Nina Bamberg (00:00):
We can go ahead. We’re really excited to be doing this, whether you all found us through U4SC or Pedagogy.Cloud. We’re really excited to be doing a webinar today that brings our joint ventures together here. And so we’ll go ahead and introduce a little bit about what we do for those of you who found us through one of these organizations but not the other, but just in general today, I’ll go over just kind of the agenda we’re going to cover, but kind of some of the particular dangers that AI presents for things like civic engagement, civics education, things like that. Also a bit going into some issues that are present in how civics is being taught and learned today, as well as the particular importance that the age of AI presents for civics education, as well as how civics education can meet the need of AI as well. And we’ll end with a discussion on ways that civics teachers and people in the civics education field can use AI for different purposes. And then we’ll end with a reflection on what we see as where this is going and what we hope for the future of civics education with ai.
(01:30)
But Priten, I’ll let you take this part away if you want to give a quick intro to everything.
Priten Shah (01:39):
Hi everybody. My name is Priten. These are both ventures that I started way too long ago. One of them right after high school, one of them during college and the United for Social Change is our nonprofit arm. And so that’s, we provide interdisciplinary civics education. So we try to provide what a liberal arts education is supposed to do at the higher ed level, at the K to 12 lever to democratize that access. And then pedagog.ai is a project of Pedagogy.Cloud, which is a education technology company that I founded. And pedagog.ai is our teacher facing resources, professional development tools and events like this that help teachers navigate the era of AI.
Nina Bamberg (02:19):
Awesome. Yeah, I mean, we kind of said all this already, but just who we are. I serve as a director of growth at Pedagogy.Cloud, as well as the director of programs at United 4 Social Change. I’ve been with U4SC going on four years now I think, and just joined on the Pedagogy.Cloud side earlier this year. And Priton is the CEO of both and as well as the author of a new book on AI and the Future of Education. I don’t know if you wanted to say any more about that.
(03:04)
So yes, we will be posting a recording of this webinar along with these slides within the next couple days. Should be by the end of the week. Okay, awesome. So this first part is a little scary but also a little fun. So I tried to include some real life examples of the kind of dangers that AI presents that we need to consider in the world of civic engagement. And this first one is a deep fake video and it has a couple different examples of politicians who are not really saying the things that they’re saying in this video.
video (04:03):
Artificial intelligence. We talk about it a lot on this show, how it’s changing its potential tonight, how it’s being used in politics. Some videos have depicted made up futures, other deep fakes have made it sound like a candidate is saying something that the candidate never said, America, you’ll blame me for interfering with your democracy, but I don’t have to. You are doing it to yourselves. We’re entering an era in which our enemies can make it look like anyone is saying anything at any point in time, even if they would never say those things. Our great national anthem may be shark, baby shark, baby shark. These are not the best current examples.
Nina Bamberg (04:51):
So those are just kind of some silly examples they included in there, but I kind of wanted to include that because at the end he said they’re or he aknowledges that these aren’t the best current examples we have available. And that also to say that this technology is only going to get better and it’s going to become less and less obvious when this kind of technology is being used. This next one is real. This was really tweeted by the GOP Twitter account X account. And it’s not a deep fake video, but it is completely AI generated content. So I’ll play a bit of it just to give you an idea of the kinds of ways that AI is being used to affect political messages and campaigns. Now,
video (05:52):
This just in, we can now call the 2024 presidential race for Joe Biden this morning an embolden science in Banks. Taiwan financial markets are in free fall as 500 regional banks have shuttered their doors. Border agents were overrun by a surge of 80,000 illegals yesterday, evenings closed the city of San Francisco this morning fighting the escalating time and fentanyl crisis who’s in charge? It feels like the train is coming off the tracks.
Nina Bamberg (06:26):
So again, this not a deep fake video, but it’s a way that AI has been used to compile those images, create that voiceover and everything like that. So that’s an example that’s more real that we might actually be seeing coming up in next year’s election. This one’s an image. We saw a bunch of these going around earlier this year, again, pretty with some close examination, it’s easy to tell that this image isn’t true or isn’t real because the lighting is weird and things like that. But again, this technology is only going to get better and better and it might be harder and harder to tell what images are AI generated and which are not. And then this is another one of Barack Obama and Mark Zuckerberg dancing outside the White House. Again, not real, I don’t know, Priten you made this one, right? Or what platform did this come from?
Priten Shah (07:30):
Yeah, this is generated on Midjourney, and this was my third prompt on Midjourney. So I didn’t do anything crazy. This wasn’t anything. I wasn’t an expert using it when I made this, I was actually trying to generate one of the images of President Trump at that time because that’s when this became, that’s when I was starting to look into the DeepFakes with Midjourney and they had blocked the ability for, they obviously had rolling out changes to prevent folks from using them in malicious ways. And so they blocked my ability to recreate some of those President Trump being arrested images. But I was able to generate an image of Mark Zuckerberg and President Obama dancing. And so this was a nice example of that. And honestly, the shadows are pretty good. This is an example of where if I was quickly looking at something, I really wouldn’t be able to dissect that this isn’t real right away. And that’s when we get into some more dangerous territory. Some of the earlier examples are a bit more fun, but this is where we started to see the potential for misuse.
