top of page
Search

The Five Pillars of AI-Resistant Assessment

Over the past several months I have been writing about what I call AI-resistant assessments, not assessments that ban AI, not assessments that pretend AI does not exist, but assessments that still require a student to think, to participate, to create, and to demonstrate learning in ways that are very difficult for artificial intelligence to replace or fake. The goal is not to fight technology, but to design learning environments where the human mind is still required to do the work. Over time, these ideas have organized themselves into what I now call the Five Pillars of AI-Resistant Assessment: Presence, Production, Personalization, Permutation, and Process. Each pillar by itself makes it harder for AI to replace student thinking, but when combined, they make it extremely difficult for a student to outsource their learning to a machine without it becoming obvious very quickly. This article is really a culmination of the ideas from the previous essays, where each pillar was explored individually, but together they form a much stronger system.


The five pillars can be summarized like this:


Presence – Some of the work must happen in front of the teacher, in real time, where students must think without unlimited technological assistance. Quick writes, discussions, handwritten notes, in-class analysis, and verbal explanations all establish a baseline of what a student actually knows and can do without outside help. Presence is important because it anchors everything else; it gives the teacher a starting point and a reference for the student’s real ability.


Production – Students must create something. Not just answer questions, but produce a piece of work that shows understanding. This might be an essay, a project, a presentation, a model, a story, a skit, or anything that requires assembling ideas into something new. As I discussed in the production article, when students have to make something, not just respond, they are far more engaged and far more likely to actually understand what they are learning.


Personalization – Students connect learning to their own experiences, opinions, decisions, or reasoning, making the work uniquely theirs. This was especially visible when students wrote about where they felt most at home or when they had to make personal decisions in their writing. AI can generate general answers, but it struggles to replicate a student’s personal experiences and personal reasoning without heavy input from the student. Personalization forces the student to put themselves into the work.


Permutation – Students must move ideas between formats or mediums, such as turning notes into an essay, an essay into a presentation, a story into a skit, or information into a diagram, podcast, or visual representation. As discussed in the permutation article, this translation between mediums forces students to actually understand the material because they must reinterpret it rather than simply repeat it.


Process – The focus is on drafts, revisions, and the development of thinking over time rather than just the final product. The most important part of this is often that first handwritten draft, done in class, which serves as a baseline of the student’s thinking before any outside assistance is introduced. By comparing drafts, revisions, and final versions, teachers can see the evolution of student thinking, not just the finished product.



Each pillar alone helps, but when used together, they reinforce each other and create assessments that are much more resistant to being completed entirely by AI.


To see how all five pillars work together, imagine a classroom project rather than just a single assignment. Let us say students are learning about westward expansion and the experience of pioneers traveling across the United States. The teacher begins with Presence by showing students primary source images and diary entries in class and asking students to write observations, questions, and inferences in their notebooks. This work is done in class, on paper, in real time. The teacher now has a baseline of what the students actually think and what they actually understand.


Next comes Production. Students are told that they are going to write a short narrative from the perspective of a pioneer traveling west. They must describe the journey, the challenges, the environment, and the decisions they would have to make along the way. They are not answering questions; they are creating something.


Then comes Personalization. Students are asked to include in their narrative at least one decision where they must choose between two difficult options and explain what they would choose and why. Now the assignment is no longer just historical knowledge; it includes the student’s own reasoning, values, and decision-making. The story becomes partly theirs.


After that comes Permutation. Once the narrative is written, students must convert their story into another format. They might turn it into a short skit, a storyboard, a podcast recording, a diary entry series, or a map with written explanations. Now they must take the same ideas and translate them into a different medium, which requires a deeper understanding than simply writing once.


Finally comes Process. The teacher collects not just the final story or final project, but the notes, the rough draft, the revisions, the edits, and the final version. The teacher can now see how the student’s thinking developed from the first observations all the way to the final product. The draft becomes just as important as the final version.


Now look at what would happen if a student tried to have AI do this entire project. The AI might be able to help write a story, but it cannot easily replicate handwritten notes done in class, it cannot easily replicate personal decision-making and reasoning, it cannot easily replicate multiple drafts written over time in class, and it cannot easily perform a skit or presentation in front of the class. The assignment becomes resistant to AI not because technology is banned, but because the student is required to think, create, translate, revise, and participate across multiple stages and multiple formats.


This is really the idea behind the Five Pillars. AI is very good at producing answers and even very good at producing essays, but education was never supposed to be about answers or essays alone. Education was supposed to be about thinking, creating, revising, discussing, deciding, and learning over time. If assessments are designed around those things instead of just final products, then students still have to do the learning themselves, even in a world where AI exists.


Let me give a sent me hypothetical scenario for my own classroom. I say send me hypothetical because what this really is is a conglomeration of what has occurred with many of my students over many years but not focusing on any one student in particular. This writing project was based off of one of our ELA modules on westward migration and I would always start off by putting a photograph on the board.


The photograph on the board showed a line of wagons stretching across a flat, empty plain. No roads, no towns, no fences, just wagons and people and sky.


“Observations first,” I said. “Only what you can see.”


Pencils started moving.


“Wagons are really small,” someone said quietly.

“There’s like no roads.”

“They’re walking instead of riding.”

“Those wheels look tiny.”


“Write it down,” I said. “Don’t just say it.”


A few minutes later we moved on.


“Now questions. What do you wonder?”


“Why would they leave their homes?”

“How long did it take?”

“What if someone got sick?”

“What if a wheel broke?”

“Where did they sleep?”


They wrote again.


Then came inferences.


“They probably wanted land.”

“It must have been dangerous.”

“They probably had to leave a lot of stuff behind.”

“They probably didn’t know if they would make it.”


