Language Practice Partner Prompt

ELL Instructions Translator Prompt

Personalize Quizzes Prompt

Difficulty Level Differentiation Prompt

Cultural Contextualization Prompt

Virtual Science Tutor Prompt

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.

Bryan D. Eldridge, M.Ed.

With an accomplished career spanning over a quarter-century as an educator, AI practitioner, and leading emerging technology educator, Bryan brings a uniquely strong passion to the area of applying cutting edge technology to the art of delivering quality instruction. Bryan holds an M.Ed. in Computer Based Education from the University of Georgia and a B.S. in Mathematics from the University of Kentucky. He has dozens of publications spanning both the educational and emerging technology realms, and has designed over fifty degree, certificate, and bootcamp programs in emerging technologies.

So just lovely to have you all here. I'm Amanda. I'm the CEO and co-founder of AI for Education.

Very excited to do a really practical session today on differentiation. You in all the work that we have done all the work that we have seen differentiation and personalization of learning and has been one of the biggest most interesting and impactful use cases of these tools.

We know that they are new and that they are, growing they don't do everything we want. We were just complaining about their inability to format things correctly.

But it is something in which we can see this ability to format things correctly about their inability to format things correctly.

But it is something in which we can see this ability to really, but it is something in which we can see this ability to really have like have supercharged our lesson planning and to better meet the ability to really see this ability to really have like have supercharged our lesson planning to better meet the needs of our students.

And, and to better meet the needs of our students. And so, what I always want to see and to better meet the needs of our students.

And so, what always wanted to see and thank you for those that are already getting started. This is your opportunity to get involved, not only with Brian and I, but with each other.

So please make sure to say hello where you're from. And also just a chat. So if you have resources, best practices, questions, their community and our chat is one of our most favorite things of this year so please use that time wisely and and these people that you have around you.

And then lastly like we're actually going to be doing a significant amount of prompting today. So we have 7 new prompts that have been designed in collaboration with Brian around everything from a student facing prompts around tutoring to a culturally responsive prompt, etc.

00:02:43.000 --> 00:02:49.000

So we love you to get involved. We'll be dropping in those prompts, but definitely have your chat QT, your favorite AI program open because we'll be doing that pretty much for the entire session.

0:02:49.000 --> 00:03:01.000

Lastly, oh my gosh, it's almost the holidays. So we only have one left here does not mean that the I launch pad series is going away.

00:03:01.000 --> 00:03:05.000

I'm a crazy person. This is number 26. We do not plan on ending these webinars because we have such a lovely group and I get to keep having great guests like Brian.

00:03:05.000 --> 00:03:25.000

But our next one is with Tom Vander Ark who is a good friend and really amazing thought leader and we're gonna be going deep and get ready for like a pretty interesting naughty conversation about durable skills of the future.

00:03:25.000 --> 00:03:31.000

What is a learner experience going forward? I promise it'll keep you on your toes. But today we're really going to be focusing on differentiation.

00:03:31.000 --> 00:03:38.000

And so we're going to do a little bit of level set in terms of like what differentiation is and how it's changed.

00:03:38.000 --> 00:03:42.000

But we want to spend the most time actually showing you how to use generative AI specifically. Chat to do this work.

00:03:42.000 --> 00:03:54.000

So I'm going to hand it over to to Brian who is going to hand it over to to Brian who is going to give us our very common first answer.

00:03:54.000 --> 00:04:01.000

Okay, great. Wow, thanks so much, Amanda. My name is Brian. I'm a geek.

00:04:01.000 --> 00:04:16.000

I'm a self proclaimed geek for the last 25 years. For the last 25 years I've had one foot solely and startups and and startup centers and working with small companies, early-stage companies and emerging technology.

00:04:16.000 --> 00:04:18.000

In the other foot I've had in education. So I've been really interested in the skills gap.

00:04:18.000 --> 00:04:27.000

Been really interested in how we can take, you know, merging in frontier technology and help all kinds of practitioners do their jobs better.

00:04:27.000 --> 00:04:37.000

And, you know, I have a real soft spot in my heart for teachers. I think there's a lot of power and throughput to come from a frontier technology like Generative AI.

00:04:37.000 --> 00:04:45.000

Very excited to be here and be part of this. The very first time that I used generative AI, wish for a demo because someone had text to me at work and say, Hey, do you know much about this?

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

You're keeping up with it. I'm like, oh yeah, I've heard about it.

00:04:51.000 --> 00:04:57.000

This is probably like last January. And, I have a theme song for each of the people I work with.

00:04:57.000 --> 00:05:02.000

So when they come on, I said, yeah, I just said, Dan, Dan, you're just the man.

00:05:02.000 --> 00:05:03.000

Okay.

00:05:03.000 --> 00:05:23.000

So yeah, so I had chat GPT come up with new theme songs for 3 people. On my on my team and it was a lot of fun and it got people interested and some people have new nicknames because of it but but it was it was really cool we started using it in anger very shortly after that to help out with creating deliverables and outlines for new courses.

00:05:23.000 --> 00:05:32.000

And we've really tried to do a lot of interesting things. With chat GBT and prompts and hopefully today we get an opportunity to share some of that with you.

00:05:32.000 --> 00:05:37.000

Hello, I love that and I love the creativity. I mean, you would, I mean, if you're a elementary school teacher out there and watching us today, go.

00:05:37.000 --> 00:05:51.000

We're getting a theme song for your students. We know we love the viral, you know, like fist bumps or you know the different salutes but getting AI to kind of have that fun with you is such a great idea and I love that that was your first use case.

00:05:51.000 --> 00:06:00.000

I always find this is such a great example of like what people like first getting you know excited about and so they kind of leaves us today because you're talking about differentiation.

00:06:00.000 --> 00:06:07.000

Actually, you're differentiating the way that you work with the different people in your team and the way you approach them, the way you say hello to them, which is actually what we're going to talk about today.

00:06:07.000 --> 00:06:14.000

And so we think about differentiation. We're gonna do, we're not gonna spend a lot of time, but it's been around.

00:06:14.000 --> 00:06:17.000

I mean, I was a teacher. Almost 20 years ago. Oh man, I back in the day and you know we were talking about differentiation.

00:06:17.000 --> 00:06:44.000

At that time this ability to meet the needs of students at different levels within the same classroom and it was really fascinating because this is you know very early on in the technology map of our schools we did not have one to one devices we did not have great connectivity and we were expected to like create these spaces in which we were able to love both instruction not only to students needs in terms of their skill, not only to students needs in terms of their skill levels but also their interests.

00:06:44.000 --> 00:06:49.000

And one of the things that I got most excited about when I thought about the students needs in terms of their skill levels, but also their interests.

00:06:49.000 --> 00:06:50.000

And one of the things that I got most excited about when I thought about this technology was that I got most excited about when I thought about this technology was that immediately it became clear to me how you could start.

