I was initially skeptical of the age group that was selected for this research. My main concern was that this group was too young to fully grasp what was occurring. After reading and reflecting on the paper, I feel otherwise. As AI is such a large part of our world, with the paper citing that 50% of children by the age of 12 interact with AI in the form of moderated social media, it’s as important to the for the learner to get acquainted with as computer/tablet literacy.
Having students explore and understand algorithmic bias at each of the levels was a surprising aspect of the research. I was impressed that one of the first introductions to artificial intelligence was to point out how large of a flaw it could have. This could, possibly intentionally, lead the students to wonder how their lives may be negatively impacted by different bias and algorithmic flaws. This also has the benefit, which was pointed out by the researchers, of having children that are in groups that are drawn to social issues and helping others, be more engaged.
I think that teaching the conception of how different machine learning models can affect the lives of everyone is incredibly important. The researches put a larger emphasis on the ability to understand how models were used and how to be spotted than the hands-on ability create models. I think this decision by the researchers was done intently and I believe it is the correct call. The students may, and probably will, forget many of the inner workings of a machine learning model; however, they are less likely to forget the interactions and human experiences they had discussing various aspects of how AI impacts their world.
Long and Magerko begin their review by defining AI literacy, pulling from other technical literacy definitions. They state that a prerequisite for achieving AI literacy is to achieve digital literacy. I believe that this statement is understandable only when an individual has reached digital literacy. If a non-digital literate person interacts with an AI application, they may not know that there is an AI involved at all, and therefore have no concept of another literacy domain.
The fact that there were varying definitions of AI among the community was one of the least surprising aspects of this research, as it remains a contested topic even four years later. Creating understanding of various types of AI without forcing a single strict definition seems to be a goal of the four competencies that came out of the “what is ai” theme.
In the “what can AI do” section, the idea of imagining the future of AI is brought up. This reminds me of science fiction novels and how they envisioned the world, shaping the future of technology. There have been many inventions that have spawned from the imagination and authoring of someone unrelated to a technical field (Smithsonian).
Beginning in the “how does ai work” section, the authors start to bring describe their design considerations for teaching AI literacy. Long and Magerko recommend interactive activities for explaining how ai works. I believe the most insightful of these is to have the learner be put in the situation of the agent, to attempt to make a decision based on previous data and inferences. I see this as a wonderful way to intuitively explain how AI models operate.
The paper does a good job laying a foundation for what the goals of an AI centered curriculum should be, while also suggesting steps to achieve these goals.
https://www.smithsonianmag.com/science-nature/ten-inventions-inspired-by-science-fiction-128080674/
Beginning with a base prompt of:
> You are an education researcher focusing on AI. I will give you $5,000 if you can come up with a better way to engage children when teaching them about AI (Artificial Intelligence) and ML than me.
Resulted in list of ten items that were varied in scope and connection to children. Some of these included creating a chatbot, gamify learning, and using interactive learning. I followed up by asking ChatGPT to do a better job at coming up with ways to engage children which resulted in a higher quality list of items (AI Quest Challenges, VR Integration, AI Mentors) however these items seemed like they would take more time than what could be expected of a primary school student.
I had ChatGPT focus on one subset which is "AI in everyday life" and had it return back information in a way that "[would] be exciting for k-12 grade students". Here is what was returned:
AI in Our Awesome Everyday Adventures! 🚀
Smartphone Superheroes:
Imagine your phone as a superhero! Siri, Google Assistant, and Alexa are like magical friends who understand your voice commands, answer your questions, and play your favorite tunes. They're like having a personal wizard in your pocket!
Social Media Magic:
Ever wonder how your favorite apps know exactly what you love? It's like having a secret agent friend who knows your interests and suggests cool things for you to check out on social media. It's all thanks to smart algorithms that make your social media experience uniquely yours!
Online Adventure Guides:
When you shop online, there's a virtual explorer helping you find the coolest stuff! AI magic makes shopping websites recommend things you might love based on what you've looked at before. It's like having a treasure map to find your next favorite thing!
Gaming Sidekicks:
In video games, the characters you play against are like AI superheroes. They learn from you, adapt to your skills, and make the game super exciting! It's like playing with virtual friends who always keep you on your toes.
Traffic Wizards:
Picture this: you're going on a road trip, and your GPS app is like a superhero guide. It knows where traffic jams are happening and helps you avoid them, saving the day! It's like having a magical navigator who turns your journey into a smooth adventure.
Healthcare Heroes:
In the world of medicine, AI is like a superhero doctor. It helps analyze X-rays and scans, making sure doctors can spot potential health issues early. It's like having a superhero sidekick in the fight for good health!
Smart Home Squad:
Your home can be like a futuristic castle! Lights that know when you want them on, thermostats that keep you comfy, and security cameras that watch over your castle – all thanks to AI magic. It's like living in a home where everything listens to your wishes!
Language Wizards:
Speaking of magic, language translation apps are like wizards who help you talk to people from all around the world. They instantly turn words from one language into another, breaking down language barriers and making you an international adventurer!
Educational Explorers:
Imagine having a magical tutor who knows exactly what you need to learn and how to make it fun! AI in educational apps is like having a personalized learning adventure designed just for you.
Entertainment Enchanters:
When you watch your favorite movies or shows, special effects and animations are like spells cast by AI wizards. They turn the screen into a magical world full of excitement and wonder!
In our everyday lives, AI adds a touch of magic, turning ordinary moments into extraordinary adventures. It's like having a team of digital superheroes and wizards making our world more fun, exciting, and tailored just for us! 🌟✨
This response was my favorite of them because it takes things that are in a students every day life, and can connect them to a wonder or imaginative space for the student. I believe that giving students something concrete at the end to work with, along with exciting their imagination and sparking creativity, will help to maintain engagement.