There is no doubt that Artificial Intelligence (AI) is going to influence the way we do school in the very near future. I have been pondering what that influence will look like. What are the implications now and what will they be in just a few short years.
Now: AI is going to get messy. Unlike when Google and Wikipedia came out and we were dealing with plagiarism issues, AI writing is not Google-able, and there are two key issues with this: First, you can create assignments that are not Google-able, but you are much more limited in what you can create that is un-AI-able. That is to say, you can ask a question that isn’t easily answerable by Google search, but AI is quite imaginative and creative and can postulate things that a Google search can’t answer, and then share a coherent response. The second issue is that AI detectors are not evidence of cheating. If I find the exact source that was plagiarized, it’s easy to say that a student copied it, but if a detector says that something is 90% likely to be written by AI that doesn’t mean that it’s only 10% likely to be written by a person. For example, I could write that last sentence in 3 different ways and an AI detector would come up with 3 different percentages of likeliness that it is AI. Same sentence, different percentage of likelihood to be AI written, and all written by me.
So we are entering a messy stage of students choosing to use AI to do the work for them, or to help them do the work, or even to discuss that topic and argue with them so that they can come up with their own, better responses. We can all agree that the three uses I shared above are progressively ‘better’ use of AI, but again, all are using AI in some way. The question is, are we going to try to police this, or try to teach appropriate use at the appropriate time? And even when we do this, what do we do when we suspect misuse, but can’t prove it? Do we give full marks and move on? Do we challenge the student? What’s the best approach?
So we are in an era where it is more and more challenging to figure out when a student is misusing AI and we are further challenged with the burden of proof. Do we now start only marking things we see students do in supervised environments? That seems less than ideal. The obvious choice is to be explicit about expectations and to teach good use of AI, and not pretend like we can continue on and expect students not to use it.
The near future: I find the possible direction of use of AI in schools quite exciting to consider. Watch this short video of Sal Hahn and his son, Imran, working with an Open AI tool to solve a Math question without the AI giving away the answer.
When I see something like this video, made almost 6 month ago, I wonder, what’s going to be possible in another couple years? How much will an AI ‘know’ about a student’s approach to learning, about their challenges? About how best to entice learning specifically for each student? And then what is the teacher’s role?
I’m not worried about teachers being redundant, on the contrary, I’m excited about what’s possible in this now era. When 80% of the class is getting exactly the instruction they need to progress to a grade standard in a class on the required content, how much time does a teacher having during class time to meet with and support the other 20% of students who struggle? When a large part part of the curriculum is covered by AI, meeting and challenging students at their ideal points of challenge, and not a whole class moving at the class targeted needs, how much ‘extra’ time is available to do some really interesting experiments or projects? What can be done to take ideas from a course across multiple disciplines and to teach students how to make real-world connections with the work they are studying?
Students generally spend between 5 and 6 hours a day in class at school. If we are ‘covering’ what we need to with AI assistance in less than 3 hours, what does the rest of the time at school look like? Student directed inquiries based on their passions and interests? Real world community connections? Authentic leadership opportunities? Challenges and competitions that force them to be imaginative and creative? The options seem both exciting and endless.
The path from ‘now’ to ‘the near future’ is going to be messy. That said, I’m quite excited about seeing how the journey unfolds. While it won’t be a smooth ride, it will definitely be one that is both a great adventure and one that is headed to a pretty fantastic destination.
_____
Update: Inspired by my podcast conversation with Dean Shareski, here.