← Back to the open letter
Classic Mode (Cognitive Abattoir) Socratic Mode (Training)
μ_MCS = 0.5
Instant generation
Zero metabolic effort
Maximum risk of FCPT
Adaptive friction
Calibrated effort
Mobility L1 → L2
Guided step-by-step resolution
Maximum metabolic consumption
Resistance to atrophy

Labora. Roga. Proba.

A level-conditioned oral verification protocol for AI-assisted homework

Students may use AI for homework, at an MCS (Mechanism of Cognitive Stimulation) level set by the teacher. The teacher then verifies understanding through random oral checks, with questions generated according to that level.

The protocol needs no server, accounts or new apps. It uses the AI assistant the teacher already has, and grading stays entirely with the teacher.

The rationale, related work and limitations of the protocol are described in the paper published on Zenodo.

How it works

  1. The teacher sets the MCS level for the assignment. If the assignment aims for the student to learn a method, the level is higher. If the method is merely a tool for something else, the level is lower.
  2. The student does the homework with the AI assistant configured at that level, via the slider extension or the MCS instructions, and submits it electronically.
  3. The teacher randomly selects a few students in the next class. No student knows whether they will be called on.
  4. The teacher generates the question with the form below and sends it to their own AI assistant.
  5. The teacher asks the question and assigns the grade. The AI formulates the question but takes no part in grading.

The extension and the MCS instructions for students are on the implementation page.

Question generator

Fill in the fields, copy the prompt and paste it into your AI assistant. The prompt updates as you type.

05

Task type
Prompt for the AI assistant

The prompt is assembled only in this browser. Nothing you type here is sent to the server.

Pinned values stay only in this browser. On a computer shared by several teachers, don’t pin them.

The second question

If the first question misses what the student worked on, the student may ask for a second one, on the same stage. So that this option does not become a free exit, the teacher chooses one of two variants:

  • Grade cap: a student who asks for a second question can obtain at most one step below the maximum of the local grading system.
  • Increased probability: a student who asks for a second question will almost certainly be called on for the next assignment as well.

What the protocol does not do

  • It does not detect AI-generated text and does not prove whether the student used the slider. It verifies understanding, not the process.
  • It does not grade. Any reference point generated by the AI is support for the teacher, not a verdict.
  • It has not yet been empirically validated. The paper describes what should be measured to validate it.

AI assistants have minimum age limits, which differ from one platform to another. Check them before asking students to use one.