Nina Bamberg (08:34):
And this one was another one that we wanted to share. This is a recent article from just a couple days ago that it was found out that Sports Illustrated was publishing AI generated stories with AI generated authors and photos and everything. So there was no human writer behind some of these stories. And so we also wanted to bring this up as a way that this is going to become a pervasive problem within journalism as well. And the kinds of information that us and our students might have access to and what they’re finding, that they might be coming across something that looks like news on a site that they might be used to being reputable and fact checking all their information only to find out that it was written by not even a real person.
(09:30)
And the reason we’re bringing this up is because it has a real impact on our society and the society that our students are going to grow up in. And knowing that so many things that we’re seeing or interacting with can be AI generated, really can lower trust in our government, in media as well as in each other, the idea that the presence of AI could erode some public trust makes it much easier to spread misinformation and disinformation by people who want to do that or manipulate information, manipulate people into thinking a certain thing is real when it is not, which we see is a real harm to the democratic process. And so we wanted to start off our conversation about the need for civics education in the age of AI with those examples, because that is the world that today’s students are going to need to be prepared to be existing in. And so we want to make sure that civics education is meeting that need.
(10:44)
And unfortunately, the next part is also a little bit more on the depressing side of things, but based on research that we’ve done through United 4 Social Change as well as some other things, keeping up to date on news, we see some very clear and present issues that exist within civics education that if it is going to be ready for the age of ai, these issues really need to be addressed. And when we take a very broad view of civics education, so when we are talking about this, we are talking about pretty much any process where members of a society learn the beliefs, responsibilities, and norms associated with being a member of that society. Obviously we are thinking here today much more about what that looks like in the classroom, but we in general take a very broad view of what civics education is and where students and everyone obtain civic knowledge from.
(11:52)
So we mostly found that historically our civics education in the United States has broadly followed two different models. One being the bureaucratic model, which focuses on the structure and the function of government. So where students learn what are the branches of government who has power and things like that. And some of the issues that we find with that type of education often is that it is taught through rote memorization with little to no discussion of how students are meant to interact with these institutions that we are teaching them about. Then on the other side, another common type of civics education is character education, which focuses on the disposition and traits of how a society defines a good citizen. And a present issue in this type of education throughout American history is that it’s been used as a tool of assimilation for poor immigrant and non-white students. So it has often been misused as well. We’re absolutely not saying that either broadly either of these types of education should be scrapped. We obviously think that students need to learn about the structure and function of government and what it means to be a good citizen in their society. We’re just pointing out the flaws in how this education has often been approached throughout American history. Priten, was there anything you wanted to add there? Alright, cool.
(13:32)
Also today, so fast forwarding up to where we’re at today in terms of civic education, what we find is that civics education is often given a low level of importance in schools and in standards. And so a project that we’ve been on a years long journey of undertaking at U4SC is what we’re calling research for social change, where we dive into the state social studies standards across the country, and we often find very little civics that is required throughout the education process and particularly at the higher levels, at the high school levels where we think it should have even more of an emphasis because that is the age at which students are about to be adults and start voting and being full members of their communities. Also, there’s some pretty deep inequalities we find in who is receiving quality civics education across the country, and we see those divides along racial and socioeconomic lines as well as a real issue in rural areas. And so study found that a majority of rural youth this country live in what they called civic deserts, basically places where there are little to no opportunities for civic and political engagement outside of the classroom.
(15:04)
And so all of this obviously has an impact on democratic society with low knowledge and engagement affecting the whole process. And because when citizens are engaged, they engage in a number or participate in a number of different activities that can help their communities. And when they’re not engaged, it all suffers. We all suffer. I’m just going to check the chat. Oh, awesome. Okay. And unfortunately, as we’ve seen up until this point, that technology has been deepening a lot of these issues. So while we have an increased access to information, while we also are living in a environment where it can often be harder to know what’s true and what’s not true when we’re seeing so many different things being shared online, also the social media algorithms have created a situation where we’re all living in echo chambers and only being fed information that we already agree with.
(16:11)
And so social media platforms are not a place where any real discourse is happening. And so that brings us to the point of why we think digital literacy skills are such an important aspect of civics education and just kind of like our thought on what civics education could and should be. And we go back to Dewey here and take a lot of inspiration from his idea that schools themselves can be sites of democratic participation. And that the civics education really needs to encapsulate a number of different things, including everything we talked about, about how the government functions, but also things like problem solving skills, how to communicate, how to collaborate media literacy as well as valuing civic responsibility and social awareness. And this comes from a book chapter that Priten and I collaborated on earlier this year on AI and education, and I’m going to let him take over from here and talk a little bit about the importance of civic education in the age of AI.
