canon to read itself. People still enjoy it because it stretches literature beyond its horizons.
Hey Hollis, you've got scammers
I strongly disagree with this. There isn't one canon and ai can't ever write literary works (at least in the sense where it pushes literature beyond its treaded boundaries).
We're at a point where need to participate in meaning and a great conversation throughout history. I'm interested in a Christian canon. We don't need to, and shouldn't, reduce the concept of literary canons to some new technology that may already be peeking. Even if ai could write literary works, it's more important what it adds to the great conversation.