00:06:50.000 --> 00:06:55.000

And one of the things that I got most excited about when I thought about this technology was that immediately it became clear to me how you could start to use this tool to start actually meeting the needs of your students.

00:06:55.000 --> 00:07:03.000

And so I'd love to, you like, so Brian, I'd love to understand what you think has changed and their ability to differentiate based on Chats VT and other Generivi tools.

00:07:03.000 --> 00:07:11.000

Yeah, I think that it's really perfect timing because if you look at the climate and some of the challenges that we have in the classroom today is we have more diverse classrooms than we've ever had.

00:07:11.000 --> 00:07:16.000

We have more more children, more students on the spectrum. We have more cultural diversity than we've had.

00:07:16.000 --> 00:07:27.000

And this isn't something that's just in big city school districts now. It's all over the country.

00:07:27.000 --> 00:07:42.000

And I think teachers also have more responsibility in the classroom than they ever have. So they need partners. You know, they need technology that will work with them can take the the innate abilities that they have and help improve those and help introduce efficiencies.

00:07:42.000 --> 00:07:46.000

And I think the generative AI is unique and then it really doesn't take a lot of setup.

00:07:46.000 --> 00:08:00.000

It doesn't take a deep understanding of all the underpinning mathematics and AI. In fact, I think Amanda you and I were talking yesterday or sometime, sometime this week about the people that tend to do really well with this are very creative.

00:08:00.000 --> 00:08:11.000

They're people that are already master. Communicators and problem solvers. They're curious, they're judgmental, they're opinionated, they have all these great things, they're good storytellers.

00:08:11.000 --> 00:08:19.000

And I think bringing those to bear to task to you generative AI is really going to have a massive throughput in education.

00:08:19.000 --> 00:08:29.000

And so I think that maybe early on people who weren't technical were a bit intimidated and said, well, I don't want to get into this because I don't understand how it works.

00:08:29.000 --> 00:08:37.000

Trust us when you don't have to understand how it works. What you have to understand is how to have these conversations, how to really embrace the empirical method.

00:08:37.000 --> 00:08:44.000

You're gonna try something. You're not gonna get the response that you want. You're gonna try something to get and then you were.

00:08:44.000 --> 00:08:50.000

And what another thing that I'm and I were talking about before the session is I started on some.

00:08:50.000 --> 00:08:55.000

Some props and got him to her and it's like what we came out with in the end in very short order with very little collaboration is something that would have been very tough to do without this type of technology.

00:08:55.000 --> 00:09:11.000

In the short time, rising that we had to actually pull that together. So yeah, so I think the generative AI to answer the harder your question, it could be a real game changer.

00:09:11.000 --> 00:09:22.000

And differentiation. And should be a tool that teachers look to when they're dealing with the base principles of different differentiated instruction.

00:09:22.000 --> 00:09:29.000

I love that and we talk about teachers being the best prompt engineers because what do we do every day all day as we ask good questions?

00:09:29.000 --> 00:09:30.000

Yes.

00:09:30.000 --> 00:09:37.000

And so all what we are doing with generative AI is we're acting good questions in the same way that you learn how to frame a question for a student.

00:09:37.000 --> 00:09:42.000

In the same way we're gonna ask you to think about framing your question to chat, GPT, which we call prompting.

00:09:42.000 --> 00:09:53.000

And always blows everybody's mind that to understand that the most common like like language right now for coding is essentially not common, but most newest is in English.

00:09:53.000 --> 00:09:58.000

Literally there is a system prop that lives behind ChatT that says, Be polite. You know, like, don't say I don't know.

00:09:58.000 --> 00:10:07.000

But keep answering if you're not sure the first time try again and it's not lines of code that's not lines of text.

00:10:07.000 --> 00:10:18.000

I mean, of numbers. What it is, it's actually literally language. And you can see from the system prompt that they've actually iterated on that system prompt because you kind of tell where they've improved it over time.

00:10:18.000 --> 00:10:38.000

And so everyone in here today that's going to be working with us or watching it later, when we ask you to prompt with us, we're asking you to be computer scientists, but it doesn't mean that you have to understand everything about this technology, but it does mean that if you learn how to ask a good prompt, how to frame it, how to essentially give it a direction, you are going to actually see a better

00:10:38.000 --> 00:10:40.000

output. And so I think this is where it gets really exciting where the work that Brian and I have done.

00:10:40.000 --> 00:10:49.000

These prompts will get you much closer to where you want to be. And so it's a starting place, but you're going to be the creative engine.

00:10:49.000 --> 00:10:56.000

You're gonna know your students best. You're going to be able to identify the areas in which you can personalize this instruction.

00:10:56.000 --> 00:11:08.000

Much better than we can. So we're just giving you the step up for that as Brian said that curiosity, that creativity and empirical like approach that a comfort with failure sometimes failure is really funny.

00:11:08.000 --> 00:11:12.000

I mean, I'm honestly like sometimes when chatting team messes up, it's one of it's actually the most fun I've had in a while because it's so silly.

00:11:12.000 --> 00:11:27.000

It's actually the most fun I've had in a while because it's so silly. But so I think that like when we think about this, like when you think about differentiation, what do you think the technology can do really well right now and what do you think it doesn't do as well?

00:11:27.000 --> 00:11:33.000

I think right now it's really good at getting to the specifics. Of a given problem that you're trying to solve.

00:11:33.000 --> 00:11:45.000

So I need to get something translated, you know, I need an idea for advanced topics or a medial topics on this certain set of objectives that I'm dealing with.

00:11:45.000 --> 00:11:49.000

I think the other thing that it does really well is scalability. Okay, cause when you look at when you look at what artificial intelligence really does, it recognizes patterns.

00:11:49.000 --> 00:11:54.000

It recognizes relationships and structures at a scale would be really tough for a human to look at 5 million slides.

00:11:54.000 --> 00:12:03.000

And say, okay, that's cancerous. So these, these cells probably aren't cancers.

00:12:03.000 --> 00:12:12.000

So, so I think getting to points where you have problem sets, we have specific types of problems and you're able to say, okay, this works for one audience that I'm differentiating for.

00:12:12.000 --> 00:12:17.000

Now I can make a couple of small changes in the follow-up prompt and do it for all my audiences.

00:12:17.000 --> 00:12:39.000

So I think the scalability aspect is something that is really powerful. But just its ability to articulate back in language and give you the format you're looking for, whether it's a haiku or whether you're looking for a song lyric or an outline and now I have it building all kinds of complex tables with different formatting for different types of internal deliverables that we're playing together. So it's really good at that.

00:12:39.000 --> 00:12:50.000

So it's good to the details and truly good scalability. What it's not really good at is, you know, this is all still in its infancy is so powerful that I think sometimes we take it for granted we're dealing with an infant.

