This is the third article in a four-part series exploring how artificial intelligence can be used to strengthen the transferable life skills often referred to as the 4 Cs: Communication, Collaboration, Critical Thinking, and Creativity. This article will focus on critical thinking.
Outside of cheating, one of the biggest concerns that teachers commonly have about artificial intelligence and generative AI is around those tools hurting the development of critical thinking skills. The reasoning follows that if AI can write essays and generate complex ideas in seconds with a simple prompt, students won’t have to think anymore; AI will do all the hard work for them.
While it’s true that students are finding ways to use AI to help with their work, it’s not necessarily true that they are no longer thinking critically. In fact, critical thinking will likely become more important than ever in the age of AI. With thoughtful learning architecture, educators can encourage students to use AI in ways that sharpen their critical thinking skills, not dull them.
In his recent Forbes article, Ron Carucci supports this point of view, writing, “In reality, critical thinking becomes even more necessary in the age of AI, both to use it properly, and to do the necessary work behind the scenes to make it a more reliable tool.”
This is not the first time that new technology has threatened to disrupt K–12 education. In fact, the release of generative AI tools has been compared to the introduction of search engines. When Google was launched in the late ’90s, some people feared it would stifle student learning and problem-solving by making information retrieval too easy. They argued that such easy access would eliminate the productive struggle needed to problem-solve and think critically during the research process.
While people were provided with significantly improved access to information, search engines didn’t end critical thinking—they just shifted it. Instead of spending valuable resources like time and energy on finding content, learners can now shift their focus to processing and thinking about that information. In many ways, generative AI feels like another Google moment. While it is forcing us to rethink how we do things and how we spend our time, it is not necessarily stopping us from thinking critically.
There are two types of strategies that you can use in your classroom to sharpen critical thinking skills through the purposeful use of AI. By implementing these suggestions, you can design learning that challenges students to use their critical thinking skills at high levels.
Student Use of Generative AI
The first set of examples puts students in the driver’s seat, challenging them to think critically while using generative AI tools like ChatGPT. By interacting thoughtfully and meaningfully with an AI chatbot, students can practice higher-order thinking and become better critical thinkers. Here are a few examples of how that can happen:
The first step to using any AI chatbot is entering a quality prompt or question. To do this, students must think critically. They need to determine what information they are seeking, and then think through potential key words and questions that will prompt the chatbot to return the best version of the information they want.
To take this to the next level, you can explicitly teach students how to generate a quality prompt, a process called prompt engineering. Once students understand the key ingredients used to construct a strong prompt, they will be able to practice an even greater degree of critical thinking, as their questions and prompts will now be evaluated against a set of criteria used to engineer prompts. This will not only get students thinking more critically but will also make them better at writing prompts and using AI in general—a win-win scenario.
Once students click enter and the chatbot begins typing out its response, another layer of critical thinking begins. Now, students must critically examine the responses received from the chatbot. They should be coached to evaluate the logic of the response, the quality of the content, and its accuracy. An expectation should also be set that they look for any hallucinations, or made-up information, and identify bias in the responses.
While students can do some of this based on what they already know about the topic, most students will need to conduct some additional research outside of the chatbot in order to corroborate the response provided by the AI. This type of vetting again requires a high degree of critical thinking, especially when students are asked to explain the process that they used to vet the AI-generated content.
This critical thinking exercise builds upon the results of the first two activities. Because prompt engineering is both a science and an art, students will need to iterate and refine their prompts to get better results. There is no single right answer. To do this well, students must reflect on the prompts they entered and the corresponding responses they received. Based on how they rate the quality of the results, they should construct meaningful and targeted follow-up prompts. In essence, they must refine the critical thinking process that they used when constructing their first prompt. The instant feedback they get during this process can guide them toward improvement and also be rewarding and motivating.
If students are working in groups, the challenge will be even greater, as they will need to reconcile differences among group members and synthesize their ideas into one or more follow-up questions. This type of synthesis requires a high level of critical thinking. Once again, it can be powerful to have students reflect on their process and share what they have discovered along the way. This is both good metacognitive and critical thinking practice.
In Embracing the Future of Artificial Intelligence in the Classroom, Yoshija Walter points out that, in addition to going through the process of using AI tools, students should be challenged to think more broadly about artificial intelligence in general. Walter suggests that thinking critically about AI helps students develop “skills to critically assess AI news, research, and claims” as well as “understanding how to differentiate between AI hype and reality.”
