Perspective: AI and the Future of Education
“That it will be of very great use cannot be questioned, but how will its uses add to the happiness of mankind?”
“You give [them] only the semblance of truth They will appear to be omniscient and will generally know nothing. having the show of wisdom without the reality.”
With all the hype around artificial intelligence, you might assume these quotes were written about the power of modern AI. However, the first, penned in a New York Times story about the earliest successful test of a transatlantic telegraph cable, is dated August 19, 1858. As communication times between Europe and the United States dropped from three days to three minutes overnight, the author was aghast at the unimaginable societal horrors surely in store. The second quote depicts Socrates’ classic concern about students’ being consumed by the dangers of the newest 370 BCE technology—the written word.
Unlike the authors of these alarmist proclamations, I am indeed writing here about AI. As we attempt to understand a future in which AI plays a prominent role, concerns about this seemingly mystical technology often occupy center stage. However, the bigger picture embraces anticipation and excitement as much as caution and concern. Where AI and learning intersect, intriguing questions—and more than a few surprises—await.
Checking our assumptions
If history reveals anything with certainty, it is that when ground-breaking new technology appears on the scene, uncertainty is sure to follow. No matter the specific example—AI, telegraphs, bicycles, umbrellas, or forks—new technology and fears about the future are close cousins. According to Genevieve Bell at Intel, moral panic sets in any time a technology changes our relationship to time, space, and people—three categories AI will undoubtedly engulf, if it hasn’t already. Given this, it’s our responsibility to search through the hype and panic to unearth AI’s true promise and to understand which concerns are justified.
In Kevin Roose’s recent exploration Futureproof, he investigates some of the crosscutting assumptions about AI and automation: AI will make things better by doing the boring work for us. Humans and AI won’t compete; they’ll collaborate. We’ll come up with jobs we can’t even imagine today.
His conclusion? The future isn’t as simple as such pithy proclamations imply—and it’s not all rosy either.
According to Roose, AI will usher in a new industrial revolution. Importantly, this one is likely to be different. In particular, it will undoubtedly be a white collar revolution—ChatGPT has already shown the ability to pass complex licensure tests, including the bar exam, with flying colors.
In addition, a view of past such revolutions shows that while they have improved many things, each has also left repetitive jobs and significant inequity in its wake.
Finding the silver lining
Still, there are reasons to be optimistic. As part of his thesis, Roose lays out a set of rules “for humans in the age of automation.” Indeed, these rules will be disruptive for many, as some jobs are clearly in jeopardy. However, Roose’s rules offer consolation to those in education. Roles most impervious to AI takeover, he predicts, are those that change significantly from day to day, hinge extensively on social interaction, and do not lend themselves easily to predictable, data-intensive recipes. That describes every teacher I know.
Nevertheless, while the art of teaching may be safe from the AI bots (for now), there’s still some pretty solid change on the horizon. Let’s look at some of the positive potential, from “adjacent possible” near-present scenarios to some wild future ideas worth envisioning.
Looking to the AI future
Teaching and learning are a complex pastiche. AI and automation have the potential to provide valuable transformation in many places—from serving as highly intelligent resources to providing solutions to currently vexing problems and opening up new synergistic opportunities.
For example, imagine educators had the information and real-time guidance to easily orchestrate extended, open-ended learning experiences while simultaneously providing tailored assistance to individual students. AI and automation could monitor and manage the logistical challenges of coordinating project-based learning, render the intricacies of differentiation and scaffolding for individual students tractable, and enable teachers to understand where each student sits along a desired trajectory of learning or instruction.
We can think more broadly as well. With better information available about learning both in the moment and over time, performance assessment, portfolios, and competency-based learning could become more effective. AI could find paths toward promoting and improving student collaboration and could gauge cross-disciplinary learning to boot.
The possibilities for improving engagement, fueling student-centered approaches, and ensuring relevance to students and society are immense. AI may finally allow us to truly view education as a broader endeavor informed by a diverse set of informational inputs, bringing the promise of equity in learning closer than ever.
Preparing learners for an AI-filled future
Of course, we must also consider students’ post-graduation future. What do learners need to be ready for? After all, the rapid innovation we’re seeing in AI is by no means limited to the classroom. AI is already actively transforming the world our students will enter. What does the future of writing look like in a world where ChatGPT can write robust, detailed essays? What’s the future of design in a world where technology can generate art, three-dimensional structures, and even full videos from whole cloth?
And what thousand things are we not yet imagining? Even now, innovators and artists are using ChatGPT to generate concepts, employing DALL-E to create an image sequence from them, then feeding the sequence into another platform to produce animated visualizations. Truly transformational tools are already difficult to wrap one’s head around. When it’s possible to combine them in multiplicative chains, the resulting potential is truly unimaginable.
When such possibilities arrive seemingly overnight, we can be left feeling adrift. But taking the broad view offers important perspective and useful starting points. While our first instinct may be to wall off students from chatbots, what if we instead take for granted their role as constant companions and begin to examine their potential? Rather than wring our hands about questions of copyright in generative art, what if we put our energy into exploring the green fields of creativity waiting to be unlocked? Just imagine the opportunities if we were to view AI tools through the lens of combination and connection, pouring our unbridled human creativity into the search for unexpected and unexplored synergies.
Generative AI has already helped design buildings and predict protein structures. Why shouldn’t it generate new classroom configurations, diagram novel learning trajectories, or uncover new ways of viewing longstanding problems in engineering, science, or mathematics? Chat-based learning models could give us the seeds for student feedback, generate unexpected lesson plan concepts, or provide novel approaches for differentiating student learning.
While we must indeed proceed with care and with eyes wide open, our roles as scientists, innovators, and explorers compel us to embrace the excitement of the unknown. As we look ahead, we must acknowledge and welcome our students as active participants in shaping the future that the existence of these tools will create. The years ahead promise amazing things. We need to stay open to their potential.
Chad Dorsey (email@example.com) is President and CEO of the Concord Consortium.