00:12:50.000 --> 00:13:03.000

Essentially, in terms of a mature technology and when you're working with things like Dolly, which I have the the paid version for GPT for it can be very frustrating because it's one thing to give something a sentence and get a sentence back.

00:13:03.000 --> 00:13:15.000

It could be one of countless number of responses that would make sense. But when you give it a prompt for an image, you're thinking about something very specific.

00:13:15.000 --> 00:13:18.000

And you start seeing when your language maybe lacks a little bit or you don't know quite how to ask for something visually.

00:13:18.000 --> 00:13:31.000

The same way you would when you expect. You know, from a language based response. So I think the, the the different modalities that you'll be able to output to because you know there's already some tools that do things like videos.

00:13:31.000 --> 00:13:40.000

I use it all the time when I'm, before I shoot a video, it creates a running shot list.

00:13:40.000 --> 00:13:54.000

So say, okay, I'm gonna teach new Python programmers how to build this type of module with Django or Flask and it will do the video and it'll have the running list and the images on the right hand side with very little editing I can output that, hold in my prompt software and have an instructor shoot a video.

00:13:54.000 --> 00:14:11.000

So that part is awesome now, but I think when it gets to the point where it can create 3D objects quite easily, it can stub out videos, it can have different audio tracks that you can swap in and out to have different languages with all your videos.

00:14:11.000 --> 00:14:21.000

I think I think all this stuff's coming next and that's going to be exciting. So it's, it's, it, there's also that bridge in between, okay, you get what you want in your prompt and you're really excited.

00:14:21.000 --> 00:14:27.000

Then as we were also talking before we came online, was then you've got a deal of a format.

00:14:27.000 --> 00:14:28.000

Yes.

00:14:28.000 --> 00:14:39.000

So I think I think having some better tools. Some better, more intuitive output capabilities with, with that having to be heavily prescriptive will be something I think will add a lot of utility.

00:14:39.000 --> 00:14:40.000

To generate.

00:14:40.000 --> 00:14:49.000

Sorry. Yeah, and I think that the chat interface can be a problem. Like, it's not necessarily always the easiest, as you said, to get the output, it's not necessarily always the easiest, as you said, to get the output that you really want to see.

00:14:49.000 --> 00:15:07.000

But I think that they're really just great opportunities to, the thing I get most excited about is I think that one of the used cases it does really, really well, which you're going to show, is this idea of leveling text or leveling questions or leveling explanations for like, high skill, medium skill, emerging skill, right?

00:15:07.000 --> 00:15:24.000

And so I think that that's something that's really, really great and we know that makes a huge difference because if you give students an explanation, I just, and instruction within their own approximal development, you're going to see them succeed much, much faster and they're gonna find challenge in a way that's meaningful for them and not discouraging.

00:15:24.000 --> 00:15:34.000

So that's something it does really, really well. And so we're going to see that today, but like as we go forward, we're going to probably see more and more tools that are able to do that at a click of a button.

00:15:34.000 --> 00:15:45.000

And then the thing that makes it magic is not that it's just, you know, a different description, but it's actually the first prompt we're going to show you in a moment is we can make it in the students interest.

00:15:45.000 --> 00:15:46.000

Yeah.

00:15:46.000 --> 00:15:50.000

We can make it in their their zone of engagement. I don't even know if that's a real thing, but if it isn't, I just made it up.

00:15:50.000 --> 00:15:51.000

Okay.

00:15:51.000 --> 00:15:52.000

So, the unparalleled development and zone of engagement. Like what is a kid most interested in doing or learning about?

00:15:52.000 --> 00:16:13.000

Like that can be really great or what's the most culturally responsive way to do this? Because you know, you just said like we have a very complex, very diverse, beautiful makeup of our country and countries that are here today and to be able to meet those students where they are both in the terms of their skill but engagement I think is really exciting.

00:16:13.000 --> 00:16:21.000

But on the other end yesterday I was actually doing a webinar on a webinar but I've training with a tribal college in Bismarck, South Dakota.

00:16:21.000 --> 00:16:27.000

And so, these are instructors that are working with the Lakota Nation. And some of them are Lakota, some of them are not.

00:16:27.000 --> 00:16:38.000

And what was really interesting is that when they were really trying like some of the teachers and structures were thinking about how do I create more culturally specific content for their their students.

00:16:38.000 --> 00:16:45.000

And it was really interesting because while it could write a response in Lakota when it was asked to create a specific activity with a Lakota connection, it used incredible cultural biases.

00:16:45.000 --> 00:17:00.000

In fact, the first one was about totems which don't exist in Lakota history and the second one was around sacred rights which we definitely don't want to talk about in a Python coding question, which is not a appropriate.

00:17:00.000 --> 00:17:19.000

But it was like the structure was really trying hard to be more culturally appropriate, but he was not Lakota, but someone else was and they were able to very, very supportively say, this actually doesn't, this is very culturally inappropriate or a lack of specificity.

00:17:19.000 --> 00:17:24.000

It's using these general stereotypes which we know are a major issue with these generative AI tools.

00:17:24.000 --> 00:17:30.000

And so it does require, I think, as we think about real deep differentiation that is culturally appropriate.

00:17:30.000 --> 00:17:39.000

We do have to be very, very cognizant and very careful. That we are not exacerbating existing biases and or creating new ones.

00:17:39.000 --> 00:17:46.000

So our new stereotypes. So I think that, you know, as we talk about in every webinar and everything I do, it's always about the balance.

00:17:46.000 --> 00:17:53.000

And as you said, this technology is a baby. It is learning to crawl. It's like kind of like a spaceship baby.

00:17:53.000 --> 00:18:09.000

Like it's like a space alien baby so we don't really exactly know what crawling means right now because sometimes it's crowing and sometimes it's running and sometimes it's skipping like and it kind of does it in a really weird pattern but it is still very new and we need to be critical users.

00:18:09.000 --> 00:18:19.000

So everyone, if you are not on chat to VT already, take out your phone, get on the web, whatever you need to do, pull up chat, we're going to use the free version like we always do.

00:18:19.000 --> 00:18:32.000

And so Brian is going to do our first show and tell of the day with one of the amazing prompts that he put together for us.

00:18:32.000 --> 00:18:33.000

Yeah.

00:18:33.000 --> 00:18:36.000

Hey, let me get my screen shared here. I too many screens open. Those of you are on zoom all the time can probably empathize with it

00:18:36.000 --> 00:18:41.000

Relate relating right now. Can you make it slightly better? I think like let's make it like 3 times bigger if we can.

00:18:41.000 --> 00:18:49.000

Okay. Okay, I've got a huge monitor here and I'm trying to

00:18:49.000 --> 00:18:50.000

That better.

00:18:50.000 --> 00:18:55.000

There we go. Okay. Like, maybe just, yeah, that's one more.

00:18:55.000 --> 00:18:56.000

Perfect. There we go. Yep.

