By now, you’ve heard about generative AI tools via some of the larger AI offerings that have launched in the last several months. While OpenAI’s ChatGPT1 has taken center stage in the discussion, Bard2 has just launched with big plans for integration across the Google Office suite, and Microsoft plans to incorporate generative AI tools beyond the Bing search engine into Word and Power point. Additionally, Meta3 is launching its own large language model, all contributing to the growing conversation around the ease of which anyone can use AI. While impressive, these generative text-based tools are almost quaint compared to generative image4 and video tools5, which can quickly pull together complete multi-model offerings that can challenge even the most diligent observer of the uncanny valley.
While there has been some backlash and criticism around generative AI tools, there is really no way to minimize AI’s impact on the future of learning. As time goes on these tools will only draw more interest and their use and application in the classroom will require more skills, knowledge, and abilities. Like many other innovations and technological disruptions, language educators are well situated to take advantage of these tools. When used right, generative AI can further unbox the power of high-quality language learning to advance communication skills while building relevant skills in critical thinking, collaboration, and creativity.
There’s a lot to learn about AI. The following five tips will provide some guidance to those who are just starting out with these tools as well as those who are already exploring or adopting generative AI as a resource in their classroom.
Tip 1: Prompting Is a Skill
One of the first things most educators need to understand about generative AI is how prompting works. Prompting is a critical skill and learning how to build prompts can help inspire instructional strategies that use AI. Larry Ferlazzo (Using ChatGPT With ELLs. Edweek)6 provides great examples of building interactive and engaging prompts that will allow you to maximize generative AI in the classroom. The challenge for educators when it comes to building a prompt is understanding that the goal is not to produce something for review, but something for your students to use. A well-constructed generative AI prompt will provide curated and tailored task-based learning content that will engage your students to use all of their digital literacy and communication skills.
When incorporating AI into a lesson, consider how students will prompt and what the results will look like to understand how to get the best learning experiences. AI will produce great volumes of text, and can do so at various CEFR7 levels and syntax complexity. This makes it possible to have students generate their own personalized AI content to work with, which can help improve progress in English while giving them valuable experience working with digital tools.
Another useful way to use generative AI is to build content like reading passages, videos or even songs, for students to work with. Students can then review and revise these AI-generated products based on classroom rubrics for various language skills. For example, in a unit on figurative language, AI can generate a random passage, poem, or lyric using figurative language. Students can then mark up the writing, give feedback to the AI author, and discuss the benefits and weaknesses of what was created. This will not only provide an interesting way to further contextualize the language you are teaching, but help students internalize the language, its use and application, strengthening their own understanding and application of figurative language.
AI Will Lie to You
But wait; “the limitations of generative AI”? What limitations? Many are looking at what generative AI tools can do and immediately seeing how these tools could replace workers and automate tasks.
In reality, machines aren’t taking over the world. You still need people with skills and experiences in specific contextual areas to understand the quality of the output of any AI tool. One of the most significant limitations of AI is its tendency to hallucinate8 or make up information that is entirely incorrect. Alarmingly, many AI programs will confidently maintain that the output is correct, even when it is clearly wrong.
This can work to great advantage for language educators who can construct engaging task-based experiences where learners prompt AI for information on topics relevant to their interests or courses of study, and then collaborate to research responses, fact check, and correct errors. In groups, learners can present their findings, demonstrate where generative tools got it wrong, and share tips to avoid incorporating false information into their work. These types of lessons can also help to curb potential cheating and plagiarism, as AI cannot be trusted to provide correct information, even though it will confidently give you false information all the time.
The current rapid adoption of AI tools will require our learners to be able to work with the information AI believes is real and truthful. The more our learners can understand the constraints of what AI can produce, know how to spot errors, and solve problems in output, the more effective learners will be in using and communicating with teams—who are also incorporating AI tools in research and projects.
Tip 2: AI Is more Than Just ChatGPT
ChatGPT has been taking up a lot of the airspace in the conversation around AI, but it’s certainly not the only tool. Larry Ferlazzo has curated a spectacular list of resources9 for using AI that will benefit any language educator. The right tool can create text, images, or video. Beyond creating, AI tools can modify existing work, turning a paragraph into a bulleted list or enhance photos with razor-sharpness.
The use of AI to generate content helps to solve one of the greatest challenges for language teachers, as it reduces the time to produce different types of engaging content. Suddenly, putting together a quick video, producing a listening prompt, or even creating dynamic interactive tasks for activities are easy to pull together with AI tools. As you bring these resources into your classroom, take time to explore video, music, and image creation tools to inspire activities for your learners.
