The header image was generated by DALL•E with the prompt “I'm writing a post on how to use AI as a physical education teacher. Can you help me with a header graphic. I'd love to have a school gymnasium in the style of a modern 3d animated movie. In the middle of the graphic, can you create a brain that is made up of a bunch of connected nodes (to represent a neural network)?”
These are the show notes for my latest episode of The #PhysEd Show Podcast. Take a listen to the episode by using the player below or subscribe to the show in your favourite podcast app!
Subscribe to The #PhysEd Show in all of its formats to make sure you never miss out!
The Podcast: Subscribe on Apple Podcasts | Spotify | Google Podcasts
The Vlog: Subscribe on YouTube
Live Episodes: Subscribe on Facebook
The only thing wilder than the rise of AI has been the brain-numbing speed at which it is occurring.
ChatGPT, the first true AI household name, launched on November 30th 2022. By March of 2023, OpenAI launched the ChatGPT API, which made it possible for software developers to bake AI chatbots into their products. Suddenly, every app I use on a daily basis had AI access just a few keystrokes away.
Fast forward to today, and we have heavy hitters such as Google's Bard (now Gemini), Microsoft Co-Pilot, and rumours of a fully revamped AI-powered Siri on the horizon. Not to mention thousands of plugins that allow you to do everything from generating videos and images, to summarizing research papers, and more.
I'm not going to lie, it's been overwhelming and it took me a while to figure out how I felt about it all.
I'm someone who enjoys producing creative work. Being creative is a big part of my identity, and it's something that I'm very proud of. To watch AI effortlessly produce work at such mind-boggling speeds felt, well, intimidating. This is a big reason why I was so late to the AI party.
However, over the last few months I've figured out how AI can fit into my life in an empowering way. I'm now a full-on AI convert and I feel excited about the potential that AI is unleashing in my work and my teaching!
In this post, I want to help you feel a little less overwhelmed by the AI revolution that we are now finding ourselves in. I would also like to share some of the ways that I am currently using AI to maximize creativity, free up headspace, and accelerate my professional learning.
Here are our learning targets for today:
Let's dive in!
As teachers, we know how important academic language is. It's what allows us to engage with learning materials by making vocabulary more accessible.
There's value in knowing the different terms (and acronyms) that casually get thrown around in discussions about AI. To that point, I thought we should begin this deep dive by defining key terms:
To make sense of Artificial Intelligence, we first need to define intelligence.
When we talk about intelligence, we're referring to the ability to acquire and apply skills and knowledge. For most of history, that kind of learning happened naturally by using our noggins to learn stuff and share it with others.
However, as computers became more and more capable and computer science advanced, we figured out ways to have these machines learn on their own. If you've played chess against a computer, used spell check to proofread a document, or asked Siri a question, then you've experienced artificial intelligence.
AI isn't something new, it's just reached a tipping point where today's technology is empowering it to become exponentially more capable than ever before.
Machine Learning, which has benefitted massively from advancements in computing power, is how artificial intelligence has become so much more robust over the past couple of years.
With Machine Learning, computers are able to learn without needing humans to intervene in their learning. Today's computers (a.k.a. machines) are able to go through insane amounts of data, make sense of that data, and make their own decisions based on the data they have access to.
Every time a machine tries something and is either successful or unsuccessful, a new data point is added that it can now use to make better decisions in the future. In other words, it has learned.
Unlike us limited humans who typically need to learn things sequentially (e.g. try one idea, then another, then another), computers have the capacity to try tons of ideas all at once. That's how they're able to go about things so quickly, including their own learning.
Language Learning Models are a type of machine learning that focuses on – you guessed it – language.
When you think of it, language is just a code. We say words, which represent ideas, which other people can interpret. This is the same way that the sequences of dots and dashes in Morse Code represent letters and numbers. A Morse Code interpreter looks at those sequences, and understands them as words which then form messages or ideas.
Computers use code too. For example, the browser you are viewing this blog post in had to read strings of HTML elements, interpret those as building blocks of a webpage, and then present it all as content you can consume.
But what if computers could read other kinds of codes that weren't originally designed for computers, like language. After all, language is insanely complex. It's full of expressions, metaphors, tones, and nuanced intentions. However, despite all of that complexity, it's still just code.
LLMs are designed to help computers learn and make sense of language. These models train computers to take into account all of the nuances of languages, beyond just the words being said. Not all LLMs are created equal, and they haven't always been great. If ever you've asked a digital assistant for something and received an unexpected/frustrating response back, then you've experienced the limited abilities of certain LLMs.
However, there have been incredible advancements in the robustness of Language Learning Models which have been possible thanks to incredible acceleration of both technology and computer science. This is why applications such as ChatGPT feel so surreal, because the LLM that it is built on is so good that it makes you feel like you are chatting with a human.