This type of higher-level thinking will help students become more informed digital citizens, and it will also improve their interactions with AI chatbots because they will have a deeper understanding of how AI chatbots work, including their strengths and limitations. This type of high-level thought exploration can be facilitated through various activities, such as class discussions, debates, and jigsaws.
With these first four strategies, it’s ideal if students can experience the use of generative AI tools first-hand. If you work with younger students, however, this is likely not an option. In those cases, the experience will need to be adjusted to be age appropriate. One option is to have the teacher model or guide the experience for the whole class. Either directly or indirectly, students should be encouraged to apply critical thinking during an AI experience.
Designing With Generative AI
You can also use generative AI to design activities that prompt students to think critically. While students might use AI during these activities, the key driver is the nature of the activity itself. The activity should be designed to challenge students to think critically. Here are five types of critical thinking activities that you can generate with the help of AI:
This activity requires students to generate their own ideas without the help of AI. Then, an AI chatbot is asked the same questions, and students are asked to compare their student-generated ideas with AI-generated responses. During that comparison, students should be asked to evaluate the results and determine which is best. They should defend their decisions with sound reasons. If there is disagreement among students, you might choose to facilitate a short debate or discussion. This type of follow-up activity extends the critical thinking process that students began with during their initial evaluation.
In this instance, you would pose a real-world problem to your students. Ideally, it has a local slant to it so that the AI has limited knowledge about the specifics. Then, students can be empowered to use AI to help them come up with solutions to those problems. When they are finished, students should be asked to explain how they used AI to help them come up with their solutions.
Ethan Mollick, a professor at the Wharton School of the University of Pennsylvania, takes this one step further and asks his students to do “an impossible thing” using AI. Not only do they need to use AI to solve a problem or complete a challenge, but they also need to stretch themselves further than they could without the tool. For example, students might be asked to code something, even if they don’t know how to code.
This is similar to real-world problem-solving but also different in at least one key way. It’s similar in that teachers are using AI to help generate the task for the student. However, it’s different because students are not solving real problems. Instead, they are engaging in “realistic” scenarios that have been created to align with the curriculum.
These scenarios can be used to prompt students to apply knowledge learned in class to solve problems, even if those problems are hypothetical. This method encourages critical thinking by requiring students to analyze situations, consider various outcomes, make informed decisions, and apply content knowledge.
AI can be used in a few different ways in this context. First of all, it can generate potential debate topics for the students to consider. Second, students can come up with arguments and evidence to support their side of the debate, and then submit them to a chatbot, asking the AI to come up with flaws or gaps in their reasoning or to respond with counterarguments to consider and prepare for in detail. This process encourages students to critically evaluate the validity of each argument and develop their own reasoned opinions. In this way, AI can be used to help push student thinking to the next level.
ChatGPT can offer suggestions for improving writing, from structure and coherence to support and strength of argument. On one hand, you might think this means that AI is doing the critical thinking for the students. However, this approach is not much different than having the teacher provide the feedback, and that feedback is often what prompts students to reconsider weaknesses in a first draft and work to improve it.
Essentially, by having the chatbot help provide feedback to the student, teachers can clone themselves. Using an AI chatbot as a feedback assistant allows students to get more frequent and timely input about their writing. The teacher should still check in personally with students, but the chatbot can fill the gaps, allowing students to move ahead while the teacher makes the rounds. As with any feedback loop, the input that students receive from either a chatbot or a teacher will require introspection and decision-making, which are both important aspects of critical thinking.
When used well, generative AI tools like ChatGPT can enhance, rather than hinder, the development of students’ critical thinking skills. As is so often the case, it is less about what tool is used and more about how it’s used.
AVID Connections
This resource connects with the following components of the AVID College and Career Readiness Framework:
- Instruction
- Rigorous Academic Preparedness
- Opportunity Knowledge
- Student Agency
- Break Down Barriers
Extend Your Learning
- Critical Thinking and Ethics in the Age of Generative AI in Education (USC Center for Generative AI and Society)
- Using AI to Teach Critical Thinking (Lucy Noble via LinkedIn)
- Enhancing Learners’ Critical Thinking Skills With AI-Assisted Technology (Cambridge University Press and Assessment)