Learning with AI: A Student Perspective

Want to know what students think about AI and education?

Check out this informative webinar where a diverse panel of students shared their viewpoints on AI in the classroom! From existing applications to future possibilities, we discussed the pros and cons of various solutions such as AI tutors or AI detection software. This student-centered discussion provided unique insights into their AI expectations and experiences in the classroom.

Some of the questions discussed:

  • In what ways are you currently seeing AI used in your education?

  • What role do you think AI should play in a student’s education?

  • What AI applications would be most beneficial for students?

  • What concerns, if any, do you have about incorporating AI in Education?

  • Do you think AI teachers will ever replace human teachers?

  • What kind of training do students need to be prepared for AI?

  • As students, what advice would you give students about using AI effectively in classrooms?

  • Amanda Bickerstaff

    Amanda is the Founder and CEO of AI for Education. A former high school science teacher and EdTech executive with over 20 years of experience in the education sector, she has a deep understanding of the challenges and opportunities that AI can offer. She is a frequent consultant, speaker, and writer on the topic of AI in education, leading workshops and professional learning across both K12 and Higher Ed. Amanda is committed to helping schools and teachers maximize their potential through the ethical and equitable adoption of AI.

    Ana Herrera

    Active writer and AI enthusiast, Ana Herrera is currently an 11th grader at Trinity Preparatory School. Beyond the classroom, she fuels her passion for language through active roles in the school newspaper and creative writing program. Ana is also the president and founder of "Read, Write, & Inspire Society," which dedicates itself to sharing her passion for literacy with the community. With her writing and organization, she advocates the responsible integration of Artificial Intelligence (AI) in education and creative expression. A believer in AI's potential to enhance student work and writing, while upholding academic integrity, Ana shares her expertise through "AI Lectures" across various English and creative writing classes, from elementary to college levels.

    Nicolas Gertler

    Nicolas Gertler is Yale University's first-ever AI Ambassador—he leads workshops on AI ethics and literacy, develops academic resources for responsible AI usage across the university, and builds equitable AI tools for the Yale community. Nicolas also researches infusing novel pedagogical approaches into AI-enabled platforms to foster more engaging and agential learning at the Yale Digital Ethics Center. Nicolas is also the Vice President of Fidutam—where he manages the organization's day-to-day operations and spearheads its global expansion—and the Chair of the AI Issue Advisory Council and AI and Education Advisor at Encode Justice, where he leads 25+ AI Advisors, providing Congress with policy memos on AI's domain-specific implications and supports educational institutions in effectively and ethically integrating AI into their pedagogy.

    Pranava Manthena

    Pranava Manthena is a Student Voice/Marketing Intern at AI for Education. He is motivated junior at Middleton High School. Pranava's achievements include winning first place in the Future Business Leaders of America (FBLA) national competition and state competition. He has led business education workshops at the national level, and spoke at global conferences such as GovAI 2023 and the prestigious IIT Global 2024 regarding the future of AI in education. Pranava is commited to bridging the gap between students and teachers regarding AI literacy.

  • 00:00:45.000 --> 00:00:50.000

    So we're gonna give everyone just a moment to file in. Say hi.

    00:00:50.000 --> 00:00:58.000

    So I see got Riverdale in Philadelphia, Ohio. So say hello and got New York City.

    00:00:58.000 --> 00:01:02.000

    Also Joe, I'm sure it's raining for you as it's rating for me.

    00:01:02.000 --> 00:01:09.000

    And so yeah, so please put in your information. Hey Scott and like good to see you again and Samantha I talked to Sam today.

    00:01:09.000 --> 00:01:14.000

    Love having you all here. And we're going to get started. So as I said, my name is Amanda Vicar. Staff.

    00:01:14.000 --> 00:01:19.000

    I'm the CEO and co-founder of AI for Education. Today we are going to have such a good time.

    00:01:19.000 --> 00:01:28.000

    We have a webinar with students focused on learning with AI from a student perspective. And this has come about because we have had a really big focus since we started AI for education last year and including student voice.

    00:01:28.000 --> 00:01:41.000

    And so we've already had a student webinar that I thought was amazing and we continued to get to work with students including doing a promptathon with students out in Seattle last week.

    00:01:41.000 --> 00:01:48.000

    And we think that it's just so important to really situate and center our conversations about Generative AI with students because it's going to impact students significantly more than it's going to impact us.

    00:01:48.000 --> 00:01:55.000

    You know, these students are going into a role that's radically transforming. And so it's really important to situate and center our conversations with them.

    00:01:55.000 --> 00:02:03.000

    So. Really excited. I've got Nick. I've got Pranava and I got Anna who are joining us today.

    00:02:03.000 --> 00:02:05.000

    But the first thing is I want to set the scene. So the first thing is get involved. So make sure that you're using the chat and Q&A functions.

    00:02:05.000 --> 00:02:17.000

    Our community that has been a part of our webinars. You might have been here for one the first time or you've been here for many, but make sure that you're setting like saying hello.

    00:02:17.000 --> 00:02:23.000

    I would say if you want to talk to each other, the chat's a great piece. If you want to talk to the panel or my self, use the Q&A.

    00:02:23.000 --> 00:02:33.000

    So if you have a question and we will have time at the end for questions. The second is we're not gonna do any prompting today, but if we do something that you want to kinda share and understand, feel free to kind of take notes and get involved.

    00:02:33.000 --> 00:02:38.000

    And the last thing is to share resources. If you have a great resource that you like to share, please do so.

    00:02:38.000 --> 00:02:55.000

    So really excited to ask my panel to come up on stage with me. I'm very excited. To get everybody in and I will say, first of all, let's give some love really quickly to our students here because we have some students that are coming just like from school had a long day at college.

    00:02:55.000 --> 00:03:03.000

    We have 2 high schoolers and a college student. And so let's give them some love. First, like everybody like can't, you know, absolutely.

    00:03:03.000 --> 00:03:05.000

    But I want to say thank you so much for being here, everybody. And what we're going to do is we're gonna start like we always do.

    00:03:05.000 --> 00:03:19.000

    So we're gonna start actually with Nick and we're gonna say like I would love you to introduce yourself and what brought you to our webinar today but also what is the first time you use generative AI?

    00:03:19.000 --> 00:03:20.000

    Absolutely. Thank you so much, Amanda. And it's great to be here with all of you.

    00:03:20.000 --> 00:03:30.000

    So I've been working in AI for over 4 years now. I actually passed past year I've been focusing on AI and education.

    00:03:30.000 --> 00:03:43.000

    I'm now a first year student at Yale setting cognitive science but during my time in AI in education I've spoken directly and advise the US Secretary of State, but during my time in AI in education, I've spoken directly and advise the US Secretary of Education.

    00:03:43.000 --> 00:03:48.000

    I've worked with Google and the US Secretary of Education. I've worked with Google and I've worked with the United Nations on this work.

    00:03:48.000 --> 00:03:56.000

    And my first time actually engaging with Generative AI was in, I believe late, 2,000, and 21, I downloaded a LLLL, to my computer and started playing with it, prompting and I told my dad like you have to see this.

    00:03:56.000 --> 00:04:01.000

    This is crazy. This is gonna change the world and you know since then it's sort of exploded.

    00:04:01.000 --> 00:04:03.000

    That was my first experience.

    00:04:03.000 --> 00:04:16.000

    That's great. And so you're an OG. So, 2021, OG, NICK is like by far the first time I met Nick I was like how have you done so much but you know someone that is so committed to the work that you do I think is just amazing.

    00:04:16.000 --> 00:04:23.000

    We're so glad to have you here today. And next we're going to Anna and can you introduce yourself and talk about the first time you used Generative AI?

    00:04:23.000 --> 00:04:45.000

    Thank you so much. Hi, I'm Ana. The first time I used AI, so just to prep this with my experience with AI, I have a school organization that we have we host AI lectures at the high school and elementary level and we just explained how it can be incorporated into education curriculum and whatnot.

    00:04:45.000 --> 00:04:51.000

    But the first time I used AI, I'll be very honest. I was very nervous. I was like typing in chat GPT and I was like, I think my hands are shaking.

