Very excited to dig into all of these (and finish up going through some of the links from the last installment) but the thing that springs immediately to my mind while glancing over the blurbs is in relation to the invocation of language and the myth of Prometheus in the Hyper Landscapes excerpt: I had a similar thought when I first encountered the tale "The Smith and the Devil"," which by some accounts is possibly the oldest surviving western story, dating back to the Proto-Indo-European era. It seemed almost immediately to me that the story— about the Devil granting the Smith the ability to bind any substance to any other substance, whereupon the Smith promptly binds the Devil to a chair, preventing him from ever collecting his debt— had to be about the development of language and possibly even the sense of our inner selves that language allows us to explore in our thoughts. Then again I may be biased because I'm a writer with an interest in the nature of consciousness.
Thank you so much for doing this--it means a lot to us readers of yours that you'd wade through so much material and actually make comments on it. It's gratifying to be included!
Appreciate that! Although I'll note I do get something out of it, which is that it encourages people on the fence to become paid subscribers. But I did honestly get a kick out of seeing what many of names I recognize from comments and so on are writing.
Yes - the strategy worked on me! Still, it's a great mine for resources and it helps others get visibility.
Re: one of the links in this one: I've been spending the last day grading papers and am happy to report that I've seen no sign of ChatGPT papers yet. Apparently, I succeeded in making clear it was a zero-tolerance policy at the beginning of the semester. Resistance isn't always futile.
That's good news. I always think that in general bans actually work, since most people will obey them to begin with, without any enforcement. Which is kind of what I think about AI in general - as much as we pretend its some sort of inevitable technological development, it's very much a choice.
I'm glad to hear not everyone is having trouble. I taught two classes in the spring, and one of them had no cheating and the other had...a lot. So I can definitely see it differing from group to group.
I am a bit concerned that students will get better at faking it the more they learn about ChatGPT. It was easier for me to catch them because they had no idea what it was doing or that it could confabulate false info. I had one student actually try to use it responsibly, but he didn't know it make up sources and so when he used it as a research tool to find quotes it misled him.
Oct 20, 2023·edited Oct 20, 2023Liked by Erik Hoel
I wonder whether "getting better at faking it" eventually blurs into, basically, how people will write in the future?
E.g., You (a student) ask ChatGPT to write your history essay for you, but it confabulates sources and covers material outside the scope of the class. So you track down some real sources, copy + paste a few blocks of original text into your prompt, and ask it to rewrite the essay citing only from those sources. But now the paper is too long and feels formulaic, so you start messing with the syntax in places and moving paragraphs around. And now that you check the source, you see it actually missed something important, etc., etc.
Obviously this isn't WRITING in the traditional sense, but it's getting much closer to writing. It definitely feels like we're moving up Bloom's taxonomy: The more compelling the fake, the more genuine cognitive effort involved in producing it—if only to develop ever-more-sophisticated prompts.
As a writer and teacher, this possible future makes me sad. But I'm not sure it's *necessarily* harmful for society and schools, at least in the long run. It sure as hell changes things though...
A few of my students said something along the same lines--they felt it would end being more work to fix a ChatGPT paper than just write their own. It might be, but I'm glad it discouraged them from over-relying on it.
In the long run you might be right, the way people write will evolve this way. I actually think this is more likely in professional and corporate settings, where fewer and fewer copywriters and editors will be saddled with even more work and will have to cognitively outsource their thinking to machines in order to keep up with the demands of "productivity." It might work, and it might work very fast, but I don't think it'll be good for either the development of students' minds or the creation of new ideas in professionals.
Great list! I found some gems that I'll enjoy with a bowl of tea. I already read one article and then a second from one of the listed writers. This is good stuff.
Thanks so much for sharing my piece on NASA and psychopathy, Erik! It's very admirable the way you've been using your platform to boost other Substack writers. You are most definitely NOT a psychopath.
Just wanted to thank you SO much for including my post in this list, I'd been hoping that you would but wasn't sure how it would measure up to all these other really great articles.
Thank you so much for sharing this Erik! Really appreciate you taking the time. Much appreciated my man ❤️
Looking forward to some weekend reading.
