A New York school district’s board recently approved about $60,000 to purchase three AI products: an AI-powered, human-like robot, a virtual AI-powered teacher’s assistant and an at-home tutoring program. While the school’s superintendent believes in classroom innovation, parental apprehension resulted in the AI products being put on hold. A San Diego charter school network spent $500,000 to purchase two ChatGPT-powered, human-like robots as a way to study how artificial intelligence can support teaching and enhance future classroom learning.
Let’s take a look back and see how we arrived at this place to better understand the future of AI and schooling.
AI: Beginnings and the future
The confusing term artificial intelligence was first coined in 1956. Here is an explanation.
AI processes massive amounts of information, such as text, images, video and sound, in response to human prompts or questions, then uses algorithms or mathematical procedures to identify patterns and differences. AI does not think. Rather, it learns from existing databases and online information, then predicts likely responses, word by word, to answer questions, offer suggestions, or generate original content.
Interestingly, a 1921 science-fiction play, “Rossum’s Universal Robot,” imagined robots powered by something like today’s artificial intelligence. A mechanical device similar to what we know today as a computer was likely created in the 18th century, with the first programmable digital computer invented in 1941 Germany. The necessary component to launching AI would now become available.
Alan Turning, known as the “Father of Computer Science” greatly impacted the AI story in the 1950s. He was one of the first modern-day scientists to explore the idea of artificial intelligence, through work with theoretical mathematics. Early AI programs were developed over the next decades.
During the 1980s, with the availability of advanced computing power and larger data capacities, AI began solving real-world problems. The 1990s-2010s saw AI become an ever-powerful tool by generating or creating content. For example, Apple introduced “Siri” in 2011, a much-enjoyed voice-activated assistant used by more than 86 million Americans. Today, AI is integrated across various fields, like medicine and education, and throughout much of our daily lives. Open AI’s, ChatGPT (released in 2022) is used by 34% of all adults in the United States to generate text, images, sound and video.
AI is becoming more pervasive in the 21st century. Although this technology is readily used, concerns are being voiced. These concerns focus on ethical considerations (like who owns AI-generated art), amounts of electricity and water required for operation, privacy, spreading false information, workforce displacement, enabled weaponry and cyber security risks. Guardrails are needed so AI will operate safely, legally, ethically, appropriately and responsibly.
We can only wonder about the future of AI and what its effects will be on human life: for example, our cultural understandings, human interactions, solving modern-day challenges, to just name a few. Today’s AI researchers are worried. Because AI is progressing so rapidly, it could “spiral out of human control— a worrying science-fiction scenario that the industry had long-warned would become a reality,” according to a recent article. Then what?
Given what we know (and don’t know) about artificial intelligence, it’s critical for us to develop a thorough understanding of this technology. But according to several surveys, this is not happening. Many Americans are aware of the ways AI is part of their daily lives, such as wearing fitness trackers, yet over two thirds of us do not realize that we are actually using AI.
Across the globe, many people trust their own country to keep them free from harm in regard to AI. Americans also believe their government and business leaders are partly or mostly responsible for their AI safety. Today, an ever-increasing number of Americans have more negative than positive views about AI’s impact on society. These concerns are certainly not being addressed by the president.
At present, he advocates for federal and state-level de-regulation that would, for example, preempt AI safety and anti-bias laws; wants only voluntary industry rules and promotes fast-tracking federal permits to build larger data centers— eliminating all hurdles to the private sector’s push to advance profits. I believe AI is built from our collective knowledge and should serve all of us, because it belongs to all of us. His dismantling of the Department of Education will only make managing AI school-use increasingly difficult in the future.
AI and K-12 schooling
K-12 schools play an important role in our society, so we must address issues associated with AI school use. These include issues such as loss of critical thinking, analytical skills and creativity; safety; inaccurate and biased information; reduced human interactions; and inappropriate/irresponsible use (like submitting AI-generated work as your own work). Students are concerned too! Many of these issues were recently part of The Students First Act, a high school student-generated document proposing a national AI policy that is being shared with over 10,000 school systems.
The other side of the coin is just as important. How do teachers use AI so that more students can think deeply about their learning and ask more insightful questions. How can students learn more and want to learn more? The good news is that, according to a national three-year survey of 1,058 districts, these districts are not rushing or resisting. Rather, they are taking an intentional approach to AI use.
Recently, Levi Bachmeier, North Dakota Superintendent of Public Instruction, put forth a planned and purposeful approach. He advocated for a balance between student safety (“our first priority”) and appropriate AI classroom use. He wants to “ensure that students receive training before they are allowed to utilize AI in a classroom setting” and provides teachers with lesson plans and quiz generators for ethical AI use.
I agree with former President John F. Kennedy who wisely said, “Children are the world’s most valuable resource [and asset] and its best hope for the future.” Because of the ramifications associated with artificial intelligence and students, their families and their teachers, we must act.