    00:04:51.000 --> 00:05:06.000

    But then I asked it come up with a synonym for I think the word happy and it provided me a list of amazing like words because I love to write.

    00:05:06.000 --> 00:05:14.000

    And just that I saw so many possibilities in that moment and for a moment I realized that I had been building.

    00:05:14.000 --> 00:05:19.000

    AI for a while and then I realize no it can be an amazing tool and there's so much capabilities.

    00:05:19.000 --> 00:05:20.000

    So yeah.

    00:05:20.000 --> 00:05:24.000

    And so can you tell us where you go to school and what grade you're in?

    00:05:24.000 --> 00:05:31.000

    Yes, sorry. I go to Trinity Preparatory School in Orlando, Florida and I'm in the eleventh grade.

    00:05:31.000 --> 00:05:43.000

    Eleventh grader, talk about a month grader doing lectures already and I and we're definitely gonna talk about that balance of villainization versus you know the potential of positive learning experiences. So that's great.

    00:05:43.000 --> 00:05:48.000

    And then Pranava, Parnava is our first ever student voice in turn here at AI for Education.

    00:05:48.000 --> 00:05:56.000

    You might have seen that announcement a couple weeks ago. Also a high schooler. So same to you, your introduction and then your first time using Generative AI.

    00:05:56.000 --> 00:06:06.000

    Sure, yeah. Hey, everyone. My name is Pranova and I'm in eleventh grade right now and I'm at Middleton High School in Tampa, Florida.

    00:06:06.000 --> 00:06:21.000

    In AI itself in this field at least I've First gotten interested or the first time I sort of, delved into AI was actually, I was working on a, like this competition with a friend.

    00:06:21.000 --> 00:06:26.000

    And we're working on together and I was typing up some of the details for the competition. It was this website we had to make.

    00:06:26.000 --> 00:06:34.000

    So it was typing up something like the about pages and things like that and we needed some pictures because we couldn't really get the image of this website that we were building.

    00:06:34.000 --> 00:06:40.000

    We needed some sort of idea of what pictures we're gonna put on and stuff. And so Friend is a quick Google search.

    00:06:40.000 --> 00:06:45.000

    And he calls me, you got to see this come here. And I walked over and I see him on a discord chat.

    00:06:45.000 --> 00:06:52.000

    And like, what do you do on a discord chat? He talks about mid journey, mid journey and the very entry level on the Discord chat.

    00:06:52.000 --> 00:06:55.000

    He put in a quick prompt and it gave so many nice images. I was like, wow, that's crazy.

    00:06:55.000 --> 00:07:06.000

    So that was my sort of first introduction to AI. And from then on, I sort of delved into AI and from then on I sort of delved into AI specifically in and education.

    00:07:06.000 --> 00:07:12.000

    And, I went and do a couple of talks. And, some other locations.

    00:07:12.000 --> 00:07:16.000

    And yeah, I'm here at AI for Education now. Super excited to be working on some big things.

    00:07:16.000 --> 00:07:21.000

    So That's spin it.

    00:07:21.000 --> 00:07:26.000

    And thank you, Supernova pulled us all together. This is his first internship moment, so we did such a good job.

    00:07:26.000 --> 00:07:38.000

    But first of all, I just want to say that what's really fascinating here is that we're not just like have a panel of high school and college students that are focused on AI, but the really interesting part is that you all are focused on AI and education.

    00:07:38.000 --> 00:07:50.000

    So you're all doing the work that we do in your own, your own applications, which think it's just so amazing because I think that is it shows how important this moment in time is and that recognize that this inflection point within schools I think is really amazing.

    00:07:50.000 --> 00:07:58.000

    So what we're gonna do is we're gonna kick off with our first major question. I just want to know more about how you're using Generative AI.

    00:07:58.000 --> 00:08:06.000

    So you're, you know, you got our first year student who's got a million things on your plate Nick but like how are you actually using it for your own practice?

    00:08:06.000 --> 00:08:14.000

    Absolutely. So I think there are lots of great ways to use AI in a responsible manner and that's something that I really try to encourage when I speak with other students and you know talk, talk with faculty about how students are using AI.

    00:08:14.000 --> 00:08:42.000

    So a great way that I use it is for example I develop a piece of writing and then I put it into a tool like Chat GPT and say can you please give me a critique of this writing on you know the style of voice or the sound of writing the voice the flow cohesion structure things of that nature and then it's able to sort of give me some really great suggestions that I then go and

    00:08:42.000 --> 00:08:48.000

    take and implement. Another great way I've used it is by, inputting my study material.

    00:08:48.000 --> 00:08:55.000

    Into it so I've you know for example at a cognitive science midterm last semester and I went and put all my material there.

    00:08:55.000 --> 00:09:02.000

    I said, please generate a short answer questions for me and grade my answer one through 5 which was the foremounted the exam and then developed original questions right and I was able to have a really interactive and dynamic conversation with AI.

    00:09:02.000 --> 00:09:13.000

    And it was a really helpful way to help me study. I didn't use that as my main study.

    00:09:13.000 --> 00:09:15.000

    You know, I didn't use as the only study tool, but I think it really complemented the other tools.

    00:09:15.000 --> 00:09:26.000

    I was using since active recall and some other strategies.

    00:09:26.000 --> 00:09:41.000

    That's great. And I think that, we always, I always love when I see like tools around you're using it for your own feedback because I mean I'm you're in college right you're at Yale and I'm sure you have support but you're in this world now where you're no longer in a high school and you're getting all this support, right?

    00:09:41.000 --> 00:09:50.000

    And so studying really is on to your left to your own devices. And so I remember that moment. So the fact that you're using these tools to partner around a steady aid is really great.

    00:09:50.000 --> 00:09:54.000

    And I think that how did you on your test?

    00:09:54.000 --> 00:09:55.000

    I ended up doing really well, very was very happy with how it went.

    00:09:55.000 --> 00:10:04.000

    Yeah. Good I have no I have no doubt but it is that is fun. Okay awesome and then, the same question to you.

    00:10:04.000 --> 00:10:07.000

    So how do you use AI in your own practice?

    00:10:07.000 --> 00:10:24.000

    So yeah, very similar to Nicholas. I'm very, I try to be very cautious with the line because obviously there's a huge conversation about the academic integrity of that and that is something very important to me but I do view AI as a super valuable tool.

    00:10:24.000 --> 00:10:31.000

    So what I do very similar things like I insert my notes and say, Hey, could you please generate some, multiple choice questions for me.

    00:10:31.000 --> 00:10:35.000

    Like it's a great studying tool. I also, multiple choice questions for me. Like it's a great studying tool.

    00:10:35.000 --> 00:10:36.000

    I also, as I mentioned before, I'm like, it's a great studying tool. I also, as I mentioned before, I'd love to write like creative writing.

    00:10:36.000 --> 00:10:40.000

    So whenever I'm trying to like, there's a sentence that's not flowing quite right.

    00:10:40.000 --> 00:10:41.000

    I'm like, Hey, what's up with the syntax? Can you give me feedback here?

    00:10:41.000 --> 00:11:07.000

    The structure, the synonyms, and it's great for that as well. I also think it's really great for, like for summarizing large chunks of work, cause you know, you're in class and you're taking all these notes and you're like, okay, really consolidate your notes and really get to the key important stuff.

    00:11:07.000 --> 00:11:08.000

    So yeah, that's how I use it.

    00:11:08.000 --> 00:11:20.000

    Okay. And, did you like, well, okay, we'll come back to this question, but I'll come back to next 2, but we had a question from the audience about like, how did you learn how to prompt?

    00:11:20.000 --> 00:11:26.000

    Anna, like did you just do it through trial and error? Did you get support from your school or were you doing it on your own?

    00:11:26.000 --> 00:11:41.000

    It was trial and error, really. And then you find the more specific you are. That's kind of how I realize, okay, that's where I need to kind of direct it towards, because again, it's a tool, it's not perfect, but I did find that it really did help once they're trial and error.

    00:11:41.000 --> 00:11:48.000

    It's great. And Nick, just before I go to Parnava, how did did you take any classes or just you just did it by kind of figuring it out?

    00:11:48.000 --> 00:11:54.000

    Yeah, so my sort of introduction to prompt engineering was trial and air when I was sort of beginning to explore the technology.