00:18:56.000 --> 00:19:08.000

Okay, one more. Okay. Great. Well, one of the things that I think is almost forgotten, which is a shame, is the ability to just personalize content, which it was such a big part of differentiation.

00:19:08.000 --> 00:19:14.000

And I know for years I was in e-learning and we're always talking about new tools that were coming out that allow you to customize, create personalized learning paths, create all this amazing branching.

00:19:14.000 --> 00:19:33.000

Yeah, turn your content into bandersnatch where you could go through and see different outcomes and get different remediation and the technology is really fallen short on that in my opinion thus far or there been standards and technology in place that were just simply too complicated for a practitioner to be able to go in and achieve that level of personalization.

00:19:33.000 --> 00:19:44.000

So at this prompt, we're going to take a very simple sample product and we're going to looking.

00:19:44.000 --> 00:19:52.000

At high level of genetics. I'm just gonna copy. Opt here. And I'm gonna go to chat GPT.

00:19:52.000 --> 00:20:02.000

I'm in 3.5, which against the free version. I'm going to post this.

00:20:02.000 --> 00:20:03.000

And notice.

00:20:03.000 --> 00:20:06.000

Do you wanna just go? Like why don't you just go through the structure of the prompt first and then we can kinda walk through the outputs.

00:20:06.000 --> 00:20:11.000

Oh, okay. I think I actually, it pasted the wrong one anyways. I think that's what happens.

00:20:11.000 --> 00:20:14.000

Okay.

00:20:14.000 --> 00:20:17.000

So.

00:20:17.000 --> 00:20:21.000

Okay, here we go. I don't know why it pasted another one there. Okay, so a little bit.

00:20:21.000 --> 00:20:23.000

Technology.

00:20:23.000 --> 00:20:29.000

Do you want me to share the one the one slide of the hierarchy you just want me to talk through it?

00:20:29.000 --> 00:20:31.000

Okay, socker, it would be great.

00:20:31.000 --> 00:20:39.000

Okay, so one of the things, you know, with a prompt, there's all these great and we have a number of this to a number of really simple one-liner prompts that you copy and paste and you reuse as is.

00:20:39.000 --> 00:20:52.000

But one of the things I think is really useful early on in using chat GPT is start thinking about your prompts in very modular ways.

00:20:52.000 --> 00:20:57.000

So you're setting yourself up for res to the prompt from the very first draft that you're creating it.

00:20:57.000 --> 00:21:06.000

So you want to think about what is it you're trying to generate what personas do you want to use what products do you want to have on the back end?

00:21:06.000 --> 00:21:25.000

And if you start thinking about that in modular ways, you can very easily take something you've created for one domain, like maybe you're teaching about Niche and there's some concept in there that's really complex that may be too complicated for a a philosophy one on one but maybe not for a graduate course, you know, in NHA.

00:21:25.000 --> 00:21:37.000

But you can take that same prompt and change a few things out and have it do something very similar. And a complex topic such as quantum physics, which is something maybe you look at here and a couple of prompts.

00:21:37.000 --> 00:21:39.000

But here we have a prompt and, the prompt is, you know, you're an expert ID with expertise and building personalized target high quality content.

00:21:39.000 --> 00:21:57.000

You have some questions stems down here. And you want to take the personal interest in Japanese anime and take the concepts that are presented here in these question stems.

00:21:57.000 --> 00:22:06.000

And come out with some questions related to that. Okay, so the other thing that I can do quite quickly is I can go back to the prompt.

00:22:06.000 --> 00:22:12.000

I can edit this prompt and I'm gonna change this to, from Japanese anime.

00:22:12.000 --> 00:22:19.000

To NBA players.

00:22:19.000 --> 00:22:24.000

Okay, so I'm gonna save and submit.

00:22:24.000 --> 00:22:38.000

Okay, and then again, it gets back to my question. Tim's I have here, for example, on twins, biological concepts of nature versus nature and nature versus nurture and then family culture, you know, what role do those things play?

00:22:38.000 --> 00:22:57.000

And I can go back and I can. Edit this one more time and I'm gonna change the interest to young adult novels

00:22:57.000 --> 00:23:05.000

Okay, so here we have this for the, the young adult novels. And I have a follow up, prompt for this and I have.

00:23:05.000 --> 00:23:11.000

And I want to go in and create a.

00:23:11.000 --> 00:23:18.000

Quiz with question stems and a question key. So I have this prompt that I'm pasting in.

00:23:18.000 --> 00:23:25.000

And it says now take, each of the examples you generate above and convert each bulleted example into a multiple choice question with 4 options.

00:23:25.000 --> 00:23:44.000

The correct option being in bold.

00:23:44.000 --> 00:23:45.000

I did not want to give you bold.

00:23:45.000 --> 00:23:49.000

So that, You didn't want to give me both. It's giving me bold every time.

00:23:49.000 --> 00:23:51.000

Yeah.

00:23:51.000 --> 00:24:03.000

But didn't want to give me bold this time around. But that but that's that's just a quick and dirty example how you can take a topic you're pulling together pop quiz, you may have different interest groups of students, different personal interests.

00:24:03.000 --> 00:24:17.000

In the classroom. There may be something fun coming out, a movie coming out, there may be any number of things that are coming that might tie into personal interest that'd like you to very quickly take some very generic question stems.

00:24:17.000 --> 00:24:34.000

Such as these here. And contextualize them around a personal interest area rather quickly and generate one quiz it could be or you could create again going back to the scalability capability you could create 5 or 6 different quizzes all on the same subjects, all covering the same learning objectives, but with a different level.

00:24:34.000 --> 00:24:44.000

Of contextualization based on the personal interest of the students in your classroom.

00:24:44.000 --> 00:24:59.000

That's great. And I love, I don't know if you like, the little bit of the, someone's been using so I don't know if everyone knows that you can change a prompt and keep regenerating and that's a great way to like kind of not not have to cut and paste or ask again, use the same structure.

00:24:59.000 --> 00:25:12.000

And so the fact that that happened, it was really great with. And so it's what I would say though just as a G, if we kept going and even though it didn't give us the bowl, didn't follow it directly, we can ask it now.

00:25:12.000 --> 00:25:17.000

Please identify and answer key. So I can go and say, you know, create an answer key for this.

00:25:17.000 --> 00:25:27.000

There we go. And so even though sometimes it doesn't act like we talked about that that you know that space child that sometimes work exactly like we want is sometimes doesn't.

00:25:27.000 --> 00:25:32.000

This is an example of how to keep rolling and so a lot of that magic happens after that. First prompt.

00:25:32.000 --> 00:25:45.000

And so when you ask a question, someone asked a question about hallucinations. There's really not a beautiful like benchmark for this, but it's somewhere probably between the latest benchmarks say around 3 to 5% for chat TBT.

00:25:45.000 --> 00:25:53.000

The both versions and more with the with Claude and then even more with Bard but now Bard has changed to Geminis.

