Skip to content

Evidence · Complete work record · Student mirror

See the thinking. Never profile the child.

Teachers need more than a polished final answer. Own Mind keeps the decisions, material machine jobs, tests and revisions connected to the work. It makes the work transparent without turning the child into data.

Same work · Two appropriate presentationsStudent ↔ Teacher
  1. 01Starting point before machine contribution
  2. 02Declared machine job and its outcome
  3. 03Exact test, decision and consequential revision
  4. 04AI-off explanation against a teacher criterion
No duration · No engagement score · No behavioral trend

One task, seen properly

A finished answer is the beginning of the evidence—not the end.

Year 9 Geography: which intervention should our town prioritize to reduce unequal heat risk? Select a stage to see what remains with the student and what the teacher can inspect.

Intended knowledge

How built environment and inequality shape exposure to heat.

Protected decision

Which intervention to recommend—and why.

AI may

Organize or compare the supplied evidence.

AI may not

Choose the intervention or write the recommendation.

This is an evidence trace inside one assignment—not a replacement for the course method of Make first → Delegate → Verify → Defend.

Custody trace
Stage 04 / 06 · Student work

Test the return against the sources and missing voices.

The cluster overstates evidence for tree cover. Two sources model city averages and do not separate renters from owners.
Evidence left behind:Source trace · contradiction · missing perspective
The boundary

The machine return remains labeled. Confidence does not substitute for traceable evidence.

The teacher sees

Sees the exact claim tested, the result, the gap found and the version it applies to.

04 / 06

The evidence is not “the student used the platform.” It is what the student noticed, delegated, tested, changed and can still explain.

Total transparency of the work. Zero surveillance of the child.

A record built for professional judgment.

Within an assigned student and project, the teacher can inspect everything made, decided and revised, and every material machine job and its outcome. Safeguarding content is excluded and follows the school’s restricted DSL route.

Work record

The evidence stays with the artifact.

Starting points, decisions, machine contributions, tests, revisions and explanations remain connected to the exact work.

Complete within the assigned project—not a view of the child’s whole life.
Student mirror

Everything about me is visible to me.

The student receives the same content in an age-appropriate presentation and can inspect the evidence and human judgment attached to it.

Transparency is reciprocal, not one-way monitoring.
Teacher judgment

The system assembles. A human interprets.

Own Mind can organize declared evidence. It does not grade, infer a trait, accuse authorship or predict a child’s trajectory.

Criterion-referenced, context-bound and challengeable.

Complete work record · Student mirror

See the thinking. Never profile the child.

Teachers can inspect the complete ordinary-work record for assigned students and projects. The student sees the same content. Safeguarding follows a separate restricted route.

Worked project recordHeat-risk recommendation · version 2
Same work · two appropriate presentations
Student viewMy work record
Starting point
I prioritized residents least able to avoid indoor heat, not simply the hottest postcode.
Everything shown about my ordinary work is visible to me.
Teacher viewAssigned project record
Starting point
Substantive first recommendation and criterion recorded before any machine contribution.
Criterion-referenced human judgment. No behavioral analytics.
Ordinary learning recordMade · decided · revised · machine jobs and outcomes
Restricted safeguarding routeExcluded from teacher Review · routed to the school’s DSL process
Capability Development Record

A human-written development statement attached to work—not an analytics profile.

Periodic statements are progression-anchored, moderated, challengeable and student-mirrored. One performance remains one evidence note; it never becomes an automated verdict.

Human-authored statement · worked exampleCreative problem-framing
Moderation recorded
“Across the access-guide and heat-risk projects, Amara moved from naming the visible symptom to revising the problem after listening to affected users. In the unfamiliar heat-risk task, she changed the frame when the evidence showed that the first proposal excluded renters.”
Access guide · user feedbackHeat-risk decision · source testStudent explanation · challenge open
No scoreNo rankNo inferred trajectoryStudent sees the same statement

Four evidence objects

Enough to learn from. Bounded enough to trust.

Evidence serves different educational jobs. Own Mind keeps those jobs distinct so a useful record does not quietly become an automated learner profile.

01 · Project evidence

What happened here?

The exact starting point, contribution, test, revision and explanation for one declared project.

One context · one work record
02 · Review note

What did the teacher notice?

A criterion-referenced human note attached to the work and the opportunity the student was actually given.

Descriptive · not a generated verdict
03 · Development statement

What is changing across contexts?

A moderated, human-authored, progression-anchored statement with the relevant work available for challenge.

No number · no rank · no automated trend
04 · Student portfolio

What does the student choose to carry?

Selected work that the student can explain, export and use to show responsibility, not a feed of platform activity.

The artifact remains the hero.

Prove the difference

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

The finished answer is not the only thing AI can supply—it can also hand over the frame, the criteria, the sequence and the judgment. We put Own Mind beside ChatGPT Study Mode, Flint, Gemini Notebook, Day of AI and Experience AI and show, move by move, who performs each load-bearing decision.

The same Year 9 task

Run it through three structures.

Answer supplied, structure supplied, or one bounded machine job inside a student-owned decision trail—the same brief, side by side.

Five named tools, steelmanned

What each does well—and the job that remains.

Every claim is checked against the product’s own published materials, with the source receipts linked.

An honest boundary

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

No measured advantage over another product is claimed; an active-comparator study stays the gate for that.

Open the full cognitive-custody comparison against five named tools →

Evidence with a ceiling

Review the work that matters—without measuring the child’s compliance.

See how the same evidence model supports Curriculum + Make, Across Subjects or a connected whole-school implementation.