"...Consequently, our inability to detect fraudulent use of generative AI requires a complete transformation in how we evaluate student work. This presents a significant challenge in itself. However, the purpose of this article is not to delve into the challenges associated with generative AI, as many other writers on Substack have extensively explored this topic..."
Any specific post about it from other authors that you recommend?
And I stand by what I wrote months ago, before ChatGPT was a thing, in my first two essays, about the role of team-work and how feedback is the most important motivation for college evaluation. I think that is still true. Even more true now, if that makes sense.
Before ChatGPT, I was already “ungrading.” Mainly, I focus on specifications grading and having students do work for real audiences.
It is nothing new ... we need to focus on problem solving, collaborating in diverse groups, and managing complex projects ... not content-based evaluation.
Richard Talbert’s Substack is useful but doesn’t necessarily focus on AI.
Well said. I think a huge thing about this is that we're all sort of viewing these revolutionary tools through the following lens: it's a new tool, therefore we need to devise new ways to use it.
I think it's far better to consider how to amplify our current framework of education, as you are both suggesting here, without trying to separate students from teachers or trying to replace all education with AI tomorrow.
I like how you frame this. I think this might explain some of the pushback I get from educators. They think that I am adapting education to fit AI ... I am adapting AI to fit education.
Lance, it seems like some folks don't understand that the very ground we are standing on is shifting under our feet.
Change isn't just for fun! It's how we survive, and ultimately thrive. And Lance, thank you for continuing a very good discussion with both this piece, and here in the comments! This is why I'm here.
I never got a chance to really applaud this major achievement in human-centric education theory and pedagogy. Alejandro and Lance, I am really impressed by the depth of vision and knowledge. First, I agree the LLM offer no new knowledge as research tools. That is critical. Second, I agree that LLMs introduce profound epistemology uncertainty into the process of evaluating student writing, resistant to all existing interventions. Third, I agree that LLMs are not transforming education, but rather are offering us an opportunity to focus on what really matters. Finally, I appreciate how the social dynamics and the guidance of the instructor feature renewed areas of focus. To this point, I feel like my own project has grown a great deal through my interactions with you two fine gentlemen!!! Keep up the good work!!!😀😀
Thanks, Nick, you're too kind. I've done nothing but write down what's been on my head for a while, but folks like Lance are the ones really spearheading this discussion. He's really putting all of these practices under a microscope and leading the way on their implementation.
Major Starting of the Downfall was Clinton’s Changing the Country from a Manufacturing base to a Services Based One! Tradesman don’t need College! & training is harder to get than Before !! Gonna have to get back All factories & such, especially in 2014!
"...Consequently, our inability to detect fraudulent use of generative AI requires a complete transformation in how we evaluate student work. This presents a significant challenge in itself. However, the purpose of this article is not to delve into the challenges associated with generative AI, as many other writers on Substack have extensively explored this topic..."
Any specific post about it from other authors that you recommend?
On the limitations? Anything by Gary Marcus basically, and Ethan Mollick too. This post is a pretty good summary: https://www.oneusefulthing.org/p/what-people-ask-me-most-also-some
I meant it on the transformations required in how we evaluate student work :)
That post also has some pretty good ideas, but Lance Cummings is my go-to reference. He's linked from the article.
And I stand by what I wrote months ago, before ChatGPT was a thing, in my first two essays, about the role of team-work and how feedback is the most important motivation for college evaluation. I think that is still true. Even more true now, if that makes sense.
Before ChatGPT, I was already “ungrading.” Mainly, I focus on specifications grading and having students do work for real audiences.
It is nothing new ... we need to focus on problem solving, collaborating in diverse groups, and managing complex projects ... not content-based evaluation.
Richard Talbert’s Substack is useful but doesn’t necessarily focus on AI.
Oh yeah I discovered Richard a while back and binged his entire substack.
Well said. I think a huge thing about this is that we're all sort of viewing these revolutionary tools through the following lens: it's a new tool, therefore we need to devise new ways to use it.
I think it's far better to consider how to amplify our current framework of education, as you are both suggesting here, without trying to separate students from teachers or trying to replace all education with AI tomorrow.
I like how you frame this. I think this might explain some of the pushback I get from educators. They think that I am adapting education to fit AI ... I am adapting AI to fit education.
Lance, it seems like some folks don't understand that the very ground we are standing on is shifting under our feet.
Change isn't just for fun! It's how we survive, and ultimately thrive. And Lance, thank you for continuing a very good discussion with both this piece, and here in the comments! This is why I'm here.
Absolutely! It's a wonderful tool and it would be silly not to use it.
Exactly. Tech is meant to improve our lives, and students need more people, more humanity, more community, not less.
I never got a chance to really applaud this major achievement in human-centric education theory and pedagogy. Alejandro and Lance, I am really impressed by the depth of vision and knowledge. First, I agree the LLM offer no new knowledge as research tools. That is critical. Second, I agree that LLMs introduce profound epistemology uncertainty into the process of evaluating student writing, resistant to all existing interventions. Third, I agree that LLMs are not transforming education, but rather are offering us an opportunity to focus on what really matters. Finally, I appreciate how the social dynamics and the guidance of the instructor feature renewed areas of focus. To this point, I feel like my own project has grown a great deal through my interactions with you two fine gentlemen!!! Keep up the good work!!!😀😀
Thanks, Nick, you're too kind. I've done nothing but write down what's been on my head for a while, but folks like Lance are the ones really spearheading this discussion. He's really putting all of these practices under a microscope and leading the way on their implementation.
Major Starting of the Downfall was Clinton’s Changing the Country from a Manufacturing base to a Services Based One! Tradesman don’t need College! & training is harder to get than Before !! Gonna have to get back All factories & such, especially in 2014!