How AI generation works

Quezmo’s artificial intelligence engine is designed to generate educational proposals and review them before the teacher incorporates them into an activity.

The overall process

1. Interpretation: identifies what is asked and for what purpose.
2. Source preparation: organizes the provided material when available.
3. Planning: distributes the content that should be covered.
4. Generation: creates candidate questions or scenes.
5. Auditing: reviews quality, correctness, and coherence.
6. Selection: chooses the alternatives that pass the controls.
7. Human review: the creator accepts, edits, or discards the result.

Why are there controls?

A question may seem convincing but be ambiguous, include an incorrect answer, or stray from the source. Therefore, generation and review must be treated as separate processes.

No AI system is infallible. The result requires a final review appropriate to the educational context.