Major technology companies are selling AI products to schools, like generative lesson plans, differentiated reading levels and written student feedback. The Ed-tech Group of Companies, including Brisk Teaching, MagicSchool AI, and School AI, entered the teacher-generative AI market a few years ago. Not to be outdone, Open AIs ChatGPT for Teachers and Google’s Gemini and Microsoft’s Copilot AI Project’s for Schools are now competing with Anthropic’s AI for K-12 teachers. AI in schools is big business!
Currently, annual revenues total $700-$900 million and are expected to grow to $15 billion by the 2030s. Michael Marinaccio, The Executive Director of the Center for Responsible Technology warned North Dakota legislators that big tech will experiment on the state’s children if state lawmakers don’t act.
About 70% of America’s teachers use AI products to help make lesson plans and create quizzes/tests (the top two uses). Three out of four teachers believe the implications of AI school use is far greater than computer and internet school use. A major AI worry voiced by teachers and administrators is that without safe, ethical, responsible and appropriate use, students will experience superficial engagement with their learning — and as a result, not learn to think for themselves. Interestingly, survey results point to students also being concerned about this very same thing. Paul Krugman, Distinguished Professor of Economics, captured this concern, when he wrote: “Students who rely on AI to answer questions won’t learn to think by reasoning through evidence to form a conclusion.”
But, what if teachers knew how to use AI in ways that helped students reason through evidence and form conclusions? This would be a great example of responsible and appropriate use. We can’t stop students from using AI. We must find ways for students to use AI to become bigger and better thinkers, like AI asking probing questions about a student’s essay or arguing against the essay’s perspective.
About half of the states (North Dakota, no, Minnesota, yes) have enacted laws to require Media Literacy Education (which includes AI literacy), defined as using, understanding and thinking critically about how AI is used in schools. This legislation is important as currently there is no federal education policy to support this kind of important learning. Because President Trump’s perspectives on AI and school use focus on global economic and security interests, there won’t soon be school-based policy support from the federal government. This lack of guidance is compounded by the fact that most districts do not provide professional development for staff.
Interestingly, Moorhead Area Schools have partnered with Minnesota State University Moorhead’s Institute of Applied Artificial Intelligence to provide professional development in AI literacy, teaching improvement plans and strategies to make AI assignments ethical and responsible. A Master of Science degree in Artificial Intelligence is set to be offered in 2027.
Fargo Public Schools (FPS) is an example of a school district providing AI guidance for its students and teachers. Fargo Public Schools supports teachers with artificial intelligence implementation by providing dedicated resources, staff guidance and participation in specialized regional training workshops like the North Dakota AI Innovators in Education.
A variety of guiding principles for K-12 teachers and students to use when putting AI into practice, can be found at the Fargo Public School’s AI in Education website. A summary follows.
Principles are based on the idea that AI use cannot be avoided and that students should be taught to become AI literate users. Both teachers and students are cautioned about inaccurate and biased information, privacy concerns, as well as responsible and ethical AI use. FPS adopted ideas from a recent “Ed Week” publication, titled “What is Age-Appropriate Use of AI? 4 Developmental Stages to Know About,” for helping teachers with big picture decision making about what is critical for students, at specific grade levels, to learn about AI. This kind of support is necessary in order for AI classroom-use to be safe, responsible, appropriate and ethical, leading to more informed learning.
For children in grades K-1-2, teachers are warned not to use language that makes AI seem human, because this group might believe that AI is human because of how the technology operates. Teachers of children in third, fourth and fifth grades need to make sure that students’ continually practice problem solving and critical thinking on their own, even when AI can readily provide answers to their questions. The struggle of thinking to learn is crucial.
Middle School students in sixth, seventh and eight grades should be able to use generative AI to complete their schoolwork, like prompting AI to outline a student’s essay after she completes the research. Evaluating AI’s responses for mistakes is also important. Teachers should be aware of age limitations associated with AI products, including when parental or guardian consent is needed.
A key outcome for High School (grades 9-12) students is developing AI literacy. Literacy lessons should help students understand how AI works and the complications it creates. The goal is for students to become critical consumers of AI products, so they know when generated information is helpful and useful and when it is biased, stereotypical, inaccurate, misleading or incomplete — and should be discarded.
Various AI school-based products are included on the website to help teachers use these products appropriately to accomplish specific tasks (like using SchoolAI to generate lesson resources) in a thoughtful manner. FPS clearly communicates to students what appropriate and responsible AI use looks like — idea creation and planning how to use these ideas in assignments, for example.
Also included are examples of prohibited AI use, such as intimidating others, overuse and copying. These two AI uses require holding two truths at the same time. Parallel communication is also directed to teachers to guide students in creating appropriate and responsible AI classroom use. For example, teachers are instructed to have students review with them why they want to use AI — and when reviews are positive, include a statement of how AI tools were best used and how the generated information was blended into the assignment.
Guidance provided by the FPS is an excellent example of how artificial intelligence should be approached by public school districts to answer the following all-important question; How can AI be used safely, responsibly, appropriately and ethically to improve student learning?
