How we use AI

What we use AI for, and what we don’t.

Alder's Book is an AI-native platform. That phrase is doing a lot of work in the market right now, so here is the specific version: what the models touch, what they never touch, and why we drew the line where we did.

The short version

Two things we can point at in the code, not two things we'd like to be true.

1 /

AI never assigns a grade.

Every correctness decision in Alder's Book is computed deterministically on our servers. No model is asked whether a student was right.

2 /

Drafted problems wait for a human.

Problems written by the generator are marked as pending approval and sit in a review queue until a professor approves them.

What AI actually does

Explains

On request, after a missed problem, or as a hint mid-problem. Hints are capped at two and point at the method rather than the answer.

Tutors

Diels, our tutor, holds a conversation aware of what the student has already mastered.

Plans

Builds each student's daily study plan from their current mastery state.

Summarizes

Turns class-level performance into readable notes for the instructor. No student names are sent to the model, only aggregate statistics.

Drafts

Writes candidate problems for human review.

What AI never does

Grade.

Correctness is deterministic and server-side, by problem type:

Mechanism steps: rule evaluators, with every grading event logged immutably.
Chair, Newman, and Fischer models: compared against a fixed stored answer.
Multiple choice and short answer: compared against a stored key.

Answer keys are never sent to the browser. A student cannot read the key out of the page, and no model is in the grading path.

Decide mastery.

Mastery is an ACT-R model: arithmetic over a student's attempt history, inspectable and reproducible. Not a model's opinion.

Where the tutor’s knowledge comes from

Retrieval over the assigned textbook. The tutor retrieves before it answers.

Courses on the platform

Exercise Physiology
Financial AccountingIntroduction to financial accounting principles and practice
General ChemistryIn previewA comprehensive guide to general chemistry fundamentals
Introduction to PythonAn introduction to programming with Python — variables, control flow, functions, data structures, and file handling.
Introductory PsychologyAn introductory, evidence-based survey of psychological science: methods, learning and memory, cognition, motivation and emotion, development, personality, social psychology, disorders, and therapies.
Organic ChemistryA comprehensive guide to organic chemistry fundamentals

Student data

Our Privacy Policy and Terms of Service are the governing documents.