00:25:53.000 --> 00:26:01.000

We don't actually know, but we know that it hallucinates around some things like sometimes it'll forget a piece, it'll drop it, sometimes it'll make up a factor or figure.

00:26:01.000 --> 00:26:05.000

It's not very, it can't work count. It's bad at a mnemonic devices.

00:26:05.000 --> 00:26:16.000

It's not very good at complex map. So there are a lot of areas in which we want to avoid kind of trusting these tools.

00:26:16.000 --> 00:26:17.000

Yeah, can I jump in?

00:26:17.000 --> 00:26:18.000

And so, but it's a great question. So do you mind if I do one now, Brian, if that's okay?

00:26:18.000 --> 00:26:22.000

So that's a great question. Who should the attribution go to and happy, do you know it's correct?

00:26:22.000 --> 00:26:23.000

So there are other tools we can look at maybe at the end like perplexity which is more generative search.

00:26:23.000 --> 00:26:40.000

It would not give you this level of like amazing, you know, output in terms of these questions, these interests, but it can give you, you know, an indication of the source and sources.

00:26:40.000 --> 00:26:48.000

There is just announced today. I don't know if you saw this, Brian, but there's a relationship now between Chats, UVT and Open AI and Axel Springer, like Springer databases.

00:26:48.000 --> 00:26:53.000

And so we will start to see for the first time sources for some of the work that's done.

00:26:53.000 --> 00:26:54.000

Oh, wow.

00:26:54.000 --> 00:26:59.000

Some of the outputs, which will be very similar to what you see on Bing and more and more of Bard.

00:26:59.000 --> 00:27:04.000

So. I would say we're getting better, Tim. It's not quite there yet, but it is getting better.

00:27:04.000 --> 00:27:09.000

So I'm actually gonna get a do some live prompt engineering with you guys. I don't want I have to get involved too.

00:27:09.000 --> 00:27:31.000

And so we're going to do this is one of the ones that I worked with. Brian on which is simplifying concepts and differentiating using AI and so we know that like one of the most like we talked about it one of the best ways to differentiate is just giving students the right level of content for what they need and that can be a reading but we also don't even get to level of like what about descriptions

00:27:31.000 --> 00:27:38.000

or explanations or questions themselves like this is where we can get even deeper and this is really exciting. So if you look at the original prompt and I'll make this slightly bigger so everyone can see it.

00:27:38.000 --> 00:27:48.000

If I can get my, there we go. Here, you're an expert instruction. We always prime it.

00:27:48.000 --> 00:27:52.000

I've, the way I've been talking about this, Brian lately, is like, Chat T is trained on the internet.

00:27:52.000 --> 00:27:58.000

It is an enormous model and it could do so many things. But it's like taking a sledgehammer.

00:27:58.000 --> 00:27:59.000

Yeah.

00:27:59.000 --> 00:28:03.000

To a needle. So what's happening is you have this sledgehammer and I'm answering this needle question.

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

What could happen is I can actually like break the needle. Like it doesn't it doesn't really work that way.

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

So the more I cut down the context of what I want it to do and give it context. And direction, it means that I get down to maybe like a ball pean hammer, like a small hammer and I can actually work on it better.

00:28:17.000 --> 00:28:25.000

And so when we do this, we always start with that framing and we are specific as possible. And so what I'm gonna do is I'm gonna take the example prompt.

00:28:25.000 --> 00:28:31.000

And I'm gonna go over to chat GBT, my buddy. Let's see where we are.

00:28:31.000 --> 00:28:36.000

And oh wait, no, we don't do that. So chat, GBT.

00:28:36.000 --> 00:28:39.000

You'll notice I used it a lot today because I did 3 h of presentations for Avon School District.

00:28:39.000 --> 00:28:51.000

Shout out Avon, you guys were great. And also did the training on these 2 where I tested them out as I was doing it so today I'm just open that new context window.

00:28:51.000 --> 00:28:58.000

Remember it's going to remember our conversation up to a certain point. In this context window, I always like starting a new one.

00:28:58.000 --> 00:29:10.000

I'm going to enter it in here and what we see is that you know your tenth grade English teacher and we're talking about satire for Jonathan Swift's a modest proposal and I'm gonna put this in and what I've asked to do is actually give me 3 levels of explanation.

00:29:10.000 --> 00:29:19.000

Of satire within this novel. So it's identified. You can see though I've been very specific, label each explanation with the corresponding skill level.

00:29:19.000 --> 00:29:26.000

What I have now is I have beginner introductory. So I have intermediate and I have advanced.

00:29:26.000 --> 00:29:37.000

And so you can see that the level of you know, conversation gets longer, more more robust, you see more examples, and you get to see more complex vocabulary.

00:29:37.000 --> 00:29:48.000

And so we see that happening pretty easily. And if I do what, you know, what we just looked at with Brian is I can go and I can say now I want to do it with.

00:29:48.000 --> 00:29:53.000

A different book. So I'm actually gonna do it now with.

00:29:53.000 --> 00:29:59.000

That's a good tenth grade book. Maybe let's do, East of Eden.

00:29:59.000 --> 00:30:03.000

By Steinbeck.

00:30:03.000 --> 00:30:06.000

Good choice. Good choice.

00:30:06.000 --> 00:30:07.000

I was a really weird kid everybody probably I'm surprising but I was also an autodidact.

00:30:07.000 --> 00:30:17.000

So I read a lot and I was left to my own devices and my eighth grade, this is a lot of frustration.

00:30:17.000 --> 00:30:24.000

My eighth grade book report, I did oh gosh what's the famous time back book

00:30:24.000 --> 00:30:28.000

My.

00:30:28.000 --> 00:30:29.000

Okay.

00:30:29.000 --> 00:30:30.000

No, the other. Great Iraq. And I wrote this whole thing, including like about the last scene, which is not very appropriate for eighth grader if you've read it.

00:30:30.000 --> 00:30:40.000

And I wrote this whole thing, including like about the last scene, which is not very appropriate for a day grader if you've read it. And I scared the living crap out of my English teacher.

00:30:40.000 --> 00:30:44.000

She was like, I don't even know how to grade this. It is not appropriate for the student to do.

00:30:44.000 --> 00:30:45.000

It's awesome.

00:30:45.000 --> 00:30:52.000

So I was a weird kid. But anyway, so what you see here again is we have new set satire, a basic introduction, about intermediate and the answer.

00:30:52.000 --> 00:31:02.000

You're going to notice is that it's actually not nearly as robust because satire is not as big of an issue or a construction or literary construct within grapes of within East of Eden.

00:31:02.000 --> 00:31:11.000

So you can see that it's already tailoring that to the type of, you know, like the type of literary construct, the book itself, and also giving you leveled explanations.

00:31:11.000 --> 00:31:21.000

And so as you can keep going, because we always want you to keep going. As you can do things like create assignments so I can do this now where I'm actually going to take an assignment.