Priten Shah (17:41):
So there’s multiple different reasons why we think that civic education has become even more important in the last year and will continue to become more important. As you can tell, we’re big proponents of why we think sophistication already has been very important and we’re behind as a society in terms of providing it. But we think that the impetus for providing it hopefully a little bit stronger now with the increase in ai. And so the three ways we’re thinking about this are that civics education can help us understand the social impacts of AI. So whether it be its impact of the job market, whether it be thinking about UBI, whether it be thinking about its impact in terms of widening or existing achievement gaps, all those kinds of things, civic education can help students and all of us process those impacts better. We can also better engage in discourse about AI related policies.
(18:29)
And so we will all be having really important conversations at the political level about what kinds of top-down approaches there will be to regulations. And all of that will involve folks being able to talk about these things, some level of literacy, but also some level of fluency with the actual civic process. And then we hope that folks can use the civic education that they get to advocate for better ethical and responsible AI use, whether that be political advocacy, whether that be advocacy directly in commercial enterprises or whether it be as an employee of an institution. But we think that in general, getting students prepared to do that, advocacy is an important aspect of civic education and the age of ai.
(19:12)
There’s a couple of different ways that we think that social education needs to adapt in order to meet the need in the age of AI, we don’t think that it can do so as it stands right now, one of the key things is to reframe how we think about digital literacy. And so this is an example of how the Council of Europe isolates out the two dimensions of digital literacy and both of these dimensions are going to need work on within our curriculum. And so the first is a technological dimension. And so our students will actually need to know what AI is. They’ll need to have some data science knowledge, they’ll have to have some knowledge about how computer science scientists create AI algorithms and what the basis of those algorithms are. And that will be the technological dimension of digital literacy. But they’ll also need to think more about the human aspect of this.
(20:00)
And so they’ll need to start thinking about what do other people need to know in order to prepare for a world of AI and the impact that it’ll have on our society? What are the ethical considerations that will come up that are both obvious? So self-driving cars is an example that always comes up in ethics classrooms, but also some more examples that won’t come out right away. And we’ll start to see the impact of which include the impact of biases within these systems and then long-term, what will the impact be on the job market and how we understand our economic systems. Both of those aspects will require deep understandings of both AI and the societies we live in. And so those are the two key aspects of digital literacy of the AI.
(20:44)
We also think that civic education will need to continue to focus on soft skills. And so this is another area where we think that civics education is really positioned to fill in an important need. And so we’ve always thought that these particular values and soft skills are an important aspect of becoming a well-rounded student at the end of either a K to 12 education or a higher ed education. And so we think that students will need to even more so develop their empathy skills, develop their decision making, critical thinking, problem solving skills, and these are the things that will eventually set apart as humans from AI algorithms hopefully for at least quite a bit of the near future. And so these are the kinds of skills that our students will need to be most adept at in order to actually adapt to whatever changes AI brings to the job market in order to position themselves well.
(21:34)
We’re an organization that focuses quite a bit on rhetoric and rhetorical skills, and we think that that will continue to be even more important in the age of AI. These skills include things like thinking about discourse, argumentation reasoning, logical fallacies, cognitive biases, all those aspects of thinking about what we’re saying to each other, how we’re saying it, how it’s being interpreted, and what we actually want to say and what we want someone else to think. And all of those things are even more important in a world where a chat bot can write like humans and can use rhetoric like humans because students will need to process this, the output that they’re seeing. And so they’ll need to be even better at finding logical fallacies at pointing out when something is solely using emotional appeals and not using logical appeals in writing they’re seeing out about, because all of this writing will be more and more tailored to individual preferences.
(22:21)
And so one of the other ways we think about civic manipulation that will occur in the age of AI is that there’ll be hyper customized propaganda. And so for example, we saw echo chambers in the age of Facebook and Twitter where folks were seeing tailored content based on their preferences. We’ll now see tailored content generated for a particular preference. And so these AI bots will know exactly what will convince you to vote in a certain way and try to give you content that will appeal to those exact same viewpoints. And we hope that we can prepare our students to be better consumers of that information, whether it be the fact checking or whether it be pointing out rhetorical mistakes.
(23:05)
We also think that at the end of the day, we don’t know what worlds our students will inherit. And so we all know that AI will change the job market. We don’t know exactly where things will settle. We’re already seeing quite a bit of that impact within the last six months. There’s new studies coming out to show that there’s a slower trickle of hiring for jobs that are easily replaced by ai. And so maybe we’re not seeing the headlines where everybody thinks that when they think in science fiction movies where you’re seeing mass layoffs, but you’re starting to see an impact on the job market in terms of folks hiring less individuals for filling roles that could be filled by ai. We’ll continue to see some impact of that. And this is only the beginning stages. It’s been just about a year since these technologies first became super popular, and we think that the amount of jobs that will be affected will only increase in the long run.