Speaking of ChatGPT, what the heck is a GPT?
GPT stands for Generative Pretrained Transformer. It's a model that is trained on mind-boggling amounts of text data. This training allows apps such as ChatGPT to pull on its broad and deep artificial intelligence (which it developed through Machine Learning and Language Learning Model it is built upon) to generate text that answers questions, produces written work, maintains conversations, and so much more.
There are many GPTs that are being built upon LLMs so that us humans can leverage artificial intelligence to learn and do more faster than ever before.
I want to acknowledge that some of you will have read that first part and thought "wow, this is amazing!". Others might be thinking "this is terrifying". A few of you, very much so like me, might be holding both opinions in your mind, and that's ok.
When I committed to using AI more to see how it fit into my life, it eventually led me to an "ah-ha!" moment. What I realized is that AI can be insanely empowering when it is used as a co-pilot. Let me explain:
One of our most powerful internal drivers is mastery. As human beings, we are hardwired to chase competency in areas we are passionate about. When we can clearly see how a task aligns with our passions and purpose, that clarity motivates us to master the skills, knowledge, and understandings required to complete the task effectively and efficiently.
Putting ourselves in the backseat and giving AI the wheel won't ever satisfy that internal drive for mastery. Tony Stark never asked Jarvis to fight Thanos for him so that he could take a nap.
That said, the road to mastery can be a frustrating one. At times, when the conditions are right, we get to experience the magical sense of flow in which time disappears and creativity pours out of every aspect of our being. Other times, we stumble over roadblocks and potholes that leave us feeling frustrated and drained.
In my experience, those obstacles in the way of mastery present themselves as gaps in my skills, knowledge, energy, and attention. Trying to push through doesn't always work, and setting the scene for another flow-driven session can feel annoying.
What I've come to realize is that AI can be used to help fill those gaps. Rather than slowing down or stalling, my use of AI allows me to maintain momentum and continue to close the gap between where I am and where I want to be.
In other words, it acts as a Co-Pilot (as Microsoft so aptly named their AI chatbot) that keeps me moving forward. With its help, I'm able to stay creative, meet my goals, and keep growing along the way.
Using AI in this way doesn't rob you of the amazing feelings that come with self-actualization. In fact, it empowers you to be at your best.
Here's an example to help you understand what I mean:
Let's say that I'm planning a fitness unit for my grade one students. I've determined the outcomes that I want to build the unit around and have started the unpacking process so that I can build the Learning Roadmap (i.e. the student-friendly, qualitative rubric) that will serve as the foundation of the unit.
At this point, I've already rewritten the curriculum outcome into student-friendly language and have determined the three learning indicators that will help me assess my students progress.
Although I'm satisfied with the outcome's wording, I'm feeling stuck regarding the language I've used for the learning indicators. I need those to be as student-friendly as possible so that my grade one students can understand them. If not, then I'm failing to invite my students to be active partners in their learning.
At this point in my planning, I've done a ton of research, writing, and mental lifting. In other words, I'm tired and running out of steam. I really want to nail this Learning Roadmap, but trying to push through the exhaustion is making me feel annoyed with the process. I mean, I am so close to producing the work that I'm envisioning in my head... I'm just stuck.
I need some help, so this is when I launch ChatGPT.
ChatGPT works best when it has a clear understanding of what you are asking it to do. This is why AI prompt writing is quickly becoming one of the most important skills you can develop. I've found that there is a lot of value in saving the prompts that you've used so that a) you can improve them over time, and b) you can quickly reuse them when they're needed.
Here is my conversation with ChatGPT. Pay attention to the way I wrote the prompt and how that prompt get used by the GPT. Also, pay attention to how I choose to be polite when working with ChatGPT as I'm not fully convinced AI won't one day rule the world (I'm only kinda joking).
Me:
Hi there! I'm a K-3 physical education teacher and I'm in the process of planning a fitness unit for my grade one class. I'm putting together the student-friendly, qualitative rubric (i.e. Learning Roadmap) for the unit, but I'm getting a little hung up on the language. If I provide you with the learning outcome and the three learning indicators that unpack it, can you help me put this thing together in a way that will make sense to my students?
ChatGPT:
Absolutely, I'd be glad to help with that! Please go ahead and share the learning outcome and the three learning indicators you have in mind for your fitness unit. Then, I'll assist in crafting a student-friendly, qualitative rubric (Learning Roadmap) for your grade one class.
Me:
The learning outcome is "I can make heathy choices in my life."
That outcome is broken down into three student learning indicators (a skill-based one, a knowledge- based one, and an understanding-based one):
- Skill-Based Learning Indicator: I can put into practice a variety of health-enhancing habits.