00:31:21.000 --> 00:31:30.000

I'm actually gonna do this one for I'm gonna have it create an assignment. And I'm gonna go back though, create an assignment for each on the suggestions for.

00:31:30.000 --> 00:31:36.000

A modest proposal, cause it's more appropriate. And now what I've got is.

00:31:36.000 --> 00:31:53.000

Now 3 different assignments. Based on the 3 different explanations. And so what you'll see is you have a 300 where short essay you've got a longer essay that has more quite like more purposes or more components and then you have a much more robust final essay.

00:31:53.000 --> 00:31:57.000

And so you can see as we can see in maybe a sets, I would say essays are really hard right now.

00:31:57.000 --> 00:32:01.000

Because chatplots could do them. So maybe we don't do an essay, so I'm gonna say.

00:32:01.000 --> 00:32:10.000

Create an assignment. That is not an essay. So let's see what they come up with.

00:32:10.000 --> 00:32:13.000

And so now I go into a satirical infographic, like how fun. And so it didn't give it to me level because I wasn't specific to that.

00:32:13.000 --> 00:32:20.000

But here's an example of how where we're starting to look at different ways. And models of doing this work.

00:32:20.000 --> 00:32:32.000

And a pretty easy, like this is 3 min, you know, and it was able to do a pretty good differentiation based on level, but also 2 books where one satire is a big part of and one where it's not.

00:32:32.000 --> 00:32:36.000

So it's a good example. So we're going to go to our next prompt. Do you want to show the next one?

00:32:36.000 --> 00:32:44.000

Yeah, sure. And while I'm bringing this up, you know, having worked with a, a lot of teachers and a lot of instructors in training settings.

00:32:44.000 --> 00:32:48.000

You know, we wanted spending a lot of time, especially in technical space, looking at students who need remedial help.

00:32:48.000 --> 00:32:54.000

So if we're teaching a cloud computing workshop or a DevOps workshop. Perhaps we have students that don't know enough Linux commands, right?

00:32:54.000 --> 00:33:07.000

But I can tell you once we started rolling this out in anger, and you're again the plasticity of these props is you're creating something for remedial students.

00:33:07.000 --> 00:33:13.000

It's a great opportunity and point of time to do the same thing for the advanced students and vice versa.

00:33:13.000 --> 00:33:29.000

So I think having tools like this where it's so rapid and so easily to make minor modifications to your prompt to reach out to another audience is something you should always be aware of and that kind of modular, plastic type, middle model for, building props.

00:33:29.000 --> 00:33:39.000

But this last one that I wanted to share is creating a virtual science tutor and a chat pot. I believe we have another one in here as well that's for a language.

00:33:39.000 --> 00:33:50.000

Tutor so one of the really cool things that people may not see when they first started using a chat GPT and using props is you can create intelligent tutors.

00:33:50.000 --> 00:33:59.000

And it is a very powerful thing to do. It's very good for, personalized learning and individual and we want to take this example.

00:33:59.000 --> 00:34:11.000

Of working with. Building a tutor for quantum. Teach students quantum computing. And in this particular instance, these students, they've already had Newtonian physics.

00:34:11.000 --> 00:34:21.000

So they need a gentle introduction and transition into thinking more like a quantum physicist. By working with this intelligent tutor.

00:34:21.000 --> 00:34:30.000

Okay, so some key aspects of this is you're an expert physics tutor. As, Amanda pointed out this has been trained on everything.

00:34:30.000 --> 00:34:32.000

So there's almost every imaginable persona that you can pull together. So that's really important for building the context.

00:34:32.000 --> 00:34:40.000

You're tutoring me and I'm an advanced high school. I be physics students. I'd like you to help me learn the core concepts of quantum physics.

00:34:40.000 --> 00:34:50.000

By contrasting, query, comparing this to Newtonian physics. Hackock and human tutors start asking me basic questions.

00:34:50.000 --> 00:34:59.000

Scale up, this last partially important. The scale complexity, subsequent questions up or down per the accuracy and completeness of the responses.

00:34:59.000 --> 00:35:06.000

In this context window. When I failed to respond to one of your questions, accurately, and completely, please provide detail and support of feedback.

00:35:06.000 --> 00:35:18.000

So you can see right away that you could use the same kind of form and construct. And go in and swap out some of the tokenized things like the student persona and the subject matter expert and the product they need to create, etc, to very rapidly reuse this.

00:35:18.000 --> 00:35:33.000

Okay, so it starts up here. How do you describe the fundamental differences between classical and Newtonian physics and I could say something like the nature of.

00:35:33.000 --> 00:35:47.000

Waves and state. Okay, and it'll go think about that and provide a response.

00:35:47.000 --> 00:35:48.000

Okay.

00:35:48.000 --> 00:35:50.000

Whoops. Is part of the fun of working live. Is the failure. Okay, so it looks like you mentioned some key aspects.

00:35:50.000 --> 00:36:01.000

For this a bit further. Additionally, you mentioned state are you referring to the concept of quantum states, blah, blah, blah.

00:36:01.000 --> 00:36:19.000

And I can say yes. I was referring. Explicitly to, super position.

00:36:19.000 --> 00:36:20.000

Yeah.

00:36:20.000 --> 00:36:21.000

Hello that this is what Brian is demoing because I do not know enough about this topic to even be tutored.

00:36:21.000 --> 00:36:28.000

This is what Brian is demoing because I do not know enough about this topic to even be tutored in this manner because it says this is, what's great about this is it's coming from a level of students existing understanding.

00:36:28.000 --> 00:36:36.000

And so for me, probably not where I would start, but like I think that this is such a great example of like meeting a student where they were and we actually got a question from high ho Lou about this idea of like students may be feeling like less or better.

00:36:36.000 --> 00:36:51.000

Or like less than because we're giving them a more simplified explanation or a different explanation.

00:36:51.000 --> 00:36:52.000

Yeah.

00:36:52.000 --> 00:36:54.000

And something that we do want to get away from, we definitely don't want to reinforce the kind of blue team or Robin or you know the different kind of leveling that happened in the past.

00:36:54.000 --> 00:36:58.000

And so it's on the path towards something more like what just Brian showed us, which is really meaning the student where they are.

00:36:58.000 --> 00:37:09.000

And even though chatting T is not going to be ideal for that yet, it is where I think we're going as a lot of companies.

00:37:09.000 --> 00:37:16.000

Sorry, sorry, I say your name, hi, how sorry I said your name, but thank you for telling me how to say it because I hope you're here again and ask another question.

00:37:16.000 --> 00:37:22.000

And so but the idea that like intelligent tutoring is more and more what we're focusing on in an tech world.

00:37:22.000 --> 00:37:23.000

Yes.

00:37:23.000 --> 00:37:32.000

And so while I do know though, and I know this for a fact because we did a web, we did a panel on Monday where we had 2 students talking about how they used AI.