(23:49)
And while there’s jobs that be lost, we’ll also think that jobs will be created. But whatever that means, there is a lot of ambiguity about what the job market will look like when a lot of our K to 12 students will graduate and even when our higher ed students graduate. And so we think that this is a great opportunity for us to embrace what some of a lot of our school mottos say, a lot of our basic philosophical texts about why we have a public education system, why we’ve built liberal arts education systems in the US all that kind of already have the basis for what we think we should be moving further towards and returning to rather. And that includes helping students learn and be educated for the sake of their own personal growth and development and not just for preparing for a particular job in the future. And so we hope that that’s another aspect of civic education and education as a whole that we can start to embrace a little bit more.
(24:45)
And there’s a couple different ways that we think this can happen immediately in studies classrooms. And so we think that using real world scenarios is a great way for students to start to role play as humans. And so human decision makers rather. And so thinking about how can they maybe role play as Supreme Court, how can they role play as a boardroom making a decision and helping them really think through what skills are necessary in those contexts, whether they’ve be negotiation skills, compromise skills, their communication and rhetoric skills, all those real world scenarios will really help them fine tune those skill sets so they can actually use them in those scenarios. In the long run, we think that debates are a great way to get students to really exercise reasoning skills, the critical thinking skills, their rhetoric skills. And so more uses of debate within all of our classrooms I think will help students be better prepared to engage in the civic process outside of the classroom as well.
(25:32)
Similarly, mock town halls, legislative sessions, anything that helps them kind of mimic the kinds of things that will remain essentially within human control. And we think policymaking will largely continue to remain within the realm of human control. We need to make sure that students are getting some exposure to us early as possible. And then we hope that students can start to get some exposure to AI within these classrooms so that we can start having these important conversations. And so whether it be sitting down and testing out different outputs that are being generated by a I system and talking about the biases inherent in it, what data went into it, what might cause these biases, how might we tackle those? All the times of conversations we’re hoping can happen in our service classroom. So the students are prepared to have these conversations outside of the classroom.
(26:12)
And an important thing to remember is that all the students who were teaching right now it be at the K call level or higher ed, are going to be those who are going to shape the AI systems in the future. And so when we’re talking about ai, there’s a lot of concern. Everybody’s talking about, oh, these are all the dangerous things that can happen and here’s how things can get really out of control or be misused. But end of the day, those things are happening because humans will do those things there. Humans will be in charge of the process by which these things happen. And so if we can prepare our students to be better policymakers and be better developers, there is some of this within human control in terms of preventing the worst outcomes that folks are worried about. But in order to do that, our students need to be having those right conversations now so they can make those decisions later.
Nina Bamberg (26:58):
So this next part is going to be a little bit more concrete. So when you’re thinking through all the ways that current AI tools and technologies, how you can actually engage with them now, both within civics classrooms with students as well as on your own when you’re preparing curricula and materials and activities and all of those things. So this next part is the more I guess practical application of AI as it stands right now. And so the first thing it can help do is create standards aligned units and lessons. So AI tools know education standards and can respond to specific inquiries. It can create entire units and detailed lesson plans as well as activities and materials. So an example here, I asked chat GPT to create a unit plan on the judicial branch of the United States government for 10th grade students. And I said, it should be for lessons of one hour each and be aligned with C3 standards.
(28:06)
So I kind of included all these details in here that you might want in a lesson plan that you’re creating. And you can see right here it included the C3 standards for me, and obviously you could go look those up and make sure that they’re accurate and then write it laid out the lessons here. Obviously I asked it for four lessons, I only included two in the screenshot, but it gave me a quick overview of what each lesson might include with objectives, activities, and homework. Obviously I could refine my prompt even further with the activities that I want. I could go back to this chat and ask it to make a more detailed plan for lesson one or lesson two or any of the lessons in there. I could give it feedback and say, actually, I don’t think this objective is appropriate. Can you change it? And all of those things. But this is just an example of what this kind of prompt can get you in a tool like chat GPT.
(29:09)
Also, it can beyond lesson planning and things like that, AI can help you design engaging assignments and activities. So if you have a creative idea and aren’t sure how to bring it to life, it’s great at that. But it can also design comprehensive activities, some of those real world scenario activities that we were just talking about. And some of those activities that we recommend for civics education and the age of AI can be done with the help of chatbots and things like that. And so here’s an example where I asked it to design the simulation activities where students role play as a local election board and their goal is to increase voter turnout in their communities. And so you could see here it tells me what materials I would need to accomplish that activity, how to set it up, and then a detailed explanation of how students can go about that process. And so this can be a great helper in designing some of those types of activities that we absolutely think students should be engaging with in the age of AI, but are often a heavier lift on the teacher’s part in order to set something else up like this or design it. And so AI can be a really big helper in making some of that prep less cumbersome for teachers.
