Jost Amman, Der Buchdrucker (The Printer), woodcut from the Ständebuch, 1568. Deutsche Fotothek.
Debates about AI prompted me to ask students to investigate earlier disputes over information technologies. Each student researched one case, from printing to social media, and wrote a short history for a shared public website. They used AI to get oriented and develop questions, then checked the account against historical sources.
Students need practice finding primary sources and evaluating historical comparisons. Prepare a sources workshop and publishing demonstration, and check that each topic has accessible contemporary evidence. Twenty sources is a substantial research workload, not a requirement for a short classroom adaptation.
Students submit the draft page, research and AI-use documentation, peer review, and revised public page. An adaptation could use an ordinary document in place of a website while keeping the source work and comparison.
Assign a technology, such as the printing press, telegraph, photography, radio, television, photocopying, personal computers, search engines, or social media. Ask students to locate a particular episode in its adoption so the project doesn’t become a history of everything the technology has done.
Each page has five parts:
A final section describes the student’s use of AI and considers what it may have missed or overemphasized. This should refer to the research they actually did.
In a sources workshop, students gather and assess about twenty relevant sources. They add these to a tool such as NotebookLM to help develop an initial account, checking its claims against the supplied materials. A general chatbot can then suggest objections, counterevidence, or other perspectives to investigate.
Drafts go on the site before lightning presentations. Students explain their topic, sources, use of AI, and remaining questions. Each also reviews a classmate’s page for coherence, evidence, the comparison with AI, and documentation of the research process.
In my course, students forked a shared GitHub repository and submitted pages through pull requests. Other publishing platforms could serve the project. Whatever the platform, allow time to learn it so the mechanics don’t consume the research work.
prompts for the sources workshop
What perspectives are missing from this account? What sources would I need to find to tell a fuller story? What historical context is missing but useful? What do we NOT know about this topic, and why?
Constructed example — no historical quotation or student result is implied. A student finds a contemporary complaint that a new information technology lets unqualified people address a large audience. They compare that concern with AI-generated explanations.
A useful account identifies who made the complaint, what authority they stood to lose, and evidence of what actually changed. It also asks whether broader distribution of human writing is equivalent to generating an answer without a traceable author. The similarity supplies a research question; it cannot settle whether today’s objection is justified.
The peer review and final assessment use the same criteria:
Readings such as Adrian Johns’s work on print can help students see why trust in a medium needs explaining. Ask which objections were justified and who benefited or lost out when the technology was adopted. The historical cases should leave room for different conclusions about AI, including reasons to oppose a particular use.
Ask where each historical quotation was found. An attributed quotation in an AI response still needs a source. Students may need help finding the primary material behind a frequently repeated anecdote.
If the comparison with AI could be attached to any technology, ask for a closer connection to the case. If the section on outcomes merely continues the chronology, return to a particular prediction and ask what happened to it.
Budget a publishing demonstration and a help session. Drafts and peer review are easier to manage when everyone can find and read the pages before class.
This sketch is licensed under Creative Commons BY-NC-SA 4.0. You're welcome to share and adapt it for noncommercial purposes, with credit, as long as you share your adaptation under the same license.
Image Jost Amman, Der Buchdrucker (The Printer), woodcut from the Ständebuch, 1568. Deutsche Fotothek. The image's own license applies to it, separately from the sketch.
Suggested citation
Fred Gibbs. "Disruptive Expertise." AI Sketchbook, Amaranth, University of New Mexico, September 25, 2026. https://amaranth.unm.edu/ai-sketchbook/teaching/disruptive-expertise.html. Licensed under CC BY-NC-SA 4.0.