Very excited to dig into all of these (and finish up going through some of the links from the last installment) but the thing that springs immediately to my mind while glancing over the blurbs is in relation to the invocation of language and the myth of Prometheus in the Hyper Landscapes excerpt: I had a similar thought when I first encountered the tale "The Smith and the Devil"," which by some accounts is possibly the oldest surviving western story, dating back to the Proto-Indo-European era. It seemed almost immediately to me that the story— about the Devil granting the Smith the ability to bind any substance to any other substance, whereupon the Smith promptly binds the Devil to a chair, preventing him from ever collecting his debt— had to be about the development of language and possibly even the sense of our inner selves that language allows us to explore in our thoughts. Then again I may be biased because I'm a writer with an interest in the nature of consciousness.
Really appreciate this Erik - we've already had a spike in visits to our site. Plus, there is some really good reading here.
Cheers!
Nice - good to hear that
Thank you so much for doing this--it means a lot to us readers of yours that you'd wade through so much material and actually make comments on it. It's gratifying to be included!
Thank you again for doing this. It shows such generosity of spirit. And the links are fascinating!
Appreciate that! Although I'll note I do get something out of it, which is that it encourages people on the fence to become paid subscribers. But I did honestly get a kick out of seeing what many of names I recognize from comments and so on are writing.
Yes - the strategy worked on me! Still, it's a great mine for resources and it helps others get visibility.
Re: one of the links in this one: I've been spending the last day grading papers and am happy to report that I've seen no sign of ChatGPT papers yet. Apparently, I succeeded in making clear it was a zero-tolerance policy at the beginning of the semester. Resistance isn't always futile.
That's good news. I always think that in general bans actually work, since most people will obey them to begin with, without any enforcement. Which is kind of what I think about AI in general - as much as we pretend its some sort of inevitable technological development, it's very much a choice.
I'm glad to hear not everyone is having trouble. I taught two classes in the spring, and one of them had no cheating and the other had...a lot. So I can definitely see it differing from group to group.
I am a bit concerned that students will get better at faking it the more they learn about ChatGPT. It was easier for me to catch them because they had no idea what it was doing or that it could confabulate false info. I had one student actually try to use it responsibly, but he didn't know it make up sources and so when he used it as a research tool to find quotes it misled him.
I wonder whether "getting better at faking it" eventually blurs into, basically, how people will write in the future?
E.g., You (a student) ask ChatGPT to write your history essay for you, but it confabulates sources and covers material outside the scope of the class. So you track down some real sources, copy + paste a few blocks of original text into your prompt, and ask it to rewrite the essay citing only from those sources. But now the paper is too long and feels formulaic, so you start messing with the syntax in places and moving paragraphs around. And now that you check the source, you see it actually missed something important, etc., etc.
Obviously this isn't WRITING in the traditional sense, but it's getting much closer to writing. It definitely feels like we're moving up Bloom's taxonomy: The more compelling the fake, the more genuine cognitive effort involved in producing it—if only to develop ever-more-sophisticated prompts.
As a writer and teacher, this possible future makes me sad. But I'm not sure it's *necessarily* harmful for society and schools, at least in the long run. It sure as hell changes things though...
A few of my students said something along the same lines--they felt it would end being more work to fix a ChatGPT paper than just write their own. It might be, but I'm glad it discouraged them from over-relying on it.
In the long run you might be right, the way people write will evolve this way. I actually think this is more likely in professional and corporate settings, where fewer and fewer copywriters and editors will be saddled with even more work and will have to cognitively outsource their thinking to machines in order to keep up with the demands of "productivity." It might work, and it might work very fast, but I don't think it'll be good for either the development of students' minds or the creation of new ideas in professionals.
Great list! I found some gems that I'll enjoy with a bowl of tea. I already read one article and then a second from one of the listed writers. This is good stuff.
Thanks so much for sharing my piece on NASA and psychopathy, Erik! It's very admirable the way you've been using your platform to boost other Substack writers. You are most definitely NOT a psychopath.
Chur bro, appreciate what you're doing aye. The Delinquent Academic.
Just wanted to thank you SO much for including my post in this list, I'd been hoping that you would but wasn't sure how it would measure up to all these other really great articles.
Thank you, Erik. Thank you!
Great roundup
Eric, what a beautiful list. I think we are on the same page, with this, no?
https://herojig.medium.com/my-project-plan-for-techbros-must-die-caa778b1c499
Let me know here!