Editor’s note: This essay is an entry in Fordham’s 2023 Wonkathon, which asked contributors to answer this question: “How can we harness the power but mitigate the risks of artificial intelligence in our schools?” Learn more.
When we started thinking about the question of how we can harness AI to benefit K–12 teachers and students, we asked AI, ChatGPT specifically, and got the following response:
Artificial intelligence (AI) can significantly enhance K–12 education by offering personalized learning experiences, automating administrative tasks, providing intelligent tutoring, and fostering engagement.
The chatbot then detailed specific ways in which AI can help to offer all those benefits within an educational context. The reality of AI in education is that it can supply all those amazing supports, and there are already many examples of personalized learning, automated tasks, intelligent tutoring and chatbot tools, and opportunities for increased engagement.
In order to narrow the scope of our exploration of the impacts of AI in the classroom, we will focus specifically on science education because of its increasing importance and relevance to our current world and the ways in which AI is able to integrate into a science curriculum not only as a tool, but also as a subject of deeper inquiry and study.
When we asked AI what AI + science equals, we got this insightful response from ChatGPT:
AI + Science = Innovation and Discovery
The combination of AI and science leads to innovative solutions, accelerates discoveries, and enhances our understanding of the world.
In today’s science classrooms, AI can provide real and innovative solutions to enhance student learning. AI can allow students to access key models and phenomena through new and diverse lenses. AI, in partnership with science educators, can be a tool that helps to make science relevant, engaging, and accessible to all students, regardless of their personal background, support needs in the classroom, or other relevant characteristics to their learning.
Some uses of AI within science education that have had positive impacts in the classroom include:
- AI-driven assessments: Assessment is an integral part of any science curriculum. To be able to understand and track student progress, there need to be assessments throughout the year. A report on the future of testing from the Center for American Progress details ways that assessment might be enhanced by AI with “intelligent tutoring, stealth assessments, games, and virtual reality.” They assert that “mini-tests built by artificial intelligence can provide a wide variety of ways to use this technology to build engaging tools.”
- Relevant examples: AI can be used to tailor examples and phenomena in the science classroom to specific student contexts and experiences to make the curriculum more culturally relevant and engaging to students. Using examples that reflect students’ lived experiences can help them to connect on a deeper level with their science learning. For example, if you ask ChatGPT to provide an example of a science phenomenon based on energy that is relevant to a middle school student in Louisiana, it will provide details and examples of solar energy. This is in large contrast to a student example in Seattle, Washington, where an example might be closely related to water because of the abundance of water resources in that area. Also, with ChatGPT, you can provide more specifics to the student and school culture, NGSS dimensions, or other characteristics related to phenomena that would help to pinpoint the closest relevant examples.
- AI grading tools: Tools like gradescope can support teachers in their grading, saving them time and providing insights on how their students are progressing. AI also has the capability to help address bias in grading by ensuring that students are graded by objective standards and that teachers’ implicit biases are not reflected within student grades.
- The MSU Denver Virtual Lab Assistant: Otherwise known as the VLA, this tool supports science learning for students with visual impairments and allows those students “the ability to control laboratory equipment in a completely independent manner via voice command.”
- Benetech: A non-profit working to develop an AI solution to create accessible equations and charts for usage with screen readers and within their audio content library, Bookshare. This increases access for students with dyslexia/dysgraphia, students with low vision, or students with other barriers to accessing the materials. Benetech envisions a science classroom where “readers will not have to take extra steps to access the complex visuals as they would already be converted into digital materials.”
- AI tutors: Emerging AI tutor solutions like Khanmingo provide tutoring support for both teachers and students alike, providing coaching through both lesson planning and student learning. Students in computer science courses can get real time feedback on their work as they code to help them identify where they might need to go back to and re-learn specific things to apply to their assignments.
With the development of tools like the ones above, we can envision a future where AI solves many of the access and equity issues plaguing science education and education more broadly. These innovative tools and uses of AI also show us areas where we can further develop AI to fill more gaps in the science classroom. In science, with the advent of innovative assessments and tasks, we are seeing that many schools and districts struggle to know and understand high-quality aligned Next Generation Science Standard (NGSS) assessments. As an organization, in partnership with AI, this provides us an opportunity to creatively problem solve for these instances and provide the knowledge necessary to support the integration and analysis of these assessments.
Also, as we consider what students are learning within the NGSS, specifically in science practices, we can use AI to help teachers expand their curriculum. For example, in MS Practice 8, it says, “[students will] gather, read, and synthesize information from multiple appropriate sources and assess the credibility, accuracy, and possible bias of each publication and methods used, and describe how they are supported or not supported by evidence.” This practice provides an opportunity to use AI as an on ramp to meeting these expectations, but also expands the opportunities for students to practice.
AI can unfortunately also present challenges for students and teachers if used incorrectly in a classroom setting. Right below the area where you input your prompts in ChatGPT is an important warning about the information that is provided: “ChatGPT can make mistakes. Verify important information.” This highlights the potential for AI to spread misinformation and bias, and even beyond that, there are additional risks associated with AI being used in ways that undermine students’ creativity and critical thinking by completing tasks for them instead of allowing them to reason through them.
The way we can begin to truly integrate AI into the classroom and mitigate the risks associated with AI is by forming a strong partnership between AI and educators. Educators and education leaders should be given the training and support to understand the benefits and potential uses of AI to enhance student learning and be given opportunities to work alongside content experts to review content generated by AI. There are several resources available to empower school leaders and educators to leverage AI in the classroom, from organizations like Code.org, OpenAI, and other leading technology and education organizations.
The use of AI within science education just makes sense. AI, as an innovation within the field of science, shows students the ways in which science has already transformed our world and the ways we think, work, learn, and do most other things. With the right guidance and collaboration from educators and students, AI can revolutionize the way we learn in a science classroom. AI can bridge long-existing learning gaps by providing innovative accessibility supports, engaging and culturally relevant examples within assessments and activities, and supporting teacher practice with enhanced grading and feedback supports. We must embrace AI in the classroom to meet the needs of this new generation of science learners.