    00:11:54.000 --> 00:12:20.000

    There is no classes available at Yale or available. I would say even any high school or college across the road just focus on AI literacy, but then as I sort of got more, engaged in the work, I started looking online for resources and I found empirically based resources and strategies to employ such as, you know, giving AI a persona, giving it really clear examples with one shot or a few shot prompting as well some other great

    00:12:20.000 --> 00:12:23.000

    That's great. And so just so everyone knows, one shot is very similar to our prompt library, which are gonna be very simple prompts with highly specific context.

    00:12:23.000 --> 00:12:31.000

    And then few shot is where you give it examples of what you want. So if a case for maybe Anna, you're giving an examples of some sentences you do like and then asking it to help you.

    00:12:31.000 --> 00:12:41.000

    Refine what you're doing. So that's really great. Okay, Pranava, how do you use it?

    00:12:41.000 --> 00:12:42.000

    I know you use it a lot and then like you learn how to do it.

    00:12:42.000 --> 00:12:53.000

    Yeah. Sure. So starting off with how I use it. So, The first thing is Chat GVT or I personally use Cloud.

    00:12:53.000 --> 00:13:00.000

    I much more, but it's almost become something like a second Google for me, right? Like Google is like, okay, cool.

    00:13:00.000 --> 00:13:11.000

    I'm gonna go on there and I'm gonna search something up that I need to like sort of figure out which websites I need to go from there and sort of dig deeper but I was currently working on some sort of research project.

    00:13:11.000 --> 00:13:20.000

    And for that, sometimes I come across some very, very specific questions. And even as a student, like when I'm learning, I come across very, very specific questions.

    00:13:20.000 --> 00:13:27.000

    And for that, Google provides too much of a general sort of list of websites that I can go to.

    00:13:27.000 --> 00:13:33.000

    So this is when I pivot and I use chat UPT and it's sort of gives me a much more concrete answer.

    00:13:33.000 --> 00:13:38.000

    Of course, I do have to fact check the information. That's something that I would definitely want to point out.

    00:13:38.000 --> 00:13:46.000

    Because Chat 2 is not always right. But that's been the sort of the main use. I'm Chat GPT and also brainstorming ideas.

    00:13:46.000 --> 00:13:51.000

    So I can shrink the time between, you know, me thinking of my own idea to me getting a whole list of ideas and just picking and bouncing off ideas from there.

    00:13:51.000 --> 00:13:58.000

    So those are in the 2 main uses and me getting into it. So when I first started, it was trial and error for sure.

    00:13:58.000 --> 00:14:05.000

    We didn't have like the resources that we're putting up in and education back then, which are really helpful now, but didn't have him back then.

    00:14:05.000 --> 00:14:16.000

    So it was just like, okay, sitting there and figuring out, okay, this worked, this didn't, this got me the right thing, this didn't.

    00:14:16.000 --> 00:14:21.000

    And like Anna said the more specific you are and the more specific that you get with what you want.

    00:14:21.000 --> 00:14:25.000

    It sort of feeds, better results, but yeah, it was definitely a trial and error.

    00:14:25.000 --> 00:14:30.000

    That's great. And actually, so there are some tools called generative search and so that kind of combine the 2 things.

    00:14:30.000 --> 00:14:33.000

    So Dick, do you want to talk about perplexity?

    00:14:33.000 --> 00:14:45.000

    Yeah, absolutely. So, perplexity is a generative search tool. So essentially the knowledge in which it, pulls in order to answer your questions or all, you know, resources online.

    00:14:45.000 --> 00:14:51.000

    It's sort of a Google, but just infused with a lot more, you know, a much clearer database.

    00:14:51.000 --> 00:15:11.000

    That's not just synthesizing information in a very in the same way Chat Tp does and then on along the research lines there are also some great tools such as Skype, Elicit and Consensus, which pull data from empirically and peer reviewed empirically based articles and peer reviewed articles as well.

    00:15:11.000 --> 00:15:16.000

    That's great. And what I'd say is what we'll do is we'll make sure that we will include those at like when we get the Q&A.

    00:15:16.000 --> 00:15:25.000

    So maybe and we'll include that in the email after as well. So these are great. And then the insight that Parava uses cloud AI.

    00:15:25.000 --> 00:15:26.000

    Do you want to

    00:15:26.000 --> 00:15:29.000

    Yeah, yeah, essentially it's like another version of Chat. It's just little more up to date.

    00:15:29.000 --> 00:15:35.000

    I found it to be a more up to date more recent. So cloud to I. It's spelled CL A UD.

    00:15:35.000 --> 00:15:37.000

    Oh, cla, clad, okay. I was like, I don't know what clock.

    00:15:37.000 --> 00:15:42.000

    Clawed. That maybe I'm saying it wrong. Claude, sorry, my apologies.

    00:15:42.000 --> 00:15:43.000

    No, no, you're good. Cause I I thought you were, I thought you were telling me something that I was like, oh, and actually clock.

    00:15:43.000 --> 00:15:48.000

    Maybe I'm saying it wrong, but, Yeah, it's Claudia.

    00:15:48.000 --> 00:15:49.000

    Was yesterday. Yeah, it's fine now.

    00:15:49.000 --> 00:15:54.000

    I think Yeah, it came up with the new, update. Huge update. Really cool.

    00:15:54.000 --> 00:16:12.000

    And I also wanted to point out Gemini, that also like Google release something where they also have like a wafer, you can click a little Google icon on Gemini when you put in a prompt and it can give you a bunch of like where the sources were from what you put in so I found that to be very useful too when I was you know doing the project.

    00:16:12.000 --> 00:16:16.000

    But yeah, those 2, I'd say I'm mainly on now, yeah.

    00:16:16.000 --> 00:16:18.000

    Yeah, that makes way more sense, cause I was like, I don't, I have heard that one.

    00:16:18.000 --> 00:16:20.000

    Yeah, yeah, yeah, Anthropic. They're incredible.

    00:16:20.000 --> 00:16:31.000

    I have spent a lot of time learning about this stuff. But no, that's great. And so actually it's really interesting that because I think that one of the questions we had is about the ethical component.

    00:16:31.000 --> 00:16:45.000

    And so we'll talk, do one more about you all before we move to like the larger let's let's go out of lens but like the larger let's let's go out of lens but like how do you so we talked about perplexity we talked about being more up to date like Claude so you know their knowledge cutoffs which is amazing so you've built some really strong literacy but how

    00:16:45.000 --> 00:16:55.000

    do you like how do you verify like we have our every framework where we talk about verifying the output right that you're the owner, like you can't say chat to me, you have to be the one that owns that.

    00:16:55.000 --> 00:17:05.000

    So how do you how do you own the way that you ensure that. So how do, how do you own the way that you ensure that the thing you're submitting or learning about, isn't, isn't hallucinated or made up?

    00:17:05.000 --> 00:17:06.000

    Nick, do you want to go first?

    00:17:06.000 --> 00:17:10.000

    Yeah, absolutely. I think this is a, you know, a literacy is perhaps, you know, becoming one of the most important skills that anyone could have.

    00:17:10.000 --> 00:17:24.000

    And interestingly, as you know, students when we enter the workforce, a lot of employers, and we're looking at these job hosting us that we have a literacy and prompt engineering skills.

    00:17:24.000 --> 00:17:37.000

    So for a framework that I employ, I like to look, you know, for sort of hallucinations and biases and you sort of could put them together but for for hallucinations what I do is every time there's specifics, you sort of could put them together.

    00:17:37.000 --> 00:17:40.000

    But, as for, hallucinations, what I do is every time there's specific claim or argument that's not general knowledge, I go and verify that.

    00:17:40.000 --> 00:17:45.000

    So I'll use Google, I use perplexity and find that claim online. For biases, those can be a little bit more nuanced and sort of embedded.

    00:17:45.000 --> 00:17:56.000

    Into the core concepts, but I'll ensure that it's important to note, of course, all AI models are trained on human input and humans are biased.

    00:17:56.000 --> 00:18:02.000

    You know, So therefore AI is biased. Doesn't emerge out of thin air.

    00:18:02.000 --> 00:18:07.000

    But yeah, I, definitely verify each claim. Using external sources.

    00:18:07.000 --> 00:18:16.000

    Yeah. And I mean, the cool thing about this group is that we're, you know, you all know that like even generative search is that we're, you know, you all know that like even generative search is going to mess up, right?

    00:18:16.000 --> 00:18:17.000

    These things, there's no full proof way because these are going to mess up, right? These things, there's no full proof way because these technologies are always hallucinating.

    00:18:17.000 --> 00:18:25.000

    We just they're just wrong sometimes and we call it a hallucinating. We just they're just wrong sometimes and we call it a hallucination or a fabrication.

    00:18:25.000 --> 00:18:27.000

    So I think that this is something that, and we call it a hallucination or a fabrication.

    00:18:27.000 --> 00:18:29.000

    So I think that this is something that though to your point, Nick is that we do like the AI.

    00:18:29.000 --> 00:18:37.000

    So I think that this is something that though to your point, for fabrication. So I think that this is something that though to your point, Nick, is that we do like the AI literacy component, which we'll come back around to is so important because you have to learn to be critical users.

    00:18:37.000 --> 00:18:40.000

    Like, and it's so easy to like take your eye off the ball or fall asleep at the wheel.

    00:18:40.000 --> 00:18:41.000

    And I think that this is why we'll, you know, we're gonna come back around to like what you think needs to happen to build that literacy.

    00:18:41.000 --> 00:18:51.000

    And I think that this is why we'll, you know, we're gonna come back around to like what you think needs to happen to

    00:18:51.000 --> 00:19:00.000

    So very similar to Nicholas. If I ever have, like, if it ever makes a claim, I'm always first to prove to check it and make sure it's correct.

    00:19:00.000 --> 00:19:13.000

    But personally what I do is I use it more I've more so find my own work and then use like AI tools and platforms really to kind of enhance my work.

    00:19:13.000 --> 00:19:25.000

    So I't say that I'm kind of asked prompting it for the information I'm more looking for myself than helping me like modify it, condense it if needed.

    00:19:25.000 --> 00:19:30.000

    So, but I always like proofread it, make sure that I'm fact checking it.

    00:19:30.000 --> 00:19:33.000

    And like that's a, like I love that though because what you're doing is you're actually choosing a use case that you feel confident in.

    00:19:33.000 --> 00:19:49.000

    And that is also really interesting is that you're kind of the same way that Pranava, you were talking about the fact that you use it as like a secondary, like you pivot to it, but you're not always going there first.

    00:19:49.000 --> 00:19:50.000

    Okay.

    00:19:50.000 --> 00:19:52.000

    I think it's really a fascinating opinion too is that you've learned that it's good for a thing, right?

    00:19:52.000 --> 00:19:59.000

    And that's the thing that you use it for, which I think is really important to recognize too is that it's not necessarily at all time.

    00:19:59.000 --> 00:20:02.000

    Everything tool yet, but a more specific like.

    00:20:02.000 --> 00:20:09.000

    And whenever I'm like presenting it, one of the key things that I always say, it's there to support student work not replace it.

    00:20:09.000 --> 00:20:11.000

    So that's kind of like my approach with it. I'm also like very cautious when it comes to that.

    00:20:11.000 --> 00:20:18.000

    But yeah, that's just how I approach it.

    00:20:18.000 --> 00:20:19.000

    That's great. Paravid, do you want to anything?

    00:20:19.000 --> 00:20:28.000

    Yeah, sure. So I've actually I've went into some of these chat lots just for fun just to see what we're like where the hallucinations does sort of happen.

    00:20:28.000 --> 00:20:39.000

    I found it does happen a lot with math. And there are some bots out there that are really good with math, but, you know, with Claude, at least it's earlier models and you've been chatting with D like 3.

    00:20:39.000 --> 00:20:46.000

    When I input a certain math question that I know the answer to, it generally, sometimes doesn't give the right answer.

    00:20:46.000 --> 00:20:48.000

    And what I find myself doing is, typing in, Hey, I think you might be wrong on this one.

    00:20:48.000 --> 00:20:52.000

    Can you double check it? And when I do that, then it rechecks and comes up with a different output.

    00:20:52.000 --> 00:21:04.000

    That's sometimes right in sometimes still wrong. So it was just a very interesting sort of experiment, that you can also try.

    00:21:04.000 --> 00:21:12.000

    But when I usually, you know, get sort of hallucinations and things, first I like to check it with the bot itself just to see what would happen.

    00:21:12.000 --> 00:21:15.000

    And then of course cross reference with Google and actual sources online just to make sure that everything's sort of, you know, right.

    00:21:15.000 --> 00:21:24.000

    In a sense. And so, Yeah, I guess similar to what Nick and to do.

    00:21:24.000 --> 00:21:30.000

    That's great. And just as a reminder, right, the reason why it doesn't do math is it's not a calculator.

    00:21:30.000 --> 00:21:39.000

    Yeah.

    00:21:39.000 --> 00:21:40.000

    Yeah.

    00:21:40.000 --> 00:21:47.000

    And it's not a thinking machine, right? And so if it gets it right, it gets it right because of like the training data and the predictive qualities and so may sense the kind of convert time zones it doesn't know time zones in a real way right and it's not able to do that math and that there are tools that are going to be better for that.

    00:21:47.000 --> 00:21:55.000

    But I love that you took that experimental fact like thing because I think this moment of seeing it not work next time you do this you're not going to use it for your math homework, right?

    00:21:55.000 --> 00:22:02.000

    No, definitely.

    00:22:02.000 --> 00:22:03.000

    And

    00:22:03.000 --> 00:22:05.000

    I, anything you're gonna use it in other ways that you feel more confident in the output. So I think that's a great way to kind of show that that that trust that needs to be built.

    00:22:05.000 --> 00:22:11.000

    Yeah, and sort of the easier way I can put it is kind of like if it did get the bath problem, right?

    00:22:11.000 --> 00:22:19.000

    It's almost as if it's seen that math problem before. Somewhere online. And that's where it can sort of draw that conclusion and say this is what I think is right.

    00:22:19.000 --> 00:22:28.000

    And so that's what's happening behind the scenes, but if it hasn't seen it or if there's a problem that's a little more odd or slightly worded different, it'll get it wrong.

    00:22:28.000 --> 00:22:32.000

    And then of course you can ask a double check and verify online. So, you know.

    00:22:32.000 --> 00:22:43.000

    Absolutely. And so those training data sets. So if, if you're following the way they benchmark, these tools, they, sometimes they benchmark it on, you know, questions that are publicly available.

    00:22:43.000 --> 00:22:53.000

    That we're most like in the training data set, but sometimes they like the the better invention marking is when they take these assessments that have not been online and that are new and that way they can actually their novel questions and that's when you can actually benchmark how good these models are.

    00:22:53.000 --> 00:23:01.000

    Okay, so next we're gonna go, we're gonna come out a little bit of a lens because you guys are interested in education.

    00:23:01.000 --> 00:23:09.000

    So what is happening at Stanford? What is happening at Middleton and in Orlando? Like, what's happening around you?

    00:23:09.000 --> 00:23:13.000

    Because you guys are early adopters, your first movers, you know, so I know that's not the case.

    00:23:13.000 --> 00:23:17.000

    So Nick, do you want to talk about what's happening at Stanford?

    00:23:17.000 --> 00:23:20.000

    Yeah, so, I'm at Yale. So at Yale, you're good.

    00:23:20.000 --> 00:23:22.000

    Oh god, sorry. Oh, man. Sorry, I saw you at Stanford.

    00:23:22.000 --> 00:23:27.000

    It's the, it's the Golden Gate Bridge behind it. Yes, yes.

    00:23:27.000 --> 00:23:35.000

    So at Yale, I'm one of the rules that I, that I have there is, a, I, ambassador, for the teaching and learning center.

    00:23:35.000 --> 00:23:41.000

    So I work closely with faculty to innovate their pedagogy to responsibly incorporate AI. So one of the things that we're really pushing for here is universal access to cutting edge.

