Registration is now open! Secure your seat today to Splash 2024.... Register Today Before Seats Sell Out!
Skip to main content
Additional Links

January 3, 2024 / AI Blog

Teacher’s Guide to AI: The Dos and Don’ts of AI in ECE

About this guide

In this multi-blog series, we will help you understand better what AI is and how it works, guide you on how to use it, outline the gaps and pitfalls, and provide practical examples to illustrate it all in the context of Early Childhood Education (ECE). While AI has great potential opportunities and applications in the classroom, it is important to understand that it alone cannot provide wholly sound and reliable materials for instruction, and teachers still need pedagogically designed curriculum and content vetted by Subject Matter Experts for effective education.

This guide was entirely written, reviewed, and vetted by a team of Subject Matter Experts in the fields of Technology, Data Science and AI, and Early Childhood Education. ChatGPT was used for assistance in idea generation, research, and generating examples.

Table of contents:


What is AI and how do Generative AIs like chatGPT work?

For some educators, Artificial Intelligence (AI) can sound intimidating or futuristic. However, it’s already being used in small but meaningful ways in classrooms around the world. At its core, AI simply refers to computer systems that are designed to perform tasks we have historically thought of as requiring human intelligence and decision making. The key is that it’s a tool – one that teachers can control and direct to fit educational goals, but one that needs to be used properly and responsibly!

Generative AI specifically is a type of artificial intelligence that can create new data or content, such as text, images, or even music, by learning patterns and generating output based on what it has been trained on. It essentially “creates” things based on its understanding of existing examples. Ones you may have heard of include OpenAI’s ChatGPT and Google’s Bard. In their cases, based on a very large library of previously created text (referred to as its “training data”), they can make relationships between words to create new text in response to prompts – be it a grocery list based on a meal list, or a lesson plan based on a specific standard.

It is like building a puzzle, one piece at a time, finding the next logical piece based on the patterns it has learned and can recognize. However, it doesn’t have real-time information or a deep understanding of the world beyond its training, so it can sometimes produce incorrect, out-of-context, or just made-up responses.

If you’ve never used an AI tool before and would like to try it, the most popular one and easiest to get started with is ChatGPT, which is free to use for the base level. If needed, here is a beginner’s guide to get you set up!
How can AI help education providers?

We see immense potential for AI in the early education space. Learning from peers in the Higher Education and K12 contexts, we also know that using AI in early education is a unique context. In higher grades, the debate usually centers around student created work and how to counteract the writing abilities of AI. In early education, we see more potential for helping educators use AI tools to make their jobs easier and to focus on the core of early education – the relationship with the littles!

Teachers can leverage AI to help adapt or personalize their lessons, jump start professional development, gather research,  get ideas for activities, help save time on day-to-day tasks, and much more! We will explore examples in Part 4 of this guide, but first…

Using AI: How to write good prompts

Writing Prompts are instructions so your AI tool can create the kind of content you want it to, and writing good prompts is key to generating higher quality responses. Improper prompting can leave too much room for the AI to interpret and fill in gaps in unintended ways, introduce biases, generate less useful responses, etc. We like using the WISER framework. © Allie Miller

Once you get used to prompting, it will start feeling pretty natural! Then you can start adjusting this to fit your needs and workflow according to what works better in the given context.

“Gotchas” of pure tech/AI tools to watch out for

AI can write great bedtime stories, organize and analyze data, and even create convincing lesson plans. But if not trained specifically for early education children and teachers, the output can often be misleading, incomplete, or wrong. We have seen stories on the internet where people prompt AI programs to be very biased, or if the AI program doesn’t have all the data, it will sometimes ‘hallucinate’ by filling in gaps of what it thinks should be there. Some of us may remember the Jurassic Park movies, and how scientists filled in Dinosaur DNA gaps with frog DNA and how well that turned out! To quote Jeff Goldblum’s character: “Scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should”…yes, that’s just a joke, but have you seen this recent story about Microsoft, or the concerns around Amazon Q? Even with the best intentions and most well-trained systems, the AI will get things wrong because it is missing the context, nuance, and the understanding a human expert would have.

There is an opinion in the EdTech industry that AI can help “equalize” education, as it can help offset the current high turnover and can give more inexperienced teachers the tools and guidance they need to bridge the gap compared to more experienced and well-trained teachers. While AI can help bring efficiencies and empower teachers in their day-to-day in amazing ways, AI alone can cause more harm in that area, widening the gap instead by not providing them with the proper information, understanding, and ability to impact sustained learning. Ultimately, AI alone cannot replace quality and efficacious curriculum that is designed and vetted to pedagogically teach and support student learning for long term success…and certainly never replace a great teacher! 

This is all very new, but don’t worry, the industry is working on solutions and guidelines, including ethical and safety considerations! Recently the White House issued an Executive Order on AI to create protections over personal data for AI tools, and the US Dept. of Education’s Office of Educational Technology released a guide on how to integrate AI into classrooms in early 2023. 

 

This is why we believe having humans in the loop both training and verifying the output with subject matter experts is essential, and usage of such technologies need to be treated not as an opportunity, but a responsibility. This can be done in a variety of ways – including standards and sub-standards in training data, using appropriate curriculum as example output, and having experts be integrated in verifying and approving content before it makes it to a teacher’s classroom and influences students.  

 Key considerations when engaging with AI 

As responsible educators, some key considerations include: 

    • Carefully evaluate any AI tools for safety, privacy, and age-appropriateness before adopting them. 
    • Using AI thoughtfully as a complement to, not a replacement for, skilled human teachers. The teacher must remain the heart of the classroom. 
    • Looking for bias in data/algorithms and ensuring AI does not inherit harmful stereotypes. 
    • Giving transparency into how AI applications work so teachers and students can develop informed views.

Does that sound intimidating? We get it, but don’t get discouraged! We came up with a simple acronym to help you think about and validate these things: VALID (get it? J). Consider this a starter set of guiding questions to ask yourself when using AI tools and their output: 

The bottom line… 

The bottom line is AI gives educators some amazing new capabilities, but it cannot replace vetted curriculum or quality educators. Like any technology, it requires watchful eyes and wise use. Teachers should explore AI with cautious optimism: Embracing the potential benefits while also a maintaining a sense of responsibility and carefulness. Together, educators can shape AI applications to truly serve teachers and students, and view AI as an empowerment tool to enhance the capabilities of Subject Matter Experts, not a replacement. Conversations should continue about how to thoughtfully integrate this technology into the classroom while keeping human experts and educators firmly in the loop.  

Apply what you learned!  

Now that you have the tools and understand better how to use them, how about some practice? We created some simple examples where AI tools could be implemented and applied the WISER© and VALID frameworks introduced here to generate and evaluate the output. We highly recommend you check them out: