To ban or not to ban? Looking at how AI can be used in classrooms from the perspective of a student and a teacher
If you are a teacher, you have heard the term “AI” thrown around like everyone was let in on the biggest secret in technology. For some educational professionals, the innovation of generative models and artificial intelligence (AI) like ChatGPT is the dawn of a new age in education, where we finally get rid of the paper and pencils for a fully futuristic classroom. For others, these models do nothing but hurt education. With more and more teachers banning the use of AI in their classrooms, others are left with an interesting choice: should I ban it or not?
The term “AI” has become the new buzzword across every industry. The clear frontrunner of the artificial intelligence (AI) industry has been OpenAI, the company behind ChatGPT, but there have been a number of competitors and corporations that have been throwing their hats into the ring over the past year. Google and Meta (formerly known as Facebook) have released their own models, Gemini and Llama 3, respectively. Anthropic released their model Claude amidst the craze, which some experts think is a better model than ChatGPT. With all these names being thrown around, it can be hard to know what exactly AI is and what it is not, especially if you are getting all this information while teaching classes.
Artificial intelligence (AI) works through pattern recognition of different data, usually words. It is made to replicate human intelligence and the way that we recognize patterns and find solutions to problems. Developers use a vast amount of data to train these models. The more data the model is trained on, the better its pattern recognition will be. Data is measured in tokens; one token is around 3–4 words. For perspective, Meta’s biggest model Llama 3 was trained on over 15 trillion tokens worth of data.
There is a lot of debate going on today about AI and the future of our world. Most familiar with the industry agree that every industry will be affected by the new tech, and education is not immune. Most teachers know AI because a large group of students across the United States began using generative AI like ChatGPT and Snapchat’s AI feature to write their essay assignments for them. Once teachers caught on, students began putting the generated essays into rewriting software to make the text sound more “human”.
From the student perspective, it is very tempting to put your prompt into a text box and hit “send”, knowing that a week’s worth of writing will be done in 10 seconds. This frees up weekends and time after school to do whatever you want. For teachers, it is incredibly frustrating and disheartening to see students submit work generated by AI. Grading sessions are rife with paranoia and distrust as you try to discern what is real writing and what is not. So what are teachers to do?
The first thing we must understand as educational professionals is that AI is a tool rather than as a conscious being. Artificial intelligence, as of right now, is not what we all saw when we sat in the theater for the release of The Matrix; it is a tool that recognizes patterns and generates what it thinks the next response should be. It has an easy time generating easy answers, and it cannot possibly know anything outside of its dataset. With that in mind, here are a few techniques and strategies to use in your classroom:
The first strategy is to have the students write their essays in class, in person. This can not only act as a safeguard against the use of AI to write, but it can also be a great formative assessment to see where the students get stuck in the writing process. This can be time consuming, though, and can take valuable time away from lectures or activities that can only be done in person.
The second strategy is to change up your prompts. It might be time to rethink the simply, tried-and-true prompts that you have been using for years. Prompts like “Examine how the Civil War began in the United States” are a cake walk for generative AI programs like ChatGPT. You could change your prompts to include images and videos as sources to analyze like the image below from The Canadian Letters & Images Project, showing a family during World War 2. Keep in mind technologies like DALL-E are being developed that can analyze and generate images and videos. Another idea is to make the students use specifics from class that a generative AI does not have in its dataset. The goal is to make your prompts as unique and thought-provoking as possible so only the student can write a meaningful essay about it.
The third strategy is the most brave one, and that is to have a serious conversation with your students about generative AI. As a student, it is incredibly tempting to use a new technology when it comes out, and it is even more tempting when you are told you aren’t allowed to do so. One of the best thing that we can do as teachers is in the name of our profession: educate our students about AI. What it can and can’t do, what it was designed for, and what it generates every time you put in a prompt and press “send”. This strategy elicits that most out of all, though, because it asks the students to trust the teacher that they know what they are talking about and the teacher to trust the students that they won’t use the programs.
Whatever way you choose, it should come from a place of empathy and wisdom. You can educate yourself on how AI works using free courses like the ones offered by Elements of AI, or if you want to go all in on machine learning, the free courses from Hugging Face. You can also use the resources from the project I started, Teaching for Tomorrow, which can provide a good starting point for incorporating AI into your classroom if you choose to do so.