(30:35)
Also, it’s a really helpful at adapting materials to meet different students’ needs. So it can help you level different assignments or edit things like text to different appropriate reading levels. So here’s one example. This one actually isn’t the best. It gave me more of a summary than a simplification, but here I asked it to simplify an article about Abigail Adams to a fourth grade reading level. Again, this might be one you want to try in some different tools, and there are some tools out there including one that we have that specifically does this. But here I tried it in Claude just to see how it did. So that’s a different AI tool you might try. And I tried it in there because Claude can handle larger amounts of texts than some of the other AI tools, even though this one was pretty short. This one I thought was interesting too.
(31:27)
I asked it to design three versions of a quiz about the legislative branch of the United States at a beginner, intermediate, and advanced level. And so here are the first two levels that it gave me at the beginner and intermediate. And looking through them, I would agree with both of these being qualified as those. And then the advanced one actually came out with instead of multiple choice as short answer questions. And so it did a pretty good job of knowing what would be beginner, what would be intermediate, and what would be an advanced knowledge as well as task related to something like this.
(32:07)
It’s also great at personalizing learning. So the next few slides have some examples of how students themselves can use some of these tools to personalize the learning experience. And so it can help with things like brainstorming ideas for projects, for papers, for even exploring more deeply about topics that are of interest to them and engaging with classroom materials outside the classroom on their own terms and having it presented in a way that meets their learning needs. There’s a few examples here for this would be a brainstorming example, presuming we think persuasive speech is a very important skill within the civics realm. So I said I need a idea for persuasive speech topic related to technology, any suggestions? And this is a very simple way that you might show students how they could engage with this technology to help brainstorm. This one is a little bit more in depth, it’s more of an outline.
(33:13)
I asked it to outline an argumentative article about climate change, and I even included in there what some of the arguments I wanted to make were. So here it helped me outline those ideas that I already had. And this is great too. I put in a pretty basic example, but this is a great way if students have a lot of ideas or a lot of ways of thinking but aren’t as good at organizing their ideas, this can be a great helper for them as well. This one’s also a cool example of how you might use an AI tool. So here I said, give me a quiz to test my knowledge of the US Civil War. If I score high, give me more advanced questions, if I score low, give me easier questions to reinforce the basic concepts. And it did exactly that. So it asked me a question about the civil war. I got a couple right, and then I purposely got one wrong and then it went back and gave me an easier question. Although I do think that jumping from what year to what century was kind of a big jump in the difference in terms of what it thought the level would be there, but still it did do what I asked. And so this can be another way that AI tools can help students engage with material at exactly the level that they’re coming from.
(34:44)
It’s also really great for things like real-time feedback and support. And so we, it’s great because it can increase the amount of feedback that students receive beyond what any one teacher would be capable of doing. So throughout an assignment process or something like that, students could use an AI tool to get constructive feedback. And so it can also analyze writing for things like grammar or content or other things like that. And so a couple of the examples I have here, I said, can you give me feedback on my essay about affirmative action focus, your feedback on grammar, clarity of writing and overall structure. So I gave it some structure there in terms of the type of feedback that I wanted it to give me, and it broke that feedback down paragraph by paragraph. And then another one here is a great way that these tools can be used.
(35:41)
One of our favorites, and this is very applicable in the civics classroom where we want to be challenging thinking and ideas and things like that where I put in an article and I said, I’m writing this opinion article about housing affordability in New York City. What are some things that I might not have considered? And so this is a great use of the tools where it looked. And for this one I used perplexity, which is another AI tool that you could potentially explore. And the reason I chose that one for this task is because perplexity is also an active search engine that gives you sources with every question that you ask it. So you can see here where there’s a link at the end of some of these. And so it actually gave me a source where it could where to back up its suggestions that it was giving me what I might not have considered.
(36:37)
And so this is a great use of a tool like that to help students expand their ideas. If you’re hosting a class debate, you could have them say, right as you’re preparing your argument, go to an AI tool and ask it what the other side might think in order to make sure that they’re prepared for any potential counter arguments. Because we know that we can all often very much get stuck in what we believe or what our argument is. And the AI tools are a great way to help expand that idea and suggest what a counter argument might be or what we might have missed in our thinking. Oh, sorry, this is another example of exactly that same thing. And so I said here, I’m writing a persuasive essay on the drawbacks of social media. What are some arguments that someone who disagrees with me might take?
