To follow up on the featured quote from Agasthya Shenoy:
What I've noticed with AI is that it is VERY sensitive to the specific wording of the prompt. So, if you are able to **exactly** articulate what you're looking for, the tools can be powerful. That works great for people who already have a strong foundation, and are just trying to resolve a small technical point. But often in the classroom struggling students have difficulty even formulating what they don't know, and no AI system is going to work well with a prompt like "I don't get it." One of the roles of good teachers is that they can pick up on student's weaknesses and figure out where they need help even when the student can't. I think ultimately AI will help those students who are already doing well, while not adding much value to those who are not. Unfortunately, that is counter to the whole premise of using AI in schools, which is to provide "personalized" tutoring to weaker students. So I'm skeptical that this is going to make much of a difference, and if it does it will probably just widen the achievement gap, not reduce it.
Exactly. The level of metacognition required to iterate on your query as well the level of expertise required to factcheck the result makes AI chatbots a learning tool for the 1% of students rather than the 99%.
Years ago I stopped saying "Please contact me if you have questions about the homework" because I finally figured out that just being able to formulate a question is a significant barrier to seeking help. Now I try to use something more neutral like "if you want to talk about the homework" or "if you want to look at the homework together."
The chatbots do not really seem to have any metacognition or memory. Yesterday, I thought I could try using copilot for some Godot help. My text was blurry on resize, using a power of 2 on a square image. I asked Copilot, which said something I had already read- about-use "multiples of the design size". I asked for the design size of two specific fonts that I am using-DejaVu Sans and Gentium Book Basic (nice pi and infinity symbols in both and not locked for commercial use) and I have been unable to find the design size. As if it had no idea it mentioned design size, it said most people use those fonts in the 10-12pt range. If I were to assume the design size was 10, 11 or 12 and require a power of two for resize my math tells me these would be some very strange numbers that would have factors of 10, 11 and 12 that were also a power of two on resize-not sure the resized image would fit in my house. These products are really not ready but they are being pushed, I suspect, so experts provide feedback necessary for training. I don't like working for free but I did give feedback.
Little Billy Gates would have LOVED a personalized AI tutor, he would have been able to leave the rest of the class far behind. I suspect that when these Techbros say "A wide range of abilities in the classroom means we need personalized instruction" what they really mean is "Kids who are like I was won't be held back by their slower peers." And then it's sold as the complete opposite.
I appreciate your rigorous criticism of AI. I can't get beyond the weirdness of it all. Or I am missing the whole point.
With the data from communication of matters of current math, responses are generated. Seems this assures teaching/learning are constrained to a perpetual loop of contemporaneity. They can never get on a path to the world the technology behind it all is taking them.
I admit to watching the Netflix show re Bill Gates today and although he was not talking about AI, he was talking about tech and health care and this quote from "What's Next? The Future With Bill Gates," which premiered today he says:
"I need to learn more because I naively still believe that digital communication can be a force to bring us together, to have reasonable debate". This man could really benefit from an in person conversation with our Dan. He would have the opportunity to learn somethings.
Absolutely. No dramatic, disruptive transformations overnight in education with tech. Not even AI. Just incremental (and important) progress over time, as it has always happened. Justin Reich's book 'Failure to Disrupt' captures this really well.
Two weeks left for the AI companies to release the tutors who will be better at teaching kids to read than humans, as Gates predicted in April 2023. I'm excited for the release!
I teach college and maybe this is different with K-12, but what I've noticed is that, when one student is confused about a particular idea, usually MANY students are confused about that same idea. I've been doing this long enough that I know ahead of time which ideas tend to cause confusion, and typically that means "confusion for more than half my class." And I think it benefits students to see they're not alone, that lots of other students also don't get this particular idea right away. Wouldn't the student laboring away with her individualized chatbot jump to the conclusion that nobody else finds this difficult? How would she know?
I was just doing a lesson from the Quantway curriculum about "percentage of women who smoke" vs. "percentage of smokers who are women" and I think it's beneficial to students to share that moment when EVERYBODY is confused.
And I’m guessing any teacher worth their salt would have seen this as it played out. Thanks for giving those thoughts a firm foundation in his fast paced but contentless words.
Has anyone found a better link for the Stanford survey results mentioned in the article? The current link goes to the main page showcasing course projects, but it's not clear to me where to find the survey details.