- Knowledge-Based Learning Indicator: I know what a habit is.
- Understanding-Based Learning Indicator: I understand the difference between health-enhancing habits and health-reducing ones.
My Learning Roadmaps map indicators out across four levels:
• Not Yet: Student is yet to show progress in their learning.
• Getting There: Student is beginning to show progress in their learning.
• Got It: Student is demonstrating learning that meets grade-level expectations.
• Wow: Student is demonstrating learning that exceeds grade-level expectations.
The three student learning indicators that I shared with you each represent their "Got It" level on the Learning Roadmap. Here's how I could use some help:
1. Can you re-write each of the indicators into language that is more accessible to grade one students?
2. For each indicator, can you write what they would look like at the "Not Yet", "Getting There", and "Wow" levels?
ChatGPT:
Certainly! Let's simplify and articulate these indicators in a way that's accessible to grade one students, and then describe what each level of the Learning Roadmap would look like for each indicator.
Learning Outcome: "I can make healthy choices in my life."
1. Skill-Based Learning Indicator:
Original: I can put into practice a variety of health-enhancing habits.
Accessible Language: "I can do things every day that keep me healthy and strong." Not Yet: I'm still learning about different ways to stay healthy.
Child-Friendly Description: "I'm finding out how to be healthy."
Getting There: I try doing healthy things like eating fruits and moving around, but sometimes I forget.
Child-Friendly Description: "I'm trying to eat good food and play a lot, but sometimes I don't remember."
Got It: I do things every day like eating vegetables and playing outside to keep my body happy.
Child-Friendly Description: "Every day, I eat healthy food and play outside to keep me feeling good."
Wow: I always remember to do many healthy things, like choosing water over soda and going to bed on time, without being told.
Child-Friendly Description: "I always remember to do lots of healthy things all by myself, like drinking water and sleeping early."
2. Knowledge-Based Learning Indicator:
Original: I know what a habit is.
Accessible Language: "I know what habits are." Not Yet: I'm still learning about what habits are.
Child-Friendly Description: "I'm learning about habits."
Getting There: I can tell you some habits, but I'm still figuring out which ones are good for me.
Child-Friendly Description: "I can name some habits, but I'm still learning which ones are good."
Got It: I can tell you what habits are and share examples of good habits, like brushing my teeth.
Child-Friendly Description: "I can tell you about habits and examples of good ones, like brushing my teeth."
Wow: I can explain what habits are, give lots of examples of good habits, and tell you why they're important for health.
Child-Friendly Description: "I can explain habits, give lots of examples of good ones, and tell you why they're important for staying healthy."
3. Understanding-Based Learning Indicator:
Original: I understand the difference between health-enhancing habits and health-reducing ones.
Accessible Language: "I know the difference between habits that are good for me and habits that are not."
Not Yet: I'm still learning which habits are good for me and which ones are not.
Child-Friendly Description: "I'm learning which habits are good and which ones aren't." Getting There: I can tell some habits that are good for me and some that are not, but I still get confused sometimes.
Child-Friendly Description: "I can tell some good and bad habits apart, but it's still a bit confusing."
Got It: I know which habits help me stay healthy and which ones don't, like eating fruits is good and watching too much TV is not.
Child-Friendly Description: "I know habits like eating fruits are good for me and watching too much TV is not."
Wow: I can give many examples of good and bad habits and explain why they affect my health.
Child-Friendly Description: "I can give lots of examples of good and bad habits and explain how they make me healthy or not."
This format presents the indicators and levels in a more relatable and understandable way for grade one students, ensuring they can grasp and self-assess their progress in making healthy life choices.
Me:
Thanks so much!
To understand what happened here, think of the last time you had to write a 20+ page paper. The most intimidating part of taking something like that on is the moment where you sit at your computer and stare at a blank screen. Getting that wheel turning is a heavy lift, which is why we often start with something easy like writing a title page, choosing a font, or making an outline. What we're really doing when we do those things is we're trying to find some momentum. We're trying to find some kind of trigger that will help us get into a flow state and produce creative work.
The answers ChatGPT gave me, although amazing when you consider the technology behind it all, weren't necessarily what I would move forward with. However, that doesn't matter because what it's answers provided me with were the raw material I needed to get unstuck and maintain my momentum. Suddenly, I'm creating again.
This is just one of the ways AI has found its place within my workflow. In the podcast, I’ll be sharing more examples of how I’m using AI, which tools I’m exploring, and work that I’ve generated by leveraging AI.
I know that AI can be incredibly overwhelming, and I hope this post serves as a primer for you! In my next post in this series, I'll get into some of the specific tools I've been exploring and how I'm currently using them to enhance my teaching.
Thanks for reading and happy teaching!
