Forecast · FC-AI-01 · August 2026
They will use it for judgment before judgment has formed.
The honest forecast is short. The young will nearly all use AI; they will use it for judgment before judgment has formed; and the school sector will adopt in full before it decides how.
Shilpa Kapur SinghForecast companion to The State of AI Use
What the data shows, and what it carries
Six questions we are answering
All six dossiersWhat should a teenager’s school day look like?
Align the day with adolescent biology: later where the bell can move, and demanding work kept out of first period. It uses the schedule you already have.
Updated 25 September 2026Open the dossier 02 · DisadvantageIs your data measuring children, or birthdays?
Some of what a school records as a child’s deficit is made by the school, and can be removed by the same lever that made it.
Updated 25 September 2026Open the dossier 03 · AssessmentWhat does one grade actually tell you?
The blended grade is unreliable. Act on that, and design any reform as a structured experiment, never a settled fix.
Updated 25 September 2026Open the dossier 04 · The devicePhones are going. What should a school expect?
Restrict on the distraction evidence, and plan for a behaviour cost that arrives first and falls unevenly.
Updated 25 September 2026Open the dossier 05 · AIWhere should AI sit in a developing mind?
It should extend a developing mind, not replace the effort that forms it, and support a young person’s relationships, not stand in for them.
Updated 25 September 2026Open the dossier 06 · DevelopmentHow should school change as a child grows?
Redesign the joins between phases, starting with the move into secondary. And refuse the closing window.
Updated 25 September 2026Open the dossierEach question has a living dossier: the current answer in plain words, what would change our mind, the papers behind it, a dated record of every change, and how to cite it.
Use the evidence
Nothing you enter is stored or sentWhich timing lever should you pull first?
Three questions. Your priority order, with the moves that need no one’s permission first.
3 questions → DT-002 · DisadvantageIs your school manufacturing disadvantage?
An audit of your structures, not your students.
12 questions · 5 minutes → DT-003 · AssessmentShould your school change how it grades?
Ends in an experiment to run, never a verdict.
15 statements → DT-004 · The deviceWill your phone policy survive a challenge?
Where your policy is most exposed, before a board asks.
12 questions · 4 minutes →Reports
The research programmeThe case for change
The school day is a design decision
The strongest, most equitable improvement most schools can make this year is not curricular.
White paper · June 2026
The school day is a design decision
A structural reading of disadvantage
A measurable share of recorded deficit is manufactured, not measured.
White paper · June 2026
Manufactured, not measured
Assessment as architecture
The assessment engine
White paper · June 2026
The assessment engine
The phone as architecture
The device
White paper · August 2026
The device
Designing by developmental stage
The developing brain
White paper · September 2026
The developing brain
The forecast, scored in public
Read What FollowsWe publish what the trend lines carry, date it, and mark it right or wrong each September, in public.
From the editor
Most of what decides a school’s results was settled before the evidence existed, and has not been revisited since.
Shilpa Kapur Singh · Founder and editor · Applied neuroscientist and education leader
Latest from the Journal: Cognitive debt: what AI quietly takes when thinking gets easier 28 September 2026
The Briefing
Five minutes on what moved. Every Monday.
Free. Written for school and trust leaders.