    00:23:41.000 --> 00:23:58.000

    Foundational model such as GPT for cutting edge foundational models such as GPT 4 or cloud 3 in order to ensure that GPT 4 or cloud 3 in order to ensure that all students, foundational model such as GPT 4 or club 3 in order to ensure that all students despite their socioeconomic status as well as faculty, right?

    00:23:58.000 --> 00:24:08.000

    Because it applies to faculty as well that they all have access to it. But I think one of the things that I'm really seeing on this sort of equity point is there a lot of professors who outright reject AIs relevance in academia at all.

    00:24:08.000 --> 00:24:14.000

    They don't even want to consider it. I heard it, you know, speaking to Professor the other day, he said it was just a plagiarism machine, right?

    00:24:14.000 --> 00:24:24.000

    I think that is born out of a place of fear and it, you know, it's sort of, I, you know, I have to admit ignorance even for how AI works in its sort of place and potential in the classroom.

    00:24:24.000 --> 00:24:27.000

    On the other hand, there are a lot of other professors. Who have taken a very, you know, growth mindset oriented approach.

    00:24:27.000 --> 00:24:42.000

    I've started, you know, exploring AI through hands-on use, doing a literacy tasks in their courses, infusing AI for problem-based learning, as well as other great strategies.

    00:24:42.000 --> 00:24:55.000

    So I think one of the things I'm really concerned about right now is this sort of inequity in classrooms that's perpetuated or caused by both the technology and the AI, level literacy that that the professor or teacher has.

    00:24:55.000 --> 00:25:00.000

    Well, that is amazing. And yeah, I think that we talk a lot about how the this is an equity issue, right?

    00:25:00.000 --> 00:25:17.000

    That some, students are gonna get a leg up and some are gonna get stopped and even potentially at Yale like just like professor by professor or by you know English department versus computer science department versus you know like that even the types of.

    00:25:17.000 --> 00:25:30.000

    Majors that you have could really impact your knowledge before you go into the workforce. I don't know, I'm sure this group is aware, but you know, the IMF says about 70% of US jobs will be impacted by Genera Bay in the US.

    00:25:30.000 --> 00:25:39.000

    And it's by far the highest in the world. And so I think that's something for you all like as educators but also as students is going to be a big impact for you all.

    00:25:39.000 --> 00:25:50.000

    So Anna, what about you? Because I know I think it's really interesting because you've already kind of signaled this a bit already that you're really cautious about getting started and really cautious about what the line is for ethical use.

    00:25:50.000 --> 00:25:54.000

    Is that something that you're seeing at school that makes you a little bit cautious?

    00:25:54.000 --> 00:25:58.000

    So yeah, great question. My school I actually have to say like with our cybersecurity like team and whatnot, they are really like invested into this program.

    00:25:58.000 --> 00:26:15.000

    Like I've spoken with them, they're fully supportive of my student voice. But when it comes to it on a teacher to teacher basis, yeah, they're exactly what I've like Nicholas was saying.

    00:26:15.000 --> 00:26:24.000

    There are some teachers that are very like completely reluctant to that. And not even just reluctant flat out, like projecting any possibility of it.

    00:26:24.000 --> 00:26:27.000

    So when it's come to, when I come to them approaching them with my AI lectures, I flat up until no, not interested.

    00:26:27.000 --> 00:26:38.000

    And or if they do say yes, I have to be very careful with how I respect their classroom rules.

    00:26:38.000 --> 00:26:50.000

    But as a whole, my school has done a fantastic job in supporting me with that. But I will say it has been a lot of student initiative of getting the conversation started, which I think is really interesting.

    00:26:50.000 --> 00:27:00.000

    And of course you have some amazing teachers. And faculty members who are there supporting and clearing the way for students like me who are interested in it.

    00:27:00.000 --> 00:27:07.000

    But yeah, exactly. I am a little bit cautious with that. Because I'm aware of both sides of the argument and I want to appeal to.

    00:27:07.000 --> 00:27:11.000

    Like mind the entire faculty. So.

    00:27:11.000 --> 00:27:14.000

    Do you think, do you think students are using it like without not like, like how do you think the students are using it?

    00:27:14.000 --> 00:27:19.000

    At your school?

    00:27:19.000 --> 00:27:42.000

    Students at my school, I will say that they are like familiar with it, obviously. The student, I will say this, the students who are using it academically like, dishonestly, we're already doing that in previously without AI, which is one of the big things like people ask me like, oh, isn't this going to encourage cheating?

    00:27:42.000 --> 00:27:56.000

    I say that the people who already were cheating were doing it without AI, so that's kind of how I view it, but there are other kids like me who are interested in using it in a academic, in a responsible and educated way.

    00:27:56.000 --> 00:28:03.000

    So I think that it's also, fostering that in excitement for it because I do think it exists.

    00:28:03.000 --> 00:28:04.000

    In a responsible.

    00:28:04.000 --> 00:28:17.000

    Yeah, absolutely. And we see from this is actually at Stanford. So Victor Lee, I was on a panel with Victor where I'm a couple of weeks ago, Victor Lee is the professor that's leading the cheating.

    00:28:17.000 --> 00:28:26.000

    He's been doing a study on cheating for the last 15 years and's been doing a study on cheating for the last 15 years and found that 60 to 80% of, you know, students have cheating for the last 15 years and found that 60 to 80% of, you know, students have cheating behaviors.

    00:28:26.000 --> 00:28:56.000

    And so there was an expectation that going back into the classrooms in 2023 when we've seen other types of behaviors that have increased like the fact that internet scams went up, like the fact that internet scams went up, 1,200% in the first year of chat.

    00:29:03.000 --> 00:29:12.000

    Yeah, so, we actually have a local university, University of Florida, and they're creating their own like AI literacy programs, they're putting out to high schools.

    00:29:12.000 --> 00:29:22.000

    And I'm right now currently just helping them sort of implemented into our own high school. So we're actually getting a AI literacy program as a part of one of our, magnet classes.

    00:29:22.000 --> 00:29:29.000

    At Middleton high school so that's been really great. And now going into the use cases that I see around me, with students, I see a full under 3 camps.

    00:29:29.000 --> 00:29:38.000

    One is the students that are completely like, no I'm not gonna use that at all. It's just something that I don't wanna get into.

    00:29:38.000 --> 00:29:41.000

    I don't even understand how it works. I'm good 2 is, well, I'm going to use it responsibly. I know how this thing works.

    00:29:41.000 --> 00:29:49.000

    I researched it and I think it's, you know, some that I could use for certain things.

    00:29:49.000 --> 00:29:56.000

    And that's a very small minority group too, but group 3 is this thing is the best thing ever.

    00:29:56.000 --> 00:30:00.000

    I'm gonna use it for everything. And so what I find is that the people who are using it for everything, they know the least of how it actually works.

    00:30:00.000 --> 00:30:15.000

    And when you find out, like, just like I was saying, the Air Hoo's nations, they I bias when these things start coming up and AI doesn't do everything, then they sort of find the like, oh wait, that that's not working how I wanted it to work.

    00:30:15.000 --> 00:30:21.000

    And it's sort of little lag there. And then I do wanna get on like cheating real quick.

    00:30:21.000 --> 00:30:29.000

    So And what Anna was saying is completely correct, right? The kids who were cheating before, they're gonna continue to cheat now and just in different way.

    00:30:29.000 --> 00:30:39.000

    And all the checkers and everything out there, sure. They work for a little bit. Once next update hits, they're not gonna work as well anymore.

    00:30:39.000 --> 00:30:48.000

    Like check, GPG zeros out there. It's one of the best checkers, but even it's struggling to sort of determine AI versus human and Kids who want to cheat will find a way to do it.

    00:30:48.000 --> 00:30:57.000

    But I would still believe the I could be used as a very, very ethical tool. And can be brought to education in a very, very nice way and we can reduce.

    00:30:57.000 --> 00:31:02.000

    That cheating just overall as like, let's stop students cheating instead of students cheating with AI.

    00:31:02.000 --> 00:31:03.000

    You know what I mean? So. Yeah, I guess that's what I'd say.

    00:31:03.000 --> 00:31:04.000

    GBT that there would be another rise in cheating. 2 things. Number one, I have a funny story, but I'll get back to that.