(37:33)
So similar idea to that last one in terms of the way that these AI tools can help expand your perspective, AR and VR are also some current but also tons of future potential here in terms of tools that you might use. An example I included here, it used to be called Google Expeditions. Now it’s called Google Arts and Culture, but it has these different things that you like historical stories and virtual field trips you can go on and things like that. And they have a whole section on American democracy on their website. They might want to explore another virtual field trip kind of tool that’s free is YouTube 360. There’s a bunch of VR videos on there that you might want to explore. But I chose this example because it had a bunch of different historical and current figures, stories, places that students could explore virtually. And also there’s learning tools out there. I took this screenshot from iCivics, which has simulation games where students can participate in different democratic processes. And so that’s another avenue where AI can make civics education more engaging. And for this section, I’m going to let Friedan take over again and we’re going to talk about the value of incorporating the act of critiquing AI into civics classrooms.
Priten Shah (39:19):
So I also realized that we’re reaching the end of the scheduled hour, so if anybody has any questions that they would like us to answer and you want to drop them in the chat right now or finished speaking, we’re happy to try to address those. And we’ll try to stick around a little bit if folks have questions as well, but I’ll make sure folks have a chance to ask those. But for now, I’m going to talk a little bit about how we think that AI can be incorporated within our civics classrooms by a lot of the students to actually critique the output it generates as a way to teach them both AI literacy and some of the core standards that we want to teach within our service classrooms. And so by critiquing AI, we can help students think critically think and participate actively by having direct input on the output that it’s generating.
(40:03)
We might also help them depart a little bit from just becoming consumers information to becoming evaluators and critical of the information they’re consuming and hopefully that makes them more active in their actual learning journey. But some examples might help illustrate this a little bit more. So the first example we have is having AI generate reports. And so this can really be used for any particular unit or subject that’s being taught. So whether it be a report on climate change, whether it be a report on historical character, but having to generate some sort of report on a topic and then having students verify refining. And so they might think about which things are being left out by the report, which things might need a source to back it up, which things they think it might not actually be accurate. It could be hallucination by the AI model.
(40:46)
So this is an example of having Chat GPT generate a model, a report on the history of confirmation hearings of the Supreme Court justices over the last 50 years. And so a very easy way to integrate AI into an existing topic you be teaching in your classroom anyways. Now we can go through and think through, okay, what might be necessary for making this a more thorough history? Are any of these things incorrect? Are these things I want to back up and look for more information about? And that really helps them get some content area knowledge and test their content area knowledge, but also develop some AI literacy skills.
Nina Bamberg (41:20):
I also want to note for this specifically used GPT 3.5 for this one on the chance that there’s less of a chance, it would be perfectly accurate. So also to note, the different AI tools you choose might also illustrate this example better, but Chad GPT is a great example of if you’re paying for it, you’re probably getting more accurate information, but unfortunately the one that the students are just coming across on their own is not at up to that same standard.
Priten Shah (41:55):
The other thing to point out with this is that Google’s bard automatically does that for you. And so there’s a little Google icon in the shape of the Google logo, and if you click that one after it’s generated any text, it will actually do the fact checking for you and give you a source of backup things and tell you things that might not actually be a hundred percent accurate. So definitely we’re playing around with in terms of seeing the potential for that, but in some ways it serves as an answer key for this activity. The other example I already talked about earlier was having it generate something, whether it be stories or articles or images, and then having students discuss the biases that they see in it, see how AI make existing societal biases worse, and also how data training or user usage could actually combat the issue.
(42:41)
And so this can vary from thing. This is an example of having to generate a paragraph about inflation as if it was a newscaster talking on tv. And this is a great example of thinking through the rhetoric choices that chat GPT has made as it mimics a newscaster. Thinking through what political angle has it taken, maybe even having to take a different political angle. And so saying write inflation as a field newscaster at Fox and then at CNN and having students compare and contrast with rhetorical choices Chat GPT makes between those two and what biases are coming out in both of those examples are all great ways to foster some critical thinking about the students, probably a fun activity for them as well while still engaging in content area knowledge. You can also have students think through compare and contrast human output and AI output. And so again, you have AI generates some sort of story article or image and then have the student evaluate that, comparing it to their own writing or some other human generated writing. So an example might be reading an article from the New York Times on a particular topic and then having Chat GPT generate the same article and having the sorting through what might be differences is exactly what this example
Nina Bamberg (43:55):
Was in the New York Times today. And I use that as an example to see if I could get chat PT to say something similar.
Priten Shah (44:07):
But this is a great opportunity to really think through what is the value of a human writer or a human journalist versus us having the AI generate that article.
(44:20)
I think all these skills are pretty, I’ve talked about them as I went along, but the skills that they’ll develop by doing this will include the research and information verification skills that we think they’ll need. So they’ll need to be really good fact checkers as AI becomes more and more ubiquitous, but also need to develop really nuanced understandings and ability to comprehensively analyze text that they’re reading. And so this allows students to really think through, okay, what are the gaps in my knowledge and what do I need to know in a better critique AI and not just produce some output of my own? And these are we think will function as vital life skills, whether it be in career spaces or in the civic space as they become voting members of their societies.
