If AI Exposed an Educational Weakness, What Replaces it?

Well, the answer is not to simply throw the baby out with the bathwater. AI exists. It is here to stay. It is obviously counterproductive to try to bring education back to the pre-technological era and assume that students are going to do everything handwritten with research books they got from the library. Besides, as I expound upon in my book, Beyond Traditional Teaching, Second Edition, AI is in a lot more places than people think it is anyway, and the number of places where AI is already being used under the hood without people noticing is growing. Do you think your spell check operates simply as a list of rules anymore? If you're using Google Docs, it is using AI to help figure out what your misspelled words are. Are you using Grammarly to check your punctuation and phraseology? That is AI-powered. Just the other day, I was ordering groceries on Amazon, and it now has an AI assistant helper that you can ask questions about your different purchase products, compare items, and even suggest alternatives based on previous buying patterns. The role of AI in society is only going to expand, not shrink. So if we cannot eliminate it entirely, then what is the best way to mitigate it so that AI does not hinder our students but instead can actually elevate them?

I believe, personally, the true promise of AI is found neither in our total surrender to it—that would be the death of all thinking and creativity—nor in a blind ostrich approach, trying to stick our heads in the sand and pretend like we can stop it from coming. My belief is somewhere in the middle, where I think that human effort, human creativity, and human capacity can be elevated by it but not replaced by it—only if the tool is used correctly. AI is fire. Use it right with the right ingredients, and you can get a beef Wellington. Use it wrong, and you can burn your house down. Many educators see only houses burning down and think that the whole product is evil. But it is just a tool, and if the tool can be learned and used well, then it can elevate an individual’s capacity even if it does not directly elevate their intelligence.

Let me explain briefly from my own life. When I wrote my first book, How to Not Hate Your Job and Get Home Early: Delivery Driver Edition, it took me over a year to finish that book. I was writing in a very traditional way without any AI assistance, just typing and editing. However, after AI came out and after playing around and experimenting with it, I realized that it could operate as a running editor, fixing things as I went, allowing me a greater capacity to simply focus on the creative aspect of the books without having to concern myself as much with the mechanics. This sped up the process massively, allowing me to write my next book, the first edition of Beyond Traditional Teaching, in six weeks. Freed from the burdens of having to track every jot and tittle of my writing and always, at all hours of the day and night, having an AI I could bounce thoughts and ideas off of, consider ramifications and outcomes, and simply use as a partner to talk through ideas allowed me to process those ideas faster and get them written down more quickly. This is a process that I refer to as synthescribing.

Synthescribing falls in the center of the Moore’s AI Progression Scale. It is the place of balance where a human is able to get the maximum benefit from using AI without sacrificing any of the important creative aspects that make writing truly human. Anything lower on the scale, and you are not getting the benefits that AI can offer you. Anything higher on the scale, and the human is no longer the one leading the work, and it drifts toward autoscribing.

But an important consideration about synthescribing is that, in my opinion, that is how one should use AI as a writing tool and a writing partner—but before anyone can get that far, they need to be taught the boundaries of AI as a tool, when it should be used and when it should not be used.

Again, as a brief example, in my fifth-grade classroom students are expected to write their entire first draft, however long it is, entirely by hand, doing traditional manual research either from assigned articles or databases online or using resources from the school’s library. In either case, they are manually writing down their thoughts and their ideas, showing and working through their thinking. No AI is allowed at this formative stage because, just like with baking, putting ingredients in the oven before they have been properly incorporated will destroy the mix. The same is true of student creativity. I need them to practice finding their own voices as authors, as writers, as thinkers. The introduction of AI at this earliest stage would stymie their ability to formulate these ideas independently and creatively.

Once they have written their first draft, we discuss it, they show their thinking, and we go through several other steps. Then I will allow them to talk over their process and thinking with AI just as a writing feedback coach. So they are getting feedback both from myself and from the AI. Between the two of us, students are getting far more writing feedback than they would normally, which helps elevate their writing because the AI and I often focus on different things. The AI most often focuses on mechanics and improving clarity. Since it has taken that job, I instead focus on content and concept, helping them see connections that elevate their writing. Using AI in this way, my students’ natural capacity is improved but without a loss of their natural creativity.

When they have reached this stage of their essay and they do have a question, I will let them occasionally ask the AI for clarification or to operate as a dictionary or thesaurus, helping them find good words that they can use in their own writing so they do not lose their own voice but are able to elevate the language of their writing. Of course, to make sure that this process stays transparent, trackable, and auditable, two things come into play for me as a teacher. First, I always request a copy of their chat log with the AI, as that lets me see how the iterative process developed and unfolded. Second, if and when they use words that the AI suggested, my students are required to write those words down in their writing notebook along with a definition so they are no longer merely passively accepting better words from the AI, but are themselves learning and owning those words, improving their natural vocabulary. In this way, the use of these new words represents growth as writers, not a replacement of their voice or their ideas at the hands of AI.

This process of synthescribing can work in full concert with my practical tips for making an assessment AI-resistant that I mentioned at the end of my last article. While I will give fuller explanations in articles to come, in a nutshell, as technology advances it is becoming increasingly difficult to make something AI-proof. However, we can make things AI-resistant, and several of those methods I have already mentioned, although not by name, here in this very article. The idea of maintaining student voice in their writing, the idea of showing the process over time through auditing their conversations with AI, and of course their writing first drafts by hand in front of me—between these three, we see personalization, process, and performance, three of the key indicators showing that AI was limited in its ability to influence the final outcome.

Now, does this mean that as a teacher you cannot go full draconian? Absolutely not. If you truly want to go AI-free, then the simplest thing to do is simply not allow it in the places where I do occasionally allow my students to use it. Do not allow them to use it to ask questions. Perhaps you want to focus more on students using manual research techniques like an old-school thesaurus or dictionary. Perhaps you want them to practice using context clues or finding words in texts they do not know and researching them to elevate their writing. It really depends on what you are trying to teach and what you are trying to assess, which will determine whether AI is an appropriate or inappropriate tool for the instruction you are using. You can limit and stop it entirely, at least in front of you in the classroom. But in some cases, such as the one I mentioned above, it can be a useful tool to help speed up certain aspects of the writing process without losing the creativity benefits.

So if AI exposed an educational weakness, what replaces it? Once more: authentic human creativity. That should be the cornerstone of any and every assignment. It is the only way to push back against the insidious effects of AI in education while preserving the benefits it has to elevate your students’ capacity to be the creative, insightful thinkers and questioners that we know they

can be.