    00:31:04.000 --> 00:31:19.000

    But the second thing is AI detectors, everybody. If you're in this audience, I am sure you know us and you know that we do not support AI detectors as anything that is a final word.

    00:31:19.000 --> 00:31:27.000

    We know that there are frost positive and false negative rates, and that they are can be potentially biased against non non-negative speaker.

    00:31:27.000 --> 00:31:31.000

    So they're definitely not there if anything a little bit of a tool kit but if you're an educator out there and your kid suddenly comes in with a different vocabulary.

    00:31:31.000 --> 00:31:44.000

    I joke about it's like they lived in, you know, London for 2 weeks, they come back with an accent in the monocle and a cane, like you know that they've changed their whole voice that usually you're able to identify that.

    00:31:44.000 --> 00:31:57.000

    But my funny story is I was in at school 2 weeks ago everybody and Someone was accused of eye cheating and I called their parents and the parent said my student didn't use AI because I helped them with their homework.

    00:31:57.000 --> 00:32:13.000

    So old school teaching is okay, but, so you know, I think that was a good anecdote that the stuff is happening and that it is something that it's gonna require some rethinking of like why students cheat, you know, that goes beyond just this moment in time.

    00:32:13.000 --> 00:32:22.000

    So we've come around this a lot and I think that you know this idea of AI literacy building there's actually a ton of really great questions that kind of come in and around here.

    00:32:22.000 --> 00:32:30.000

    So we know that we want students to use these tools responsibly. We also want to know that teachers feel confident having these students use these tools responsibly.

    00:32:30.000 --> 00:32:37.000

    So Nick, like when you think about what you're doing at Yale, like how do you build AI literacy and students like what has to happen?

    00:32:37.000 --> 00:32:41.000

    What would you like to happen at Yale to get to a place where you're getting out of this.

    00:32:41.000 --> 00:32:45.000

    You know, kind of push and pull that's you see happening around you.

    00:32:45.000 --> 00:32:53.000

    Yeah, absolutely. I think it's excellent question. I think the first step in that is ensuring that we're transparent about these.

    00:32:53.000 --> 00:33:04.000

    Conversations that there's not taboo or around talking about AI so all professors address it and you know first day of class or whatever that looks like.

    00:33:04.000 --> 00:33:18.000

    And another thing is you know ensure that you know there's that all professors go through rigorous process in which they consider AIs benefits and which they consider AIs benefits and limitation sort of as it's own thing and then as well as okay rigorous process in which they consider AIs benefits and limitation sort of as its own thing.

    00:33:18.000 --> 00:33:21.000

    And then as well as, okay, would it be a good idea to integrate in my course? Can I do a pilot?

    00:33:21.000 --> 00:33:26.000

    You know, can I see how this works? I mean, then going back to the sort of inequity part, you know, ensuring that all students have access to the same technology, right?

    00:33:26.000 --> 00:33:38.000

    Because, you know, some students are able to afford that $240 a year for chats you can keep it others that's you know that is not something that is feasible for them.

    00:33:38.000 --> 00:33:47.000

    As well as just ensuring that there's clear AI policy coming from the top, right? That's flexible enough so that professors have agency over how they incorporate it.

    00:33:47.000 --> 00:34:08.000

    Yeah, at the same time has some clear guidelines, right, about academic integrity and privacy. And security, which is another huge thing that we haven't even talked about, but you know, what students put, into the bots and, you know, is that going to get trained or is, you know, company and easy to train as well as what, you know, professors or teachers put in the bot, you know, that could break if

    00:34:08.000 --> 00:34:15.000

    they put personal person identifiable, student information that could break, but and other regulations there.

    00:34:15.000 --> 00:34:20.000

    And so amazing Nick and all those things, but like we had a question about A. Literacy.

    00:34:20.000 --> 00:34:21.000

    So what like when we think about AI literacy, you talked about capabilities and limitations, you talked about appropriate responsible use.

    00:34:21.000 --> 00:34:35.000

    So the other things that you would want to see if a, like if a student was going to be AI literate or a teacher or a faculty member, what would they know?

    00:34:35.000 --> 00:34:47.000

    Yeah, I would say, they would know how to effectively parse, you know, AI generated attacks and see where the biases and hallucinations are and they would also know AI tools be on just shot GPT.

    00:34:47.000 --> 00:34:59.000

    They would know about mid journey. They would know about. It was like synthesia which is does you know video generation and also that they would know how or what a responsible way looks like in terms of for example project based learning.

    00:34:59.000 --> 00:35:09.000

    So there's a great example when you know I was at the Stanford conference that's where a man and I met in person and one of the professors there Chris Peach who teaches introduction to computer science.

    00:35:09.000 --> 00:35:29.000

    He told us how he reinvented his course. So like before you still teach his students the fundamentals really really well but then after that he enables all his students to use as much center of AI as they would like whether that Chat KPT or Thanasia or mid journey in order to create really cool projects that resonate with their own interest right and what that's doing is that's enabling

    00:35:29.000 --> 00:35:33.000

    student agency and creativity and really getting to the heart of why we even learn.

    00:35:33.000 --> 00:35:46.000

    Yeah, absolutely. And I'm just gonna add a little bit of a point. I think Just foundationally understanding that artificial intelligence is not intelligent and that these are computing probabilistic like models and not thinking models.

    00:35:46.000 --> 00:36:00.000

    I think that that is probably, I see that happen all the time that because we call it AI and because it has it could have a personality and it's first person and even can have a name that people tend to trust these tools as if they are thinking.

    00:36:00.000 --> 00:36:05.000

    And so I think that's an underlying piece too that there's just a step back that we need to get to.

    00:36:05.000 --> 00:36:17.000

    So that's really great. Anna, what do you think needs to happen? Especially from a high school perspective and you also work with elementary school students so I'd love to know what you think needs to happen in terms of building that AI literacy.

    00:36:17.000 --> 00:36:30.000

    Yeah, of course. Well, like, Nicholas said, I think teachers leading that conversation and becoming familiar with, AI platforms outside of Chat GPT is fantastic.

    00:36:30.000 --> 00:36:39.000

    And that, the teachers who are a little bit hesitant because I find when you learn more about something more well informed you are about it and the less we fear it.

    00:36:39.000 --> 00:36:45.000

    And then I also think that, and then explaining to students, here's the right way to use.

    00:36:45.000 --> 00:36:53.000

    I think having a conversation and setting those boundaries, those limitations, but also allowing for there to be a space where AI can be used.

    00:36:53.000 --> 00:36:57.000

    I think that them and also explaining to them, okay, you can't use it or not necessarily you can't use it.

    00:36:57.000 --> 00:37:16.000

    But you need to be. Cautious with it and here's why explaining why there are the concerns rather than just outright banning it and then explaining to them here are the risks with it such as the databases, your, not just your academic integrity, but also your intellectual property.

    00:37:16.000 --> 00:37:24.000

    But saying there is a way that it can be used and here is how you can use it. So also providing students the responsible way.

    00:37:24.000 --> 00:37:26.000

    To use it as well in their classrooms.

    00:37:26.000 --> 00:37:35.000

    Absolutely. All of those things are important. And I think we think of it as like, like the idea of transparency in itself is really important for ethical use.

    00:37:35.000 --> 00:37:44.000

    We have to feel confident that we can talk about it and not hide it. And if we can do that, then I think it really allows us to have better conversations to your to your point, which is great.

    00:37:44.000 --> 00:37:48.000

    And then last thing is, Pranava, what do you think? Like, what do we have to do?

    00:37:48.000 --> 00:38:00.000

    Especially now that you're working with us, like, but like we've talked to some Pranava did this whole thing where you put down like everything you could think we could do and did some great work on thinking about how to bring student voice into that literacy conversation.

    00:38:00.000 --> 00:38:02.000

    So what do you think we need to do?

    00:38:02.000 --> 00:38:04.000

    Yeah, for sure. So initially I guess we can just go back to the very, very head of like a classroom.

    00:38:04.000 --> 00:38:20.000

    So if we're bringing in artificial intelligence to a classroom, the person needs to know first is the teacher and like Nick and Ana mentioned The teacher is the one who's sort of distributing it and sort of explaining to the classroom.