(45:02)
These are some examples with Socrat.ai. This is our chatbot that allows teachers to assign particular example assignments to students that are tailored to a particular need. And it’s another great way to have students engage with AI within the civics classroom but in a controlled environment. And so some different ways include using it as having a Socratic dialogue. And so if you’re teaching about climate change, students can have an individual Socratic dialogue about climate change with Socra. Socrat will continue to ask questions to provoke their thinking and a great way to have individual interactions with each student getting individual interactions and be able to take the intellectual journey down the path that they’re going down based on how they answer the questions. That’s one of multiple different ones. Another example is using it to have students role play with a particular character. So this example of Socrat role playing as James Madison and then students having a conversation with James Madison about the Bill of Rights. And so another way to make some of the history elements of civics more engaging and get students the opportunity to go through and talk to historical characters and think through what might they have thought about this, what questions might I ask better understand this topic area.
(46:11)
The next example is a debate. And so we talked about having debates in classrooms. Students also be having individual debates with AI bots as this example again with Socrat where they’re debating about whether or not to end the SAT and Socrat models really respectful debate. And so if you notice some of the rhetorical choices Socrat is making, it includes, I understand your point, however I must say I see your point, but I respectfully disagree, really modeling healthy disagreement and debate while still staunchly pushing the student to think critically about the stance of this student is taking. And so will take the exact opposite stance of whatever the student decides to argue about. And so make sure that the student’s own viewpoint is being tested as much as possible.
Nina Bamberg (46:52):
Those are all the examples I put in there of Socrat. But if we wanted to talk, if you wanted to bring up the other assignment types as well. I think probably I just missed study buddy in language, which are probably less relevant here, but Study Buddy obviously would be helpful.
Priten Shah (47:08):
Yeah, yeah. Study Buddy is a socrat feature that allows students to have a generalized conversation about anything that they want to learn about. And so for example, if they are trying to test out, they want to know the three practice of government and they can just go back and forth, Socrat will custom tailor that to test their knowledge out a little bit and give them some more examples to test that out. So definitely another way use that in the civics classroom. But in general, there’s lots more assignment types that will be coming out by then of the school year. And so just following along with our Instagram or with Socrat.ai is a great way to get posted on any more developments that come out that are civics related.
(47:44)
A quick note about the future, one of the things that we’re really excited about working on are civic learning pathways. And so this is where we’re imagining is making civic education really meaningful to individual students so that they feel engaged and they feel that what they’re learning is exactly killer to what they kind of want to achieve in the world. And so the way we imagine this is that, for example, if you’re take a unit on the US government, a student who is passionate about climate change might learn it a little bit differently than a student who’s very subject is history. And so for the student who is passionate about climate change, they might first learn about the judicial branch through a Supreme Court case and about the EPA. They might then learn about state governments by following this climate bill through the state legislator.
(48:26)
And then they might finally learn about how elections work by following the advocacy work of a climate organization during a presidential campaign. So there’s still learning key concept that you would learn in a civic classroom, but really tailored towards the student’s particular interest in climate change advocacy. Whereas the student interest in history might learn about the structure of the US government by participating in a mock constitutional convention so they can go through the historical basis of the constitution. They also learn about the electoral college by reading experts from their federalist papers. And so really again, tying in modern day things that the civic classroom to teach from the angle that the is interested in and then finally might learn about the role of citizens in a democracy by following the fights for voting rights throughout US history. And so again, giving them a chance to think through the role of citizens and voting rights, but from the perspective of history because that’s what the student is interested in. So we hope that these kind of solutions continue to pop up. We hope to be part of some of these solutions, are excited to see other players in the field also come up with interesting ways to engage students in the civic process and become more engaged in general with the civic process outside of the schools as well.
Nina Bamberg (49:32):
Questions or thoughts reflections?
Speaker 5 (49:42):
Yeah. Well, a quick reflection. I do have a question, but I’ll let others jump into that last slide about I loved that the climate change path ended up talking about elections and in reality as a citizen, they are interrelated, but it might not be obvious, it might not even be in a textbook. And I just think that that’s really interesting that it can go. So I mean obviously AI can, but that the student experience can go so broad like that.
Priten Shah (50:18):
Yeah, I think that that’s where we see the easiest way to tie in different student. I think the key for assess providing any of that generalized knowledge and then finding what is tied into the student’s interest. And so we know we need to teach about elections, but how do we get them to engage in that? So this student who’s interested in current modern day climate change policy isn’t really, maybe not super interested in reading with the Federalist papers and the electoral college through that angle, but maybe getting ’em to think about elections from the climate advocacy policy stance. We’ll actually get them to think about the process a bit more because they might think that they might be convinced that working outside the system is the only way, and so they don’t really care about the election system. And so this is one way to bring them back into at least learning about it for the sake of learning so that they know what’s to be more informed. But this is where we see the real potential of long-term personalized learning for students powered by ai that actually allows us to still meet the curricular goals that we want to meet, but really meeting our students where they are and where their intellectual journeys are.
