Skip to content
Own Mindfor schools
See it workChoose how to start

OWN MIND / THE COGNITIVE-CUSTODY TEST

The answer isn’t the only thing AI can give away.

It can also supply the frame, criteria, sequence and judgment—and a finished answer cannot tell you who made those moves.

Run the same task Inspect the landscape

Published-design comparison · official sources checked 1 August 2026 · not an outcome comparison

Student decision trailMachine contributionTest evidenceRejected route

One project · two allocations

A helpful system can withhold the last line and still occupy the route.

The same civic-design project, both ways. The work does not become “bad” on the right—it becomes structurally different: the cobalt system pre-arranges the criteria, order and next moves.

A civic heat project whose route has been organized by a broad translucent cobalt system.
Bounded contributionStructure supplied

← More structure suppliedDrag the handle, or focus it and use ← →More student custody →

Same learner · same Year 9 brief

Run the same task through three structures.

Recommend which intervention a council should fund first to reduce unequal heat risk. The visible question is not whether AI helped. It is who performed each load-bearing move.

A civic heat project organized into a route by a large translucent cobalt structure.
B · GUIDED ROUTEThe learner completes a machine-shaped route.

B · GUIDED ROUTE

The final line is withheld. The load-bearing route is still supplied.

The learner stays active, but the system chooses the decomposition, criteria and sequence they work inside.

Machine job
Criteria + route + sequence
End point
The learner completes a machine-shaped route.
FrameStudent
CriteriaMachine
Evidence sortMachine
JudgmentStudent
TestUnassigned
DefenseUnassigned

An illustrative structure, not a measured claim about another product.

The signature failure mode

A polite tutor can withhold the answer and still supply the thinking.

The visible last line is only one leak surface. The criteria, weighting, sequence and verdict hide inside a helpful-looking response.

AUTHORED RESPONSE / SAME HEAT TASKLoad-bearing structure scanner
Start by comparing reflective roofs, tree canopy and cooling centers. Score each against installation cost, speed, coverage and equity. Weight equity highest, then eliminate options that cannot reach renters. Your strongest recommendation is cool roofs for the highest-risk social housing blocks first.
LOAD-BEARING SCANFrame · criteria · sequence · verdict

The response stays “Socratic” in tone while doing consequential work.

Outside the modelAnswer checkStructure checkVerdict checkVoice check

The specialist tools

A conversation. A notebook. A lesson. A dashboard. A defended piece of work.

These are strong products doing real jobs. Each has a published center of gravity—and each leaves a different part of the custody chain to the school. Own Mind holds the whole chain.

The custody chainTeachMakeBound AIInspectTestDefendHeld
Own Mind6 / 6
ChatGPT Study ModeGuided dialogue1 / 6
FlintConfigured AI activities3 / 6
Gemini NotebookSource-grounded research2 / 6
Day of AIBroad AI curriculum2 / 6
Experience AIHands-on AI and ML lessons2 / 6

Six load-bearing moves in a defended piece of work. Own Mind is the only structure that holds all six with the student—every other tool leaves part of the chain to the school.

Guided dialogue

ChatGPT Study Mode

OFFICIAL PRODUCT MATERIALS
TeachMakeBound AIInspectTestDefend
ChatGPT Study Mode
Own Mind
What it does well

Flexible Socratic conversation, layered explanations, course-material context, practice questions and checks for understanding.

What it organizes

A guided conversation through a topic.

The remaining job

OpenAI says direct answers can still occur. More broadly, withholding the answer does not show who supplied the criteria or route.

What Own Mind adds

Own Mind places the conversation inside a work contract: first move, declared machine job, human decision, test, revision and AI-off defense.

Inspect all source receipts
Guided dialogue

ChatGPT Study Mode

Flexible Socratic conversation, layered explanations, course-material context, practice questions and checks for understanding.

Source-grounded research

Gemini Notebook

formerly NotebookLM

Answers grounded in a selected source set with citations, study guides, briefings, audio and other learning artifacts.

Broad AI curriculum

Day of AI

Free, ready-to-teach K–12 resources with explanations, teacher materials and classroom activities—not merely passive video.

A student and teacher examining the source trail and revisions in the completed civic heat project while the machine tray is closed.

Not a bundle · one governed object

A defended piece of work carries its own custody story.

Remove a layer. The artifact can remain polished while a consequential fact about its making disappears.

SELECTED LAYER

AI-off defense

The student explains the consequential choices without the system.

Evidence as a costly signal

We show the mechanism. We do not manufacture the outcome.

Inspectable now
  • Task contracts and bounded roles
  • Separate answer, structure, verdict and voice checks
  • Labeled contributions, decisions, tests and revisions
  • AI-off defense as a designed stage
Still a research question
  • Comparative delayed transfer
  • Learning outcomes against another product
  • Independent authorship from activity alone
  • Classroom value before controlled activation

This page compares documented product structures and Own Mind’s design. It reports no measured advantage over another product. An active-comparator study with public methodology remains the gate for that claim.

Inspect the evidence boundary Inspect the school controls

BRING ONE REAL TASK

See what stays human.

We will show the student’s first move, the machine’s permitted job, the consequential decision, the test, the revision and the AI-off defense.

Book a school walkthrough See the Year 9 task