February 14, 2023

“What happens to learning when we move from the stable infrastructure of the 20th century to the fluid infrastructure of the 21st century, where technology is constantly creating and responding to change?”

Douglas Thomas and John Seely Brown

In A New Culture of Learning, Thomas and Brown wrote that when people think about learning, they usually think about schools. “And when people think about schools, they usually think about teachers.”  Yet, they say, the kind of learning that will define the 21st century is not taking place in a classroom. “Rather, it is happening all around us, everywhere, and is as powerful.” 

Artificial Intelligence is everywhere around us and it is most certainly powerful. AI has supercharged an automation process that now extends into advanced human tasks. At any given moment, AI is writing the articles we read online, creating sophisticated musical works in the style of iconic composers, learning how we cook and devise recipes, driving commercial trucks, beating us at our favorite board games, dancing like a human, designing new molecular proteins, assisting people with disabilities, creating poetry, and talking to us in a friendly tone. We’d been told for years that AI would disrupt our lives, but the striking speed and wide-reaching impact of AI have caught many technologists, academics, and scientists by surprise.

Add teachers to that group. Educators across the nation are flummoxed by the introduction of ChatGPT, an artificial intelligence assistant developed by OpenAI that can, among other things, write essays, craft poetry, produce song lyrics, write and debug code, explain difficult concepts, identify incorrect premises, and roleplay in imaginary scenarios. The world’s most advanced chatbot, ChatGPT has sparked widespread fear at schools and universities that students will use it to cheat on their homework and be unmotivated to learn.

(And if ChatGPT wasn’t enough, there are AI networks that can produce a combination of text, images, video, or sound from a simple prompt by the user. For instance, DALL-E is an AI assistant that can create original art based on your description of what you want to see and Google’s unreleased MusicLM can create music from text. Moreover, OpenAI, Google, and others are building systems that can generate videos of people and objects from text.)

The angst wrought by ChatGPT has led many schools and districts to ban its use, most notably the New York and Seattle public school systems. Other schools are opting to mitigate its impact by eliminating technology-aided writing. But banning GPT is fruitless and avoiding ChatGPT is a stop-gap solution. GPT technologies will only become more powerful and widespread and students will access them on home networks or on their cell phones. And avoiding AI is counterproductive in a digital society where those who learn to leverage AI effectively will displace those who cannot. 

So, I can only hope that the conversation surrounding ChatGPT is evolving from ban/not ban to a broader discussion of teaching and learning in this “fluid infrastructure” of the 21st century:

How do we educate our children to thrive in an automated and accelerated work environment? 

More broadly, how do we educate children to become productive and creative citizens in a complex, technological landscape?

AI is rapidly changing society, requiring us to rethink the student learning experience.

We need a framework for understanding what artificial intelligence is, what it can and cannot do, and how student learning is situated within that context.

For one, it’s misleading to label AI capabilities the way we do human cognitive processes. AI systems don’t possess a human brain, after all, and their output is derived from statistical calculations embedded in algorithms. AI systems typically engage in a series of extended trial-and-error processes and then perform self-analysis to determine how they can improve. AI manifests itself in ways that imitate human intelligence—reasoning, language use, learning from experience, etc.—but even OpenAI chief executive Sam Altman does not believe AI is intelligent in the way humans are. “It is like an alien form of intelligence,” he said in a recent New York Times article. Nonetheless, alien or not, AI has made tremendous progress in tackling advanced cognitive tasks and, in the process, has wrestled jobs away from humans. 

Despite its advances, we know that AI cannot match humans in two important capacities: adaptability and social-emotional understanding. AI systems cannot handle completely novel situations. They are ultimately limited by the data they possess, however extensive it may be.

So, humans have a distinct advantage over AI. Humans are simply better innovators. We can adapt and thrive in entirely new and novel environments. Our thinking and acting are not boxed in by data inputs; when faced with the complicated nuances and changing dynamics of human problems, we develop creative, human solutions. In a recent article Dr. Colin Allen, a professor at the University of Pittsburgh who explores cognitive skills in animals and machines points out that: 

“A conscious organism — like a person or a dog or other animals — can learn something in one context and learn something else in another context and then put the two things together to do something in a novel context they have never experienced before. This [AI] technology is nowhere close to doing that.”

Furthermore, AI is not sentient; it does not have feelings. AI is not conscious of itself and not aware of the surrounding world. It lacks the social-emotional skills and competencies for engaging and cooperating with others—self-awareness, fairness, and kindness, to name of few. And AI does not understand the notion of fairness, nor the notion of kindness. (It may even carry algorithmic biases that work against fairness and kindness.) AI is getting better at interpreting human gestures, and nonverbal cues, but is not imbued with a sense of what is right and what is wrong. It lacks a moral compass. 

So, ChatGPT reinforces the importance of making schools more human by emphasizing uniquely human skills. It means placing students in situations where they must work in creative and innovative ways. It involves students working collaboratively with others to build supportive relationships featuring empathy and personal responsibility. 

It also means personalizing learning by conversing with students more frequently and integrating student interests, hopes, strengths, and needs into our units and lessons. It means putting the Person before the Product by building a learning process that includes constant interaction and feedback. Ultimately, if we know how our students think and feel we are in a better position to help them craft fresh ideas and new perspectives.

AI can be leveraged as a partner in the creative process. AI can help students break through difficult stages of the creative process in their activities. For those students who don’t know how to begin, ChatGPT can offer suggestions. For those students “stuck” somewhere in the creation process, ChatGPT can generate ideas for their next step. All the while, teachers and students can analyze the usefulness and appropriateness of the information ChatGPT provides.

AI will continue to present challenges and opportunities. The history of education technology is replete with periods of fear and uncertainty over the introduction of certain tools and platforms. But banning or avoiding AI is unrealistic and counterproductive. Instead, we should be adapting in times of change and leveraging our uniquely human abilities to forge a pathway forward.