Speaker 5 (51:25):
And the question I had is, are you all seeing any promising trends related to governments mandating civics?
Nina Bamberg (51:35):
A few. Some states it really ranges by state about whether there’s anything required and if we think that what is required is sufficient in terms of an accurate measure of civic knowledge and engagement. And so yeah, it’s a range in terms, but it’s it’s very state by state and it’s very different depending on where you are.
Speaker 5 (52:13):
That makes sense. Yeah. Thanks.
Nina Bamberg (52:19):
Yeah, Sue?
Speaker 6 (52:21):
Yeah, sorry to bother you. This has been great. I can’t get access to the AI notes and I’m wondering if when you send the link to the slides, you could send the link to the AI notes. Am I maybe not being
Nina Bamberg (52:53):
Clear? I’ll get an email later whenever we hang up this call. I’ll get an email with an AI summary of what happened. And yeah,
Priten Shah (53:08):
I think the notes that Sue was asking about are the ones that were generated by Otter, one of the other participants had included Otter, but I believe that now that Allison has left, it has disabled access. So I did just send an email requesting access and if I get that, I’ll send it to, but otherwise Zoom has its own built-in version of this that we will get access to. We’ll make sure to email that out. But yeah, it’ll be really interesting to see what the Zoom version of this says too. Zoom tends to summarize a bit more and not just be a transcript. And then we can also include the Zoom transcript as well. So if that’s of interest, we can definitely send that out.
Speaker 6 (53:43):
That would be great. I have two more quick questions. Can you recommend a source that can help me refine my prompts? Maybe is there an article or something I could read about how I could make my, what’s in my head actually come back from chat GPT or the other tools, something deeper than be specific?
Priten Shah (54:14):
Yeah, so we provide a lot of these resources, so I’ll do a quick pitch for those, then I’ll give you some more general ones. I think Chris is sharing some of their chat as well. So those are a great place to start chat. And then OpenAI and Claude both also have their own, so if you just search, I can try to find a link in a second. But if you just OpenAI Prompt engineering, they have their own thoughts about how to do this. We have education specific education specific craft course up that talks a little bit about prompt engineering. And then we have a book out called AI and the Future of Education that also has a section on thinking about prompts with lots of example prompts as well. And then a free account on the pedagogue website also gives you access to our 600 and something prompt examples. And that might be a good way to brainstorm some of the weight, different ways you can write those prompts. The last one is completely free, so that’s definitely worth checking out at some point.
Speaker 6 (55:04):
Great, thank you. And final question, in one of the slides there was something about using, or there was something in the remarks about using Google to confirm an AI generated report, and I wanted to know if that means just googling the topic or if there was some specific Google tool you had in mind to be used to confirm that other AI report.
Nina Bamberg (55:46):
So I can show you right
Priten Shah (55:47):
Now. Yeah, perfect.
Nina Bamberg (55:50):
Google’s chat bot is called Bard and now what was that? I can even use that same report prompt I used in the slides to show how it, oh no, that’s all the way back to the beginning to show how it works. Sorry, this is being very slow. What was it a report on?
Priten Shah (56:12):
I was going to say maybe while you’re generating that, the only thing to note is that this actually only works for Output produced by Bard, and so it wouldn’t really work if you took Chat CPTs output and then asked bard to do it. I’ve tried it tries. It does an okay job, but not a great job. The best thing it does do is anything that Bard says to you, it does and it says exactly what ENA is showing that it is a really good job of verifying and it’s a button right here, so
Nina Bamberg (56:45):
I’m trying to use the same one as I use in there hearings of Supreme Court Justices over the last 50 years. And so this is all you have to do and anyone with a Google account can easily create a bared account and I’ll show you here at the bottom after it creates its output and there’s different things you can click on at the bottom. One of them is this Google icon that says double check response. And when you click on that, it goes basically line by line giving you a source where you can back up that information.
Speaker 6 (57:46):
Awesome. Thank you. Thank you very much. This was really helpful.
Nina Bamberg (57:52):
I’m glad. Yeah, so that’s a cool Each of these, yeah, exactly, Chris, right, each of these different tools, we only had a chance to mention a few like chat, GPT, Claude, this one bard, perplexity. And so each one kind of has one kind of unique thing it does or a couple of unique things it does that the others don’t. And so it can be fun to play around with a few and then kind of get to know what each one does and which one you most like interacting with. Overall. We definitely still find Chat GPT to be the most comprehensive, but again, that’s when using the paid for version of Chat GPT for, whereas all the other ones you’ve seen have been purely all the free versions that are openly available.
Speaker 6 (58:50):
Thank you. And I do have the paid version of Chat GPT, and I’ve enjoyed playing with it.
Nina Bamberg (59:01):
Awesome. Yeah, chat, GPT is still definitely our go-to for most things, but some of the others do have some of those fun features.