    00:38:20.000 --> 00:38:26.000

    So teacher workshops, I'm letting you just sort of understand what kids are doing right now.

    00:38:26.000 --> 00:38:36.000

    And this, workshop itself is a great example of that. Anything to get to teachers more on the same page with what's going on with AI and what's going on with students and different things like this.

    00:38:36.000 --> 00:38:44.000

    That can definitely help to just make a decision on. How to integrate AI into education because let's be real for a second.

    00:38:44.000 --> 00:38:49.000

    AI is not going away. It's not something that's just going to stay here for a couple of years and then just take a backseat.

    00:38:49.000 --> 00:38:55.000

    It's here, it's happening, so we might as well implement it in a very, very responsible way.

    00:38:55.000 --> 00:38:59.000

    And the next things comes down to the students. And I believe the biggest thing here is knowing how it works.

    00:38:59.000 --> 00:39:09.000

    See, at the very beginning when I was using AI, I had no clue how it works. So I'd use it for like all sorts of things thinking that'll give me the right stuff all the time.

    00:39:09.000 --> 00:39:21.000

    But once I sort of understood that, you know, these models, they work based on data sets and, and that entails some like biases and hallucinations there, something's not in the data set and things like that.

    00:39:21.000 --> 00:39:28.000

    Once you understand how it works, the back end of it, I've definitely become much more responsible within the way that I use it for sure.

    00:39:28.000 --> 00:39:38.000

    I'm using it for more things that I know are going to be correct. And so that AI literacy aspect is very, very key for students.

    00:39:38.000 --> 00:39:42.000

    If we can teach that to the teachers. Then it can all flow top down. And so that's the way I see with that.

    00:39:42.000 --> 00:39:47.000

    That's great. So okay, so I have a question. Have any of your teachers use JI in a cool way that you're aware of?

    00:39:47.000 --> 00:39:57.000

    Like in your, like, like have you seen anybody at Yale or that's done something cool that you want to share?

    00:39:57.000 --> 00:40:00.000

    Yeah, so, I was, I was just talking about this earlier actually, in an earlier panel I did here at Yale.

    00:40:00.000 --> 00:40:10.000

    So there was a great example of a professor at Yale in the school of the environment who you know, is like every year he has his students to a research project at the end.

    00:40:10.000 --> 00:40:28.000

    So what he did was how to students input their research project, what they've run so far into chat GPT and have a good feedback on the work they were doing and the sort of comments they had as well as after it to bring new angles that they could consider.

    00:40:28.000 --> 00:40:42.000

    Sort of attacking this problem or confronting this problem as well as some counter arguments, right, that they could also address, which were, you know, of course, augment the quality of the research project.

    00:40:42.000 --> 00:40:51.000

    Another great example is Yale, we actually, in one of my courses we're integrating internal, GPT.

    00:40:51.000 --> 00:40:56.000

    So, train exactly on the, on the data of the course, which is really cool.

    00:40:56.000 --> 00:41:06.000

    And then we're going to be doing in the course as corner with my professor about this. We're going to be doing an activity where my professor is the one, you know, writing the prompts into the computer, but the prompts that we, that, you know, he ends up writing are all guided by the students.

    00:41:06.000 --> 00:41:17.000

    So right, really engaging students with bats and they see it and while he's doing that he's going to explain about how it works at a foundational level and things of that nature.

    00:41:17.000 --> 00:41:25.000

    So just really engaging it, bringing it to the forefront of the class rather than being taboo, which I feel like is a common thread between what we've all set here.

    00:41:25.000 --> 00:41:33.000

    Absolutely. And I, you know, I think one of the things that's really interesting is that you probably, it's like fasting to see like the good stuff, right?

    00:41:33.000 --> 00:41:40.000

    But then also you probably have the ones that are like, you know, pit and paper. We're gonna start doing out like, you know, the blue books again, back when I was a student back in the dark ages everybody.

    00:41:40.000 --> 00:41:52.000

    And so we have like those kind of reactions as well. So what about you, Anna? Has anybody like have any of your teachers actually done something cool with you all?

    00:41:52.000 --> 00:41:58.000

    I wish it could I could say that I have a really cool story. But unfortunately, no, not a lot of them have.

    00:41:58.000 --> 00:42:10.000

    There was one teacher for a paper we had to write. She introduced us to Gemini and that one and she specifically used it for providing feedback on our thesis like we could.

    00:42:10.000 --> 00:42:27.000

    Ask it for information for inspiration, but outside of that, that was really the only one instance. And as I said before, it's been a lot of student like led initiatives being like, we want this to be a part of it and then reaching out to like the cyber security team who have their own.

    00:42:27.000 --> 00:42:33.000

    Goals as well. But yeah, like in individual classrooms, not a lot, no.

    00:42:33.000 --> 00:42:43.000

    Do, do you think, Anna, that's me, and I can say your name, Anna, do you think that is because You guys, your the kids are so much more ahead of the teachers is that what's happening is that like you guys are kind of leading the charge.

    00:42:43.000 --> 00:42:51.000

    And teachers are kind of following behind a bit.

    00:42:51.000 --> 00:43:01.000

    I would say there's some teachers that I would say are genuinely just as passionate. And really do want to incorporate this into their curriculum.

    00:43:01.000 --> 00:43:14.000

    But then I, like you said, I do think there are some teachers that are also a bit hesitant just because they're not aware or not like fully educated on the subject creating that bit of kind of like hesitance with it.

    00:43:14.000 --> 00:43:22.000

    So exactly as we've been saying before, I think if it is, it isn't this taboo conversation that we really do bring it up and we use in the right way.

    00:43:22.000 --> 00:43:37.000

    I think that it doesn't have to just be a student. Situation it can be something for both teachers and students to collaborate on but for the time being yeah does at times feel like a lot of student.

    00:43:37.000 --> 00:43:41.000

    Absolutely. And so what about you, Pranavo? Has anybody used it?

    00:43:41.000 --> 00:43:47.000

    Yeah, I've actually had this one really cool guy, his, He's always up to date with everything that's going on now and things like that.

    00:43:47.000 --> 00:44:10.000

    He's very like, you know, looking at the news all the time, checking out what's the newest in tech and this type of thing and he's an aerospace engineer teacher and he was teaching us like aerospace engineering this type of thing we're actually building rockets just little model rockets and he was like you know I want you guys to go into mid journey and design your own rockets and see

    00:44:10.000 --> 00:44:16.000

    what it'll actually look like and show me, you know, your final project, that'll be a certain grade.

    00:44:16.000 --> 00:44:22.000

    And I was like, dude, that's really cool, right? I was like a whole like 30 min of everyone just sitting there looking at everyone's pockets.

    00:44:22.000 --> 00:44:27.000

    It was amazing. But besides that, Not much else. It hasn't really been used anywhere else that I've seen at least in a creative manner.

    00:44:27.000 --> 00:44:42.000

    And, and I guess like on a said that's due to the curriculum also just because in our current code, like public school curriculum, as not really a big thing, as it's just sort of getting introduced.

    00:44:42.000 --> 00:44:52.000

    Even to my CSIT teachers and different things like that. They're currently teaching like, you know, networking, how do computers talk with one another, you know, Python basic languages, Java.

    00:44:52.000 --> 00:45:06.000

    But I think as becoming a bigger thing now. So slowly we can continue to integrate, you know, lesson plans like we're doing an education into the public school system, which I think will really help.

    00:45:06.000 --> 00:45:11.000

    But other than that, one little incident, not much else as it's just not a part of the system yet.

    00:45:11.000 --> 00:45:29.000

    It's really interesting. I mean, I think we see at AI for education, right? We do see a bit of like a push and pull that there are maybe some early doctors that are that are like either students or teachers and it kind of goes back and forth but where there's a big middle of the bear bell curve of people that don't quite know what the next step is, whether their students or teachers, which is really interesting.

    00:45:29.000 --> 00:45:31.000

    Yeah.

    00:45:31.000 --> 00:45:35.000

    So I know we're at time every once if you like want to hang out we have a ton of good questions.

    00:45:35.000 --> 00:45:43.000

    I've been trying to filter them to everybody before we go. But we just want to say thank you for everyone that's joined and and Dan will drop everyone's LinkedIn.

