Work You Can Explain
Practical AI, computing, and learning for students
term
Ask focused questions, preserve independent reasoning, review generated work, and follow course AI policies.
Use this path to get useful help from AI while keeping the reasoning, verification, and final decisions under student control. After deciding whether assistance is permitted, protect the material before sharing it with the student-data pre-sharing guide.
Educators evaluating a school-level instructional model can use the separate teaching note on testing one component of an AI-supported mastery model.
Use the ordered guides as a recommended route. Begin with the first guide for a systematic introduction, or enter where your current task begins and follow prerequisite links as needed.
Articles
Use interpreters to discover questions, builders to find firsthand context, and primary evidence to reach a bounded conclusion about an AI claim.
Articles
Use hints, questions, layered explanations, practice, and feedback to keep AI-assisted work centered on learning.
Articles
Use a known–unknown–attempt record to preserve independent reasoning and ask AI a focused, verifiable question.
Articles
Build a permitted minimal reproduction and ask AI for hypotheses and investigative steps instead of replacement code.
Articles
Review generated code line by line for requirements, assumptions, dependencies, tests, security, scope, and explainability.
Articles
Recognize invented citations and APIs, sample-only code, hidden assumptions, and unnecessary complexity in AI output.
Articles
Translate course-specific AI rules into clear boundaries for tutoring, substitution, disclosure, privacy, and published solutions.
Articles
Reduce an AI request to the minimum permitted evidence, replace real values when possible, and stop when sanitization cannot protect the material.
Articles
Use a small set of prompt patterns for hints, investigation, explanation, tests, reasoning review, quizzes, and manual verification.