00:37:32.000 --> 00:37:50.000

Both of them used it for high level tutoring one for multivariable calculus, which maybe shouldn't have maybe, maybe J, 3.5 is the great for that but another student use it for economics and so students are using this to help them study to reinforce to meet them where they are more and more.

00:37:50.000 --> 00:37:54.000

And I think that we have to recognize that as a very common use case.

00:37:54.000 --> 00:38:00.000

Yeah, I think 2, man, I'm gonna point out, yeah, I worked with a remediation systems.

00:38:00.000 --> 00:38:16.000

For teachers in the state of Georgia who did not score well on differentiated instruction and we found one of the obstacles was the fact that they had to give different assignments to different students and the parents would say, well, hey, Joey is doing a different assignment than Billy, you know, next door here.

00:38:16.000 --> 00:38:32.000

So the ability to explain differentiated instruction in the ability to understand like I was crushed when I was in Glee Club.

00:38:32.000 --> 00:38:33.000

Okay.

00:38:33.000 --> 00:38:37.000

I was in fifth grade because I got pushed to the blackbirds. I need the blackbirds. I need the blackbirds with the people that couldn't sing. I do that.

00:38:37.000 --> 00:38:54.000

I live with that till this day. And so it still makes me sad. But I think, you know, that's why to when you're looking at these prompts is to have a 3 60 is like okay I'm looking at this audience okay and I think that's a great place to start with the prompt is looking at addressing a specific issue for a given audience and then continue to look around the room at your different

00:38:54.000 --> 00:38:59.000

audiences because that's what it's all about, right? It's understanding of different fountains of knowledge.

00:38:59.000 --> 00:39:01.000

You have different fountains of life experience and culture and all these different things and you may be seeing something that's really visible.

00:39:01.000 --> 00:39:16.000

On the surface that I can I can go grab this but that's a great opportunity to say okay I've focused in on this what's in the periphery what's adjacent to this.

00:39:16.000 --> 00:39:20.000

What are some other groups that I may not be thinking about. What what are some of their individuals in my classroom?

00:39:20.000 --> 00:39:24.000

I mean, I'd be thinking about and how can I start playing with that within my prompt empirical session sooner rather than later.

00:39:24.000 --> 00:39:31.000

And in the development of that

00:39:31.000 --> 00:39:53.000

Absolutely, and I think that also sometimes it's about choice as well, like you know, like how do you like how are you on this?

00:39:53.000 --> 00:39:54.000

Yeah.

00:39:54.000 --> 00:40:00.000

Are you emerging with your knowledge, you feel really confident, do you like what level of challenge and that you work with students around meta cognition and self reflection where you're not letting them kind of stay in the blackboard because you know what a lot of people can move out of blackbird right because they do training and maybe they take it a different way, is that also giving students voice and agency where you start, taking away the stigma

00:40:00.000 --> 00:40:16.000

of having different levels of like meeting you where you are because the most important thing is not being like everybody else but being the best version of yourself the best version the most the person that's most supported to improve because that's going to really get us and that's what the zone approximately development is all about which is that idea.

00:40:16.000 --> 00:40:22.000

It's really being meaning you where you are and giving you like an equity play instead of a quality play.

00:40:22.000 --> 00:40:23.000

Yeah.

00:40:23.000 --> 00:40:28.000

And I think that this is where, but it does require you know, desegmentation, it requires very thoughtfulness.

00:40:28.000 --> 00:40:43.000

It requires us not putting kids into like buckets so they can never leave, right? And so we don't we want these to always be like, so one student might need an emerging piece for quantum physics like me, but then not need it for Jonathan Swift.

00:40:43.000 --> 00:40:49.000

Like, and that I like, and I'm moving at the rate in which I need to. And again, these tools like chat to BT are.

00:40:49.000 --> 00:40:56.000

Very early stage. They are not going to be the interface of the future. The interface of the future will be really, really different.

00:40:56.000 --> 00:41:01.000

It's just that right now we can actually start doing some of this stuff without having to wait for that new interface.

00:41:01.000 --> 00:41:11.000

And so last prompt for today, which is the one in which is we talked about already cultural pieces. So those are my text messages.

00:41:11.000 --> 00:41:12.000

Thank you.

00:41:12.000 --> 00:41:24.000

So let's not do that. Friend is good is his PhD student and is about to go home so let's go to my, go back to here we go.

00:41:24.000 --> 00:41:31.000

There we go. Resume share. And so culturally contextualized activities or culturally responsive activities.

00:41:31.000 --> 00:41:51.000

And so what we have here is like we teach tend to teach a very, like not only do we teach a very global North inspired curriculum and about ourselves, but we also our chat bots are also very global north so I saw some people from Malaysia and like Thailand it's going to be quite different and so what we want to do here is like we actually see and you'll see that there's that like

00:41:51.000 --> 00:42:00.000

a formatting piece as well of actually the way we formatted the prompt or there's some space and some like some real strong like signals as to its 2 different tasks.

00:42:00.000 --> 00:42:07.000

But what we wanted to do is actually create essentially, let's see, go and check T, new context window.

00:42:07.000 --> 00:42:21.000

I'm gonna drop this in. Gonna make sure everyone can see it but you know we're talking about Independence Day for the US and we want to see we want to look at students and if you're from these places let us know if there's some bias in this but are stereotypes.

00:42:21.000 --> 00:42:34.000

But the first task is about some creating 3 2 3 paragraphs on similar Independence Day events for these these non-US countries and then a social activity and then what I put here at the bottom is be careful not to use cultural stereotypes.

00:42:34.000 --> 00:42:41.000

And your answer and if you're not sure about students culture, do not create a response for that culture.

00:42:41.000 --> 00:42:47.000

And the reason why I did this is that there is research that shows that it was just released by anthropic.

00:42:47.000 --> 00:42:52.000

If you tell chat to TT or Claude not to be biased, it will be less biased.

00:42:52.000 --> 00:42:58.000

And if you tell it's okay to not have an answer, it will avoid answers that are potentially fraught or not good.

00:42:58.000 --> 00:43:05.000

So what we have here is we have The first task, we have Thailand, we have Australia, we've got Mexico, Guatemala.

00:43:05.000 --> 00:43:23.000

And I will say I lived in Australia and so Australia Day is actually not considered to be it's reconciliation as the other term for it, which is like this idea that the other term for it, which is like this idea that Australia Day was actually about the decimation of the other term for it, which is like this idea that Australia Day was actually about the decimation of the indigenous peoples and

00:43:23.000 --> 00:43:27.000

the fact that Australia Day was actually about the decimation of the indigenous peoples and the fact that it identifies this in the bottom.

00:43:27.000 --> 00:43:31.000

And the fact that it identifies this and the fact that it identifies this in the bottom, it's a day of efforts towards reconciliation is a good example that it' And then also here we have the task.