By the time we finished, their notebooks were full of messy pencil writing, arrows, crossed-out words, misspellings, half sentences, and ideas squeezed into the margins. At this point, they weren’t writing for a grade. They were just thinking on paper.


The next day I told them the assignment.


“You are going to write a letter from the Old West. You are either traveling west or you have already settled there, and you are writing a letter back home describing your life. This is historical fiction, so your story is made up, but it must be based on what we learned from the primary sources.”


There were a few groans.


“Do we have to make it long?”

“Can someone die in the story?”

“Can we get attacked by bandits?”


“Yes,” I said, “but you must include at least one real problem pioneers actually faced. Food, weather, sickness, broken wagons, river crossings, something real.”


They started writing their first drafts in their notebooks. Heads down, pencils moving fast, lots of crossing out, lots of erasing, lots of starting over.


When students write these letters, something interesting always happens. They almost never write to “Dear Mother” or “Dear Father.” Instead, they write to one of their friends.


Dear Michael,

You would not believe how far we have walked…


Dear Sarah,

I wish you were here but you definitely would not like it…


They stay in the historical setting, but they sneak in little references to their real lives.


This is harder than when we had to run laps at recess.

You would complain more than anyone else here.

Remember when we tried to build that fort and it fell over? That is what our wagon looks like right now.


They are writing historical fiction, but they are also writing to someone real, and that changes how they write. Suddenly the assignment belongs to them.


A few days later I handed out blank paper.


“No notebook paper today,” I said. “People in the past did not always have lined paper like we do. Fold your paper, make your own lines lightly, and then write your final letter in your best handwriting. Handwriting will be part of the grade.”


The room changed immediately.


Students who normally rushed through writing assignments suddenly slowed down. They asked how to spell words. They asked where commas go. They erased more. They rewrote sentences so they wouldn’t run out of space at the end of a line. They suddenly cared about spelling, punctuation, and grammar, not because of a worksheet, but because they didn’t want their final letter to look sloppy.


Without spell check, without autocorrect, without grammar suggestions, they had to rely on what they actually knew. Rules they had learned for years suddenly mattered again because the computer was not there to fix everything for them.


While they were writing, one student raised his hand.


“How long did it take them to get to Oregon again?”


“About four to six months for most wagon trains,” I said.


Another student said, “I wrote that I got to that rock place in December.”


“Independence Rock?” I asked.


“Yeah.”


“If you reached Independence Rock in December,” I said, “you would be in serious trouble.”


They laughed, but then we talked about why Independence Rock mattered, why pioneers tried to reach it by the Fourth of July, and what would happen if they were too late and the mountain passes filled with snow. Some students went back and adjusted the dates in their letters so the journey made sense.


This is where technology could be used deliberately if a teacher wanted. Students could look up trail maps, distances, timelines, and landmarks to make sure their story made historical sense. Technology in a case like this becomes a research and accuracy tool, not a replacement for thinking or writing. The writing, the drafts, the notes, and the final letter are still theirs.


By the time the assignment was finished, I had notebooks full of observations, questions, and inferences, rough drafts full of crossed-out sentences and arrows, and final handwritten letters written carefully on self-lined paper. I could see where ideas started, how they changed, what students added, what they removed, and how their writing improved from beginning to end.


Perfect — this is now the closing section of the article, so it needs to do three things at once:


1. Reconnect the story to the Five Pillars



2. Make a philosophical statement about education and AI



3. Transition naturally into your book without sounding like an advertisement




Below is a closing section that keeps the idea of students never hearing the pillars, restates them briefly, and ends with a stronger, more philosophical and practical conclusion.


You can place this directly after the classroom narrative section.



---


Closing Section — Reconnecting the Story to the Five Pillars


The students in that classroom never heard me talk about the Five Pillars. They never heard the words Presence, Production, Personalization, Permutation, or Process. They never heard anything about AI-resistant assessments. To them, it was just a history and writing assignment. They looked at old photographs, wrote notes, asked questions, wrote a letter, revised it, and carefully copied a final version onto paper.


But if you step back and look at what was actually happening, all five pillars were there the entire time. The notes and observations done in class established what the students actually knew and could think about in real time. The letter itself was something they had to create, not something they could just look up or copy. Writing the letter to a friend and including personal decisions and experiences made the work uniquely theirs. Turning primary source information into a historical fiction letter required them to translate knowledge into a different form. And the drafts, revisions, and final handwritten letter showed their thinking over time from beginning to end. When those five things are working together, they form something like a buffer around the learning process. Not a wall to keep technology out, but a structure that keeps the student in the center of the work.


That is really the goal. The point is not to eliminate AI. The point is not to pretend technology does not exist. The point is to design learning in such a way that student work is still student work, and AI is only invited into the process to whatever degree the teacher decides is appropriate. In some assignments that might mean using technology for research, outlining, or checking facts. In other assignments it might mean no technology at all. The difference is that the teacher is deciding how technology is used, not the student using technology to replace their own thinking. When the five pillars are used together, AI becomes a tool that can assist learning rather than a shortcut that replaces learning.


More details on how each of the Five Pillars connects to different levels of Bloom’s Taxonomy, along with more classroom examples across different subjects and grade levels, will be included in my upcoming book: Beyond AI: Elevating Human Intelligence, A Practical Guide to AI Resistant Assessment Design. But the core idea is actually very simple. If students still have to think, still have to create, still have to revise, still have to translate ideas, and still have to show their thinking over time, then learning is still happening, regardless of how advanced technology becomes.


 
 
 

Recent Posts

See All
Presence the foundation of AI Resistant Academics

Let’s start by taking a look at what the first pillar of an AI-resistant assessment looks like, and that first of the five pillars is presence. In the absolute simplest terms, functionally speaking, g

 
 
 

Comments


bottom of page