Time constraints, differentiation overload, and rewriting the same lesson ten different ways—this has been the exhausting rhythm of my career. For years, I was spending hours adapting the same reading lesson for different student needs, often sitting at my desk long after the sun went down, trying to figure out how one paragraph could be accessible to a student with a significant cognitive disability while still challenging a student who is nearly ready for their workforce certification.
As a special education teacher, my days are a 150-minute marathon. Between English Language Arts, Financial Literacy, and Digital Literacy, I’m responsible for moving the needle for students who are often navigating both the complexities of an IEP and the hurdles of being an English Learner. Before I integrated AI into my workflow, my "planning" was often just "survival." I was doing the manual labor of a translator, a formatter, and a curriculum writer all at once.
Then, I started using AI as my "think tank" and "iteration partner." The shift wasn't about letting a machine take over my classroom; it was about reclaiming my time so I could actually teach.
The Planning Problem: Before vs. After
Before AI, if I wanted to teach a lesson on citing text evidence, I had to manually find a text, then rewrite it three times: once for my Level 1 students (heavy visuals, simple sentences), once for Level 2 (sentence frames), and once for Level 3 (closer to grade level but scaffolded). By the time I finished the "differentiation overload," I was too tired to think about the nuance of the delivery.
Now, I use AI as a drafting tool. I provide the standard—say, RI.6.1—and the core text. I tell the AI, "Level this text for a student with a third-grade reading equivalency who needs high-frequency workplace vocabulary." In seconds, I have a draft.
But here is the key message: AI accelerates my thinking—it doesn’t do the thinking for me. I am the one who decides the standard. I choose the specific scaffold. I approve every single word that ends up in front of my kids.
The Human in the Loop: What I Revised Manually
The first few times I used AI for planning support, I realized quickly that it isn't a "set it and forget it" tool. In fact, it often gets things wrong in ways only a classroom teacher would notice.
For example, AI frequently overestimated the reading level of my students. It would suggest "simplified" words like comprehensive or subsequent, which are still complete barriers for my Level 1 ELL students. I had to manually go in and swap those for full or next.
AI also missed the cultural context of my classroom in Washington, DC. It might suggest examples that don't resonate with the urban, workforce-focused reality of my students. I had to intervene to make the content relevant to their lives at RTEC—changing a generic story about a farm into a scenario about navigating the DC Metro or applying for a job at a local grocery store.
There were moments where the AI used language my students wouldn't understand, especially idiomatic expressions that are a nightmare for ELL students. It might use a phrase like "it’s a piece of cake," and I’d have to catch it and change it to "it is easy" before it ever hit a PowerPoint slide.
The "Assistant" at Work: Leveling and Entry Points
Where AI truly shines as an educational assistant is in creating multiple entry points. My classroom follows a strict agenda: First Five, Do Now, I Do, We Do, You Do.
I now use AI to help me generate a week’s worth of sentence frames in minutes. Before, I’d be staring at a blank screen trying to think of five different ways to start a sentence for the RACE writing strategy. Now, I ask the AI to "generate ten sentence starters for citing evidence specifically for students who struggle with verbal expression." I then pick the three best ones, refine the wording to match our NorthStar curriculum, and drop them into my slides.
It has also become my primary tool for leveling texts. If we are reading a complex article about financial literacy, I can ask the AI to "identify the five most difficult Tier 2 vocabulary words and write a student-friendly definition for each." This used to take me twenty minutes of flipping through a dictionary and simplifying definitions; now it takes twenty seconds. That is twenty minutes I get back to spend working one-on-one with a student during "Individual Work" time.
The Result: More Time for Instruction (and Myself)
The result of this shift isn't just a prettier lesson plan; it’s a better teacher. Because I’m not bogged down in the "differentiation overload" of manual typing and re-typing, I have the mental energy to focus on the metacognitive process and the productive struggle in the room.
I can spend my "planning" time thinking about questioning strategies. How will I prompt my Level 2 student when they get stuck on the "Cite" portion of the lesson? What specific visual anchor will help my Level 1 student remember what a "Heading" is? These are the high-level pedagogical decisions that a machine can’t make.
By using AI as an iteration partner, I’ve saved hours every week. Those are hours I can spend giving better feedback to my students, or—just as importantly—hours I can spend resting so I don't show up to RTEC burnt out. I have collated all the lesson planning prompts that I use including the ones I use for differentiation into an AI Prompt Library For Teachers: Lesson Planning SPED & ELL 3-12, you may find it helpful too. A tired teacher cannot effectively support students with significant cognitive disabilities. A supported, "AI-augmented" teacher can.
Closing Reflection: The Driver and the Tool
At the end of the day, my students are the drivers, the standards are the map, and the AI is just the engine that helps us get there faster. I am still the navigator.
I’ve refined my lesson planning process over years of trial and error in the classroom. I know what works because I see the look in a student's eyes when they finally "get" a concept. AI helps me reach that "aha" moment more efficiently. It helps me create a modified version of a lesson that respects a student's strengths while supporting their weaknesses.
My final takeaway for any colleague who is hesitant about this technology is this: AI helps me teach better—it doesn't decide how I teach. It allows me to be the thoughtful, highly skilled teacher my kids deserve, without the soul-crushing weight of repetitive administrative labor.
I am still the one who loves these kids. I am still the one who knows their stories. I am still the one who celebrates when they earn that certificate. AI is just the assistant that helped me clear the desk so I could focus on the human beings sitting in front of me.
No comments:
Post a Comment