back to the handbook

remix it for a new audience

Reviewed June 2026

Most writing outside of school has a real audience. School assignments often don't, which leaves students little reason to think hard about genre, design, and purpose. A model can make that thinking visible: feed it one text, ask it to remix the same content for a completely different reader, and the choices that audience forces become obvious.

Below is one paragraph from an education article, written for educators. Pick a new audience and watch the model rewrite it. The genre changes everything: the vocabulary, the rhythm, even what the piece seems to be about.

same idea, a different reader

try it: change the audience

Choose who the rewrite is for, then press the button. These are real AI outputs, lightly edited for length.

the original · an education article, written for educators

When we don't reduce people to simple rule-obeyers, teaching takes nuanced, creative approaches, a craft without a fixed set of step-by-step strategies. It's an art, not a technical role. Teachers can't run their classrooms like Amazon warehouses or a call center. Just as workers fight back against the dehumanization of labor, educators have to create spaces that bring human flourishing to their community.

The model is genuinely good at this. It catches the register of each genre: the soft abstractions of self-help, the simple images of a picture book, the cadence of a speech. What it tends to flatten is the argument. Watch the self-help remix quietly drop the labor politics, the Amazon warehouses and the fight against dehumanization, and turn a structural critique into a story about personal growth. Whether that trade is acceptable depends on what you were trying to say, and noticing it is the work worth handing to students.

Remixing also has an edge the model can't cross on its own. AI is good at reproducing the surface conventions of a community's writing, but it has little sense of social context: the references, slang, and inside jokes that signal you actually belong. It might rewrite a homework assignment as a Taylor Swift song, but it doesn't know how to talk like a Swiftie. Students do. Let the model draft, then have them breathe life into it.

breathe life into a fan text

spot it: what a real fan would catch

Here's a real AI draft of a blog post from a Taylor Swift fan at the Eras Tour, next to what a Swiftie actually brought to the same prompt.

Write a blog post from the perspective of a Taylor Swift fan sharing her experience at the Eras Tour.

A real fan would also catch what the AI gets wrong: the tour doesn't close on "Love Story," and fireworks aren't the signature moment. The model can copy the shape of a fan's voice, but the references, the slang, and the lived details are the part only a member of the community can supply.

This is the move worth practicing with students. The model handles the surface, the grammar and the genre, and frees students to do the part that's actually theirs: the knowledge, the context, and the voice that comes from belonging to something.

remix something you're an expert in

try with students

Students are insiders to communities most adults aren't. This activity uses that expertise to show exactly where the model's fluency runs out. Students do the remixing and the noticing themselves; the one place a chatbot helps is marked optional below.

  1. remix by hand first 10 min

    Put a short text the class knows on the board and pick a wildly different audience. In pairs, students remix it themselves, no AI yet, so the genre choices are theirs.

  2. compare with the model 5 min

    Run the same source and audience through a chatbot, or use the remix tool above. Where did the model match the genre, and what did it quietly drop or soften?

  3. draft a fan text 10 min optional · AI use

    Get a deliberately generic draft about a fandom, game, sport, or community a student belongs to. With a chatbot, have it write the post; without one, a student writes it as a clueless outsider would. Either way, keep the bland draft, flaws and all.

  4. breathe life into it, then name what was missing 15 min

    Students mark up the draft with the references, slang, and personal details a real member would add, and flag anything the AI got flat-out wrong. Then turn the conversation to the real question the activity is building toward: what does AI writing lose? Use everything the students had to put back, the lived knowledge, the voice, the sense of belonging, to name what the model left out in the first place.

For moving one idea across languages, reading levels, and tone instead of audiences, pair this with translating and rephrasing.

back to the handbook