I spent my whole day today trying to save a $200 Chromebook. When I think about the hourly rate, it really wasn’t the best use of my time. School starts in less than a week, and I had many other things on my to-do list. But alas, Gemini made me feel like I could solve it, so I kept troubleshooting, and before I knew it, it was 5:00.
This is an ironic pattern I have noticed lately as I use chatbots more and more for work. It makes me feel like I can do anything. Things I wouldn’t have ever considered even attempting (read: most tech-related projects), I suddenly feel like trying because there is no harm in asking. And when it works, it does feel so good. “I did it myself!” I cry out like a small child learning to tie her shoes.
But there have been a few times now where I realized that just because ChatGPT makes me feel like I could doesn’t always mean that I should. Today, I needed my infinitely patient tutor to tell me when it was time to quit.
As an educator, I have felt mostly excited about AI for students. If kids, like me, can feel more willing to try new things because they know they have access to a chatbot to talk them through it, that is amazing.
One day, my own two sons were complaining that they couldn’t get something to work on their computer. Normally, they would wait for their dad to come try to figure it out with them—they would feel dependent. But on this day, they asked ChatGPT. Before long, their drivers were updated, and they were off and running with their new program. I thought, kids today can truly learn anything they want! I loved how a chatbot empowered them with agency in this moment.
Now, let’s view that experience through the lens of one of the common fears around AI: cheating.
Is it cheating for a learner to ask a teacher how to do something? No.
Is it cheating for a learner to follow steps given in a task list? No.
Is it cheating for a learner to go back to a mentor to say something didn’t work and ask why or what next? No.
Those are all steps to learning. So did my kids “learn” to update their drivers? Perhaps. Will my kids be able to pass a test about updating drivers now? Probably not without a chatbot.
Which feels akin to the calculator argument from my school days. Teachers would argue, “You need to know how to do this because you won’t always have a calculator in your pocket!” And what do I have in my pocket now basically at all times? A calculator. And now, also a chatbot.
Obviously, there are still things we need to “know” in the form of memorization or the ability to solve something on our own. But the question is: Which things, and why?
If virtually anything can be learned or solved using a chatbot, what things are worth solving? If kids in school are going to chatbots to get quick answers, quick essays, quick homework help, they are essentially saying, “This isn’t worth the time it takes to solve.” I actually think my adult brain could have used some of that judgment today when I decided to spend all day working on my Chromebook!
Perhaps if we shift our energy from policing AI use, we can ask some fundamental questions about education:
- What things are worth solving? When is struggle productive?
- What skills and information need to be learned in order to operate in the world students actually live in, the one where they do have a calculator and a chatbot in their pocket at all times?
- What things need to be known without assistance?
- What experiences are valuable precisely because they take time?
- What things are kids wanting to learn, and how can we leverage AI to help them learn it?
The runway to achieving skills (or at least the product of a skill) is getting shorter each day, so we need to decide which runways are worth pursuing. It can feel uncomfortable to those of us who spent years learning a skill or training in a specific field. We may look at someone who AI-generated something and think, “But you didn’t learn how to do that.” And that may be true. But just because their path is different doesn’t mean that it is wrong.
So if the fear is that AI will make kids not learn or not try, I present my Chromebook experience as evidence that the opposite might be true. Sometimes AI will make you feel so capable that you think just one more question, one more set of instructions, one more try, and you will have it. The problem may feel solvable long after it is worth solving, but the tutor will never tell you to quit. And sometimes, we want to cultivate that perseverance. But other times, we should just buy a new Chromebook.
