Frequently asked questions

Short answers to questions readers ask most often. Entries here are drawn from reader feedback and grow over time. For a correction to the book, see ERRATA.md; to ask something new, see how to give feedback.

Note for the author: these entries are seeded starting points. Review and adjust the wording before publication, and replace or expand them as real questions arrive.

General

Is this book only for mathematicians or computer scientists? No. It grew out of graduate mathematics teaching, but the core idea is not a STEM idea: a learner should be able to explain their work, verify it, and take responsibility for it, whatever AI helped produce. Readers in the sciences, the social sciences, and the humanities will find the framework applies, with some adaptation to their own field.

Does the book tell students not to use AI? No. It assumes AI is part of the toolkit and asks a different question: how to use it so that learning still happens. Used well, AI moves the work from execution to judgment rather than removing it.

How should I cite the book or these materials? See CITATION.cff in this repository.

For students

What is the “translation test”? A quick check of whether you own an idea: can you restate the same argument with the notation or terms changed, without help? If you can, you understand it; if you cannot, you have recognized it but not yet learned it. Part I develops this.

Isn’t asking AI for help just cheating? It depends on what you do next. Getting an explanation and then reproducing the work yourself is learning. Handing in what the tool produced, without being able to explain or verify it, is not. The book draws that line carefully.

For instructors

Do I have to become an AI expert to teach this way? No. The instructor guide (Part II) is built around assessment and course design that work whether or not you use AI yourself, with worked examples you can adapt.

How do I grade work when students have AI? The book’s answer is to assess understanding rather than the artifact: ask students to explain, defend, or extend their work in ways that require the understanding the work implies. Part II gives assignment and assessment models.

For departments and institutions

What is the “Institution’s Trilemma”? The competing pulls an institution faces: staying viable, defending the credential, and serving its educational mission. The institutional guide (Part III) shows why any two can be satisfied at the cost of the third, and how to reason about the tradeoff.

Can we solve this with a detector? The book is skeptical of detectors as a primary solution and treats their equity problems directly. Part III argues for policy that protects learning rather than policing tools.