Is AI an Existential Threat to Teachers?
Will teachers really be made obsolete by AI?
Summary
Artificial Intelligence will undoubtedly change education
It (probably) won’t replace human teachers in the classroom
Teachers need to adapt and change to demonstrate our value beyond the capabilities of AI
The Fear and Hype around AI in Education
“We’re making ourselves obsolete!” One teacher exclaimed in a department meeting about the applications of AI. I’m the last person to ever talk myself out of a job which is exactly what some teachers fear will be the eventual outcome of using AI in the classroom.
To help assuage those fears, let’s start with the most cynical take: a large portion of the US population views teachers as nothing more than babysitters. Unless there are some fundamental changes to legislation, there will always be a need for a certificated educator to be a warm body in the room. As an educator, I’d like to believe that I also add slightly more value to my students’ lives than your everyday babysitter. Besides that, AI is still quite a bit away from coming close to replicating the value of a human mind leading the classroom.
So what could the role of teachers be in a future where AI is ubiquitous? I envision both student and teacher facing AI that provide immediate assistance for the more time consuming aspects of learning environment.
What is possible right now?
The current iterations of AI may not be perfect for the classroom but they can still amplify our ability to educate our students. The current onus for differentiation in the classroom lies entirely on the teacher meaning that it’s up to us to assess and anticipate student needs in order to accommodate different learning styles and abilities. However as any teacher knows, there will always be a stumbling block that we couldn’t possibly anticipate. Right now there are two simple ways AI could streamline these aspects of the classroom.
The first and most obvious use-case in this situation is for teachers to use AI to anticipate roadblocks and brainstorm solutions when planning lessons or assignments.
The second and perhaps more controversial solution here is to have students interact directly with AI. Even with the most powerful tools available, teachers will never be able to anticipate every single question or roadblock a student might face. Why not encourage students to become more independent by using AI as a first step intervention? Right now, a student could easily copy and paste an assignment into ChatGPT and ask it any number of questions. Khan Academy has been working on Khanmigo which is a student-facing chatbot using a modified version of OpenAI’s platform.
Rather than fear that students will become less dependent on teachers or that teachers might even be replaced, why don’t we provide students with the tools to answer preliminary questions without having to wait for the teacher to get to them?
Sometimes students will spend ten minutes waiting for a teacher to answer a simple question because they just don’t know how to get started on an assignment. All they’d have to do right now is paste the assignment into ChatGPT and ask it to re-explain, summarize, or provide brainstorming assistance. They could do that in a matter of seconds which would then allow them to have more in-depth discussions with teachers so refine their work.
One of the most transformative aspects of AI right now is to reduce learned helplessness. Sure, some students will still just sit there staring blankly but AI can reduce the barriers to entry for students who want to be productive. In turn, this can free teachers up to spend a little bit of extra time trying to find ways to help the students who aren’t as motivated.
Current Use Cases
Generate notes from a teacher’s presentation
Summarize notes for students
Provide feedback to students, reword directions, generate outlines, brainstorm ideas
What could be possible in the future?
Imagine a classroom where AI is integrated into multiple platforms and aspects of the learning environment. Imagine a platform that could pull from a database of lessons, scaffolds, and state standards in order to tailor projects to students. Rather than relying on teachers to make best guess efforts at providing individualized lessons that challenge and engage students across the spectrum, we could develop tools that allow students to tailor their own lessons and assignments that fall within predetermined parameters or meet certain key standards.
Future Use Cases
Provide feedback and guidance without the need for third-party apps or websites
Help teachers troubleshoot behavior issues and generate interventions (funnily enough, I am working on a model to help with this)
Create robust differentiation driven by student input
AI could provide interactive guidance to students: rather than simply answering questions AI could
How are you using AI currently or how would you like to use it in the future?