00:43:31.000 --> 00:43:33.000

So we actually now we're going to go into the different components and we're going to start to think about what this looks like.

00:43:33.000 --> 00:43:44.000

So go into pairs or small groups and then think about what this looks like within your culture. And I think that that's a great way to like, So do this, right?

00:43:44.000 --> 00:43:56.000

Because what we've done is A, I'm using like my knowledge, right? And I would double check this and again, like, you know, make sure and we've already signaled to the bot not to reinforce stereotypes or biases.

00:43:56.000 --> 00:44:10.000

But we wanna make sure that this is something that our students feel comfortable as well. So what we did is that because we believe in AI literacy being not just for you all but for students, is that actually one of the the takeaways is have students evaluate the outputs to see if there is bias or a lack of cultural understanding and have that be a rich discussion in your classroom.

00:44:10.000 --> 00:44:21.000

And so these are opportunities like we're not expecting. We're not going into this saying, hey, this is perfect, right?

00:44:21.000 --> 00:44:28.000

But we're saying is that this is something that can get us sparked and thinking and may actually carry with it extensive biases.

00:44:28.000 --> 00:44:38.000

So let's have a conversation about these technologies. But you can keep going and the fact that in one prompt you did essentially 4 different explanations and a full task.

00:44:38.000 --> 00:44:46.000

With one prompt is a great example of how with structure But, Brian talked about the modularization of the proms.

00:44:46.000 --> 00:44:54.000

You can actually get to really targeted really nice outcomes that are pretty complex with very little, you know, very little effort, so to speak.

00:44:54.000 --> 00:45:05.000

So. We're gonna come off share. Like so we know that we, you know, we always try to end at 45 if you have to head off and considering it's the middle of the night for some of our audience.

00:45:05.000 --> 00:45:06.000

If you have to head off and considering it's the middle of the night for some of our audience.

00:45:06.000 --> 00:45:10.000

It's the middle of the night for some of our audience. And we're gonna drop, so Brian's, LinkedIn, and we're gonna drop, so Brian's LinkedIn and my LinkedIn in the chat because that's one of the best ways to interact with us.

00:45:10.000 --> 00:45:24.000

But we're gonna like take some questions. So if you any questions about differentiation, we have 3 more prompts that we didn't do.

00:45:24.000 --> 00:45:25.000

Yeah, remember.

00:45:25.000 --> 00:45:28.000

One was a language tutor. One was what was the other weather 2? One was Yeah. So many, when was the language tutor?

00:45:28.000 --> 00:45:36.000

One was, hold on, we'll get you 2. The language tutor.

00:45:36.000 --> 00:45:51.000

We have a. And instruction translator for ENL students. And a, like you can part, I think, yeah, and then, oh, a language practice partner.

00:45:51.000 --> 00:45:56.000

So there were 2 so language is a really big part that we didn't talk a lot about but Thank you.

00:45:56.000 --> 00:46:01.000

Language is like we have language tutor and language partner and sorry, language translator and a language partner.

00:46:01.000 --> 00:46:08.000

We know that like conversational like AI, which is what a chat bot is, is designed to have a conversation with you and at least with some of the very core common languages like Spanish, sometimes French, others that they do a really good job.

00:46:08.000 --> 00:46:19.000

And that conversation. So if you have any questions, let me know, but I'm gonna ask the first question because you were talking about formatting earlier, Brian.

00:46:19.000 --> 00:46:27.000

So how how can you take an output like this and put it into a slide? Is there is there a trick or a hack to do that?

00:46:27.000 --> 00:46:33.000

Is it possible?

00:46:33.000 --> 00:46:34.000

Yeah.

00:46:34.000 --> 00:46:37.000

You know, I'm a sadist, so I've taken to doing all the formatting myself, but there are tools.

00:46:37.000 --> 00:46:45.000

I think one of them is called rotible or ridability and a couple other tools that allow you to paste the output as markup.

00:46:45.000 --> 00:46:59.000

Into this tool and then you can copy it out of that tool. With formatting that carries over to word. And I think it's, I think it's really ironic how powerful this tool is.

00:46:59.000 --> 00:47:06.000

And I spend the majority of my time with formatting. You know, again, the prompt getting the contents not a problem, but getting it to look the way that I need to sometimes can be.

00:47:06.000 --> 00:47:11.000

So I, I think, I think there'll be a bevy of tools in 2,024 and things built into Word and PowerPoint and built into your browser that will allow you to convert that mark up.

00:47:11.000 --> 00:47:19.000

That that's a big one.

00:47:19.000 --> 00:47:34.000

Yeah, and it's, if you want to look at, so gamma, almanac, or Kieropod have these essentially you can create slides they all use chat to VT API and other APIs from other chat bot tools.

00:47:34.000 --> 00:47:43.000

They're all right now getting better like like everything they're gonna have issues, especially around image generation, but that's kind of the way to do it right now.

00:47:43.000 --> 00:47:50.000

Like, You know, the 80 20 weirdly is maybe like 95 5 and that other 15% is formatting.

00:47:50.000 --> 00:47:51.000

Yes, yes.

00:47:51.000 --> 00:47:55.000

It's getting into the right format, making it work, but that will get better and better as we see these tools natively in the productivity suite.

00:47:55.000 --> 00:48:01.000

Like Microsoft Co-pilot, which hopefully will get significantly less expensive over time. It'll be something we just use as part of our kind of daily lives.

00:48:01.000 --> 00:48:15.000

So it looks like we don't have any other questions. I just want to say like it's so amazing to have like So much like deep knowledge, like, you know, I love that Brian comes from.

00:48:15.000 --> 00:48:21.000

He's like coding Dojo. He's talking about quantum physics and Linux, but it's all about being an educator.

00:48:21.000 --> 00:48:27.000

It's all about thinking through the way we prompt and the way we meet our students need regardless of the level they're in.

00:48:27.000 --> 00:48:33.000

So I just want to say thank you so much, Ryan, for your wisdom, your effort, the new prompts and your time today and your friendship along the way.

00:48:33.000 --> 00:48:34.000

Okay.

00:48:34.000 --> 00:48:36.000

So just really appreciate you.

00:48:36.000 --> 00:48:38.000

My pleasure. Thank you for having me.

00:48:38.000 --> 00:48:39.000

Absolutely. And then of course, like everyone that is here, I'm so glad that you're here.

00:48:39.000 --> 00:48:54.000

We love working with you guys and I hope to see you next at our next webinar. Please go to benefits the middle of the night or have an amazing day in Australia or have a good night.

00:48:54.000 --> 00:49:03.000

We love having you and join everyone here. So have a great night and if I don't see even for the next before the end of the year. So have a great night.

00:49:03.000 --> 00:49:04.000

And if I don't see you for the next, before the end of the year, have a great holiday season.

00:49:04.000 --> 00:49:06.000

Thanks, everybody