    00:45:43.000 --> 00:45:47.000

    I think I, Anna, did I make you get a LinkedIn or did you already have one?

    00:45:47.000 --> 00:45:48.000

    Oh no, I had one.

    00:45:48.000 --> 00:46:00.000

    Good, good. Yeah, so young people with LinkedIn doing amazing things. But what I wanna do is before we go to the questions, I want to do a rapid fire of like, what is your advice to an educator that is thinking about getting started and embracing Generative AI and how it can support student learning.

    00:46:00.000 --> 00:46:13.000

    Like, what do you think, Nick? What is your, what is your one key advice before we go to the questions?

    00:46:13.000 --> 00:46:23.000

    Hey, take the step to change and innovate pedagogy in order to support students in developing critical thinking and creativity skills.

    00:46:23.000 --> 00:46:26.000

    And so snappy, love it, Anna.

    00:46:26.000 --> 00:46:32.000

    I think just have the conversation with your students. I think that it there are a lot of students there who are excited to use this platform in a responsible manner.

    00:46:32.000 --> 00:46:42.000

    And if you explain to them the reasons why there might be some initial hesitations, but the way that you can use it in the right way.

    00:46:42.000 --> 00:46:47.000

    I think that there's endless possibilities that can exist there.

    00:46:47.000 --> 00:46:48.000

    And Panava.

    00:46:48.000 --> 00:46:52.000

    Yeah, and I'd say, just Get familiar with AI, you know, sort of integrate it, you know, see if you can put it into your lesson plans and stuff.

    00:46:52.000 --> 00:47:01.000

    And an education definitely has a bunch of things there. But you know just go online, see what you can do to bring it in.

    00:47:01.000 --> 00:47:06.000

    There's a lot of Reddit forms and everything out there that just give you a bunch of information about it.

    00:47:06.000 --> 00:47:08.000

    And then take leap.

    00:47:08.000 --> 00:47:20.000

    Great. Okay, so we have this a couple of minutes because we have like all the things, but I'm gonna ask the one question we didn't quite get to and first of all, thank you to the audience for being so engaged is what kind of toll are you most, what either okay you can pick.

    00:47:20.000 --> 00:47:30.000

    What is the tool you're most excited to start using if you are that you like to use or what is the use case that you're most excited about that's coming.

    00:47:30.000 --> 00:47:36.000

    Like what is the thing you're most excited about, whether it's a tool or like a way to use it, that's on the pike.

    00:47:36.000 --> 00:47:38.000

    What do you think, Nick?

    00:47:38.000 --> 00:47:43.000

    I would say, Open AI, I think it's really cool and seeing how that may develop.

    00:47:43.000 --> 00:47:47.000

    And it looks like Pranava. I think it's really cool and seeing how that may develop.

    00:47:47.000 --> 00:47:50.000

    And it looks like Pranava, not as say that same thing, Yeah.

    00:47:50.000 --> 00:47:56.000

    Okay.

    00:47:56.000 --> 00:47:57.000

    That's

    00:47:57.000 --> 00:47:59.000

    Yeah, so sorry for everybody. It's like text to video generator that is like all over TikTok and is it's like Gemma I can't like Google was like we have a 1 million context window and then like open I was like, and we can make beautiful videos and 10 words.

    00:47:59.000 --> 00:48:06.000

    But is that your 2 Pr that you're excited?

    00:48:06.000 --> 00:48:12.000

    Yeah, that was mine as well. But you know, I'll think of another one once that goes.

    00:48:12.000 --> 00:48:13.000

    Yeah, yeah, yeah.

    00:48:13.000 --> 00:48:17.000

    Okay, we'll get, back to you. So Anna, what do you think? What do you either both excited to be able to do or a tool you want to use.

    00:48:17.000 --> 00:48:21.000

    You know, I have to be honest. I don't have one in particular that I'm like very specific.

    00:48:21.000 --> 00:48:35.000

    I one of the really cool things about AI I think is that it's constantly evolving. So I think in so many different aspects it is going to grow and and I'm so excited to see the stuff that I wouldn't expect.

    00:48:35.000 --> 00:48:38.000

    And specifically, I'm excited in the writing field because that is where because I love creative writing, especially like poetry.

    00:48:38.000 --> 00:48:48.000

    So and a lot of people think that like with creative writing works, it, needs to be solely like the core of it needs to be like, human originality, like the authenticity of it.

    00:48:48.000 --> 00:48:58.000

    But I also do think that AI, like with its potential, can help enhance that creativity. So I'm really excited to see what the future.

    00:48:58.000 --> 00:49:07.000

    I don't have one in particular that I'm thinking of, but. Yeah, I'm excited about that.

    00:49:07.000 --> 00:49:26.000

    I'm ready to see Anna's like collection of poems that comes out like in the next couple of years like I love I as someone that like struggles with writing I always love people that love writing because I love reading and so the fact that you see it as a creative like opportunity I think is amazing and those English teachers in the audience that are like, hey, is only for cheating.

    00:49:26.000 --> 00:49:33.000

    Ana is a perfect example of it. That is not the case. In fact, this is something that is supercharging her creativity. And I think that's amazing.

    00:49:33.000 --> 00:49:35.000

    Okay, last Penava, did you come up with something?

    00:49:35.000 --> 00:49:41.000

    Yeah. I definitely do wanna say I would wanna check out the, the, cloud, new updates to it.

    00:49:41.000 --> 00:49:46.000

    That's for sure. Because they're claiming some big things. I just want to test it out.

    00:49:46.000 --> 00:49:59.000

    So that's one tool that I definitely want to use and other than that though the Sora AI I mean Nick just you know he said it all it's incredible what they're doing now with the text video and stuff that I'm seeing on the actual website.

    00:49:59.000 --> 00:50:07.000

    That is like a game change. Like it's really big. So I definitely wanna, if users get the hands on it, cause I know that's another big, thing that people are having now like.

    00:50:07.000 --> 00:50:14.000

    Should be, just commercially that's the thing. So if it does happen, then that's definitely one of the things that I want to get my hands on.

    00:50:14.000 --> 00:50:19.000

    Well, there's also, you guys see there's a Google one that you can create to play your games.

    00:50:19.000 --> 00:50:20.000

    Hmm.

    00:50:20.000 --> 00:50:24.000

    So you can use Jenny, I'd actually create, like, multiplayer games now that are really cool.

    00:50:24.000 --> 00:50:30.000

    So I think we're gonna see a lot of really interesting use cases around that creativity. I just want to say thank you.

    00:50:30.000 --> 00:50:39.000

    We're going to let everybody go, whether you're, you know, going to dinner like Nick is or like probably getting a talk to your mom and dad and they'll ask how you guys did for Davana.

    00:50:39.000 --> 00:50:57.000

    I just wanna say thank you so much to our panel. Can we give them such a shout out? These are, first of all, let's remember we have a freshman in college and 2 high schoolers who are so if you don't believe that kids can tell you and teach you something in this moment, they are a perfect example.

    00:50:57.000 --> 00:51:00.000

    And I know that they are these are students that care about this and the intersection of A and education, but we have to listen to students.

    00:51:00.000 --> 00:51:07.000

    That's one thing that I want to say if I have one piece of advice. If you are not listening to students and you're not partnering with them in this moment, then you're not doing a the right way.

    00:51:07.000 --> 00:51:16.000

    And I feel confident saying that. And it doesn't mean that you're not trying. But you have got to include student voice.

    00:51:16.000 --> 00:51:21.000

    So I just want to say thank you to our panel. You keep in touch with us. We'll have this recorded.

    00:51:21.000 --> 00:51:28.000

    This will not be our last webinar with students. Parnava is going to be with us. I hope Nick and Ana will continue to work with us as well.

    00:51:28.000 --> 00:51:32.000

    But just appreciate everyone, whether it's the morning, the evening, or you know, like it's time for bed.

    00:51:32.000 --> 00:51:34.000

    We appreciate you being here and I hope everyone has a lovely time and just thank you again to our panel.

    00:51:34.000 --> 00:51:37.000

    We just absolutely appreciate you being here.

    00:51:37.000 --> 00:51:43.000

    Thanks