To follow up on the featured quote from Agasthya Shenoy:
What I've noticed with AI is that it is VERY sensitive to the specific wording of the prompt. So, if you are able to **exactly** articulate what you're looking for, the tools can be powerful. That works great for people who already have a strong foundation, and are just trying to resolve a small technical point. But often in the classroom struggling students have difficulty even formulating what they don't know, and no AI system is going to work well with a prompt like "I don't get it." One of the roles of good teachers is that they can pick up on student's weaknesses and figure out where they need help even when the student can't. I think ultimately AI will help those students who are already doing well, while not adding much value to those who are not. Unfortunately, that is counter to the whole premise of using AI in schools, which is to provide "personalized" tutoring to weaker students. So I'm skeptical that this is going to make much of a difference, and if it does it will probably just widen the achievement gap, not reduce it.
Exactly. The level of metacognition required to iterate on your query as well the level of expertise required to factcheck the result makes AI chatbots a learning tool for the 1% of students rather than the 99%.
Years ago I stopped saying "Please contact me if you have questions about the homework" because I finally figured out that just being able to formulate a question is a significant barrier to seeking help. Now I try to use something more neutral like "if you want to talk about the homework" or "if you want to look at the homework together."
The chatbots do not really seem to have any metacognition or memory. Yesterday, I thought I could try using copilot for some Godot help. My text was blurry on resize, using a power of 2 on a square image. I asked Copilot, which said something I had already read- about-use "multiples of the design size". I asked for the design size of two specific fonts that I am using-DejaVu Sans and Gentium Book Basic (nice pi and infinity symbols in both and not locked for commercial use) and I have been unable to find the design size. As if it had no idea it mentioned design size, it said most people use those fonts in the 10-12pt range. If I were to assume the design size was 10, 11 or 12 and require a power of two for resize my math tells me these would be some very strange numbers that would have factors of 10, 11 and 12 that were also a power of two on resize-not sure the resized image would fit in my house. These products are really not ready but they are being pushed, I suspect, so experts provide feedback necessary for training. I don't like working for free but I did give feedback.
Little Billy Gates would have LOVED a personalized AI tutor, he would have been able to leave the rest of the class far behind. I suspect that when these Techbros say "A wide range of abilities in the classroom means we need personalized instruction" what they really mean is "Kids who are like I was won't be held back by their slower peers." And then it's sold as the complete opposite.
I appreciate your rigorous criticism of AI. I can't get beyond the weirdness of it all. Or I am missing the whole point.
With the data from communication of matters of current math, responses are generated. Seems this assures teaching/learning are constrained to a perpetual loop of contemporaneity. They can never get on a path to the world the technology behind it all is taking them.
I admit to watching the Netflix show re Bill Gates today and although he was not talking about AI, he was talking about tech and health care and this quote from "What's Next? The Future With Bill Gates," which premiered today he says:
"I need to learn more because I naively still believe that digital communication can be a force to bring us together, to have reasonable debate". This man could really benefit from an in person conversation with our Dan. He would have the opportunity to learn somethings.
Absolutely. No dramatic, disruptive transformations overnight in education with tech. Not even AI. Just incremental (and important) progress over time, as it has always happened. Justin Reich's book 'Failure to Disrupt' captures this really well.
Two weeks left for the AI companies to release the tutors who will be better at teaching kids to read than humans, as Gates predicted in April 2023. I'm excited for the release!
I teach college and maybe this is different with K-12, but what I've noticed is that, when one student is confused about a particular idea, usually MANY students are confused about that same idea. I've been doing this long enough that I know ahead of time which ideas tend to cause confusion, and typically that means "confusion for more than half my class." And I think it benefits students to see they're not alone, that lots of other students also don't get this particular idea right away. Wouldn't the student laboring away with her individualized chatbot jump to the conclusion that nobody else finds this difficult? How would she know?
I was just doing a lesson from the Quantway curriculum about "percentage of women who smoke" vs. "percentage of smokers who are women" and I think it's beneficial to students to share that moment when EVERYBODY is confused.
And I’m guessing any teacher worth their salt would have seen this as it played out. Thanks for giving those thoughts a firm foundation in his fast paced but contentless words.
Has anyone found a better link for the Stanford survey results mentioned in the article? The current link goes to the main page showcasing course projects, but it's not clear to me where to find the survey details.
It isn't a survey. It's a collection of projects.
THIS is very valuable.
Thank you.