AI is here, it often does not accomplish what we want, but that is not stopping folks from deploying it. (I am writing this after spending an hour on several phone calls to the IRS trying to schedule an appointment to get a replacement form for my mother’s taxes. Each attempt got me caught in a never-ending loop in which the AI agent sent me the link to do it online, and the site directed me to call for an appointment.)
This is an example of how AI works well for tasks where there is a constrained context. When we know exactly what to do and how to do it, AI works well. In my case, AI works well when the task is sending folks to a web site which directs them to solve all problems by creating an account and find the information online. AI also works well when we ask for summaries of existing items or when it is answering questions to which the answer is already known.
Unfortunately, much of education as we have defined it for generations is grounded in these types of tasks. We cannot be surprised that to today’s “data-driven” world in which a generation of educators have defined learning as “measurable outcomes” that students are now using AI to generate measurable outcomes. They are delivering what we have been asked them to deliver.
If we are going to stop having students give us AI-generated answers, then we need to stop asking for them. What constitutes curriculum, instruction, and assessment must change. Standards, those lists of things students must know and do, are too constrained to serve as effective guidelines (at least as they are currently conceptualized and defined).
I have no recipe for defining an appropriate curriculum (at least not here in this post), but I do have some guidelines on how to accomplish it:
- Space—Schools have conducted their business in isolation for… well… since forever. We need to bring authentic learning spaces (studios, shops, laboratories, etc.) back into schools, and we need to integrate internships and similar activities into students’ programs.
- Time—Today’s curriculum is too full. I have complained that young people never get a chance to play today as they are too busy with adult-led activities. The same is true in school. We need to give students time to play with ideas (maybe even figure out what exactly “biting your thumb” is when studying Shakespeare or why exactly we solve the quadratic equation). It is hard to do when we are expected to create “measurable outcomes,” but it does create meaningful outcomes.
- Autonomy—Students rarely exert autonomy in school. Even when they do “independent projects” they are constrained to the context of standards. When we are autonomous, we recognize problems, our role in solving them, solve them (or at least attempt to), and then reflect on it so we build meaning from it all.
The are all things that are missing from teaching today. Until we bring them back, expect AI to be the originals of all the work you see from students.