What AI helps me do as a language teacher
I teach English to adults and develop ESL_Master alongside that work. The workflow runs on Claude and Cowork. The tools continue to change, so I document the teaching method separately.
My aim is to connect preparation with what actually happened in the previous lesson. That requires a record of the learner’s goals, vocabulary, recurring difficulties, and independent performance.
Where AI helps
- Drafting lesson materials and practice variations from a teacher-selected goal.
- Preparing dialogues and examples for my review.
- Summarizing lesson notes and suggesting patterns to investigate.
- Helping build interactive materials with audio and feedback.
Routine scheduling, vocabulary tracking, and repeatable checks can be handled by scripts. AI-generated text and interpretations need a different kind of review: I check them for meaning, accuracy, level, and relevance.
What I keep in my hands
I choose the learning goal and decide which draft activities belong in the lesson. During class, I listen, adjust the pace, explain, and respond to the person in front of me. Afterward, I review the evidence before accepting changes to the learning plan.
Writing down teaching principles has made these choices easier to inspect. It also makes disagreements visible: if a generated activity follows a rule but does not suit this learner, I need to change the activity or reconsider how the rule applies.
What I look for in the result
A polished lesson is only the starting point. I want to know whether the learner understood the task, could use the language without a model, and had a useful next step for home practice. Those observations matter more to my teaching than how many files the system can produce.
The following notes describe parts of that workflow: before, during preparation, and after a lesson, and reviewing my own teaching from a transcript.