Tip 3: Task-Based Learning Improves with AI Assistance
Generative AI is a phenomenal resource for those who have struggled integrating task-based learning because of the time necessary to plan solid integrated learning tasks. With a few well-crafted prompts you can have everything you need for a dynamic jigsaw-style lesson, or an engaging project activity to focus learners and drive language to productive use.
Try out this prompt with ChatGPT. Complete the gap with relevant context for your students. Include different CEFR levels and topics in your prompt to explore a variety of lessons.
“As my teaching assistant, I would like you to put together a task-based reading jigsaw lesson at an ____ level on the topic of _____. Provide the lesson plan. Please provide the reading texts I will give to students.”
Generative AI is fantastic for this kind of creation. If you have any outdated task-based materials, AI can make it easy to modernize some of your favorite content. Just prompt your selected AI tool with your specific parameters. Try this prompt: “As my assistant I want you to review these reading text and refresh them for students in 2023.” If you have your lesson plans and content ready to cut and paste, within a few clicks you could easily generate several new versions that you can immediately test and refine with your learners.
Tip 4: AI Makes Differentiation Easy
One of the best things about generative AI is that it quickly and easily enables both differentiated instruction and adaptability in assignments. Rather than constructing text at different levels to support students, you can provide prompts to students to help them adapt the content to a level appropriate for them. As you are updating previous content, you can generate new materials at differentiated levels, immediately expanding the resources you have on hand to engage your students.
Using generative AI can also allow more exposure to idiomatic language and historical information in relevant areas and give learners powerful tools to make more meaningful connections with the content they are working with. AI can also explain idioms, putting language into context and making it easier to understand.
Tip 5: AI Is Well-suited for 21st Century Learning
One of the most important ways we can incorporate generative AI into language learning is in 21st century learning. With high quality 21st century learning, digital literacy and digital tools are built into the project activity that engages the 4Cs, requiring high quality group interaction and collaborating, and engaging critical thinking and creativity. Generative AI opens up so many new ways in which to support this in language learning, while significantly reducing the work to the educator.
However, it does require educators to have a better understanding of prompting and output to navigate how to help students engage well with AI. From there, a number of tasks are possible that can help drive success for learners.
Here is a quick table that shows how you can crosswalk more traditional task-based activities into those that incorporate generative Ai tools. Given that 21st century learning delivered through tasks is one of the best ways to engage students, this opens up a number of opportunities for integrating generative AI successfully into task activities.
| Traditional Activity | Generative AI Enhanced Activity | 21st Century Skills Developed |
| Collaborative Report | Collaborative Report Review with Analysis | Collaboration, Critical Thinking, Communication |
| Summative Jigsaw | Research and Summative Jigsaw | Research, Collaboration, Presentation |
| Information Gap Activity | Information Review Activity | Information Analysis, Problem-Solving |
| Graphic Organizer Activity | Generative Graphic Organizer Activity | Creativity, Critical Thinking, Organization |
Traditional Activities:
- Collaborative Report: Students work together to research a topic and compile their findings into a report.
- Summative Jigsaw: Students are divided into groups to become experts on different topics, then they regroup to share their knowledge and complete a larger task.
- Information Gap Activity: Students receive incomplete information and must collaborate to fill in the missing details.
- Graphic Organizer Activity: Students use visual aids to organize information and concepts.
Generative AI Enhanced Activities:
- Collaborative Report Review with Analysis: Students utilize AI-generated reports, and as a group, they critically analyze and discuss the content, identifying strengths and weaknesses.
- Research and Summative Jigsaw: Students employ AI-powered research tools to gather information and then participate in a jigsaw activity to share and consolidate their knowledge.
- Information Review Activity: Students review AI-generated information sets and assess their accuracy and reliability.
- Generative Graphic Organizer Activity: Students use AI-powered tools to generate graphic organizers based on input data, fostering creativity, and enhancing critical thinking skills.
This is Only the Beginning
Everything we’ve seen so far over the past year has been the tip of the iceberg of technological change that is coming for so many industries—including education. As language educators, it’s critical to incorporate AI tools in our work to help support our students toward success. Our learners need to know how these tools work, how to use them, how to advocate when they are harmed by these tools in various ways, and most importantly, understand how AI works as a resource and collaborator. In this moment, we will play a pivotal role in preparing our students for the future, and successfully equipping them with the skills to navigate and harness the power of AI responsibly and effectively will be the foundation for success through many new technological innovations to come. By integrating AI into our teaching practices, we can foster a dynamic learning environment that embraces technological advancements and prepares our students for the challenges and opportunities ahead.
Links
3/ https://ai.meta.com/blog/large-language-model-llama-meta-ai/
4/ https://openai.com/dall-e-2
7/ https://www.coe.int/en/web/common-european-framework-reference-languages/level-descriptions
8/ https://www.nytimes.com/2023/05/01/business/ai-chatbots-hallucination.html
