What we are hiring for
Composure, patient questioning, and the willingness to leave a student uncomfortable long enough to think. None of the three is visible on a résumé, and all three are visible within twenty minutes of watching someone work.
Governance — guide selection
The Socratic Advisory Panel
Certification exams measure whether someone can describe good practice. They do not measure whether an adult stays regulated when a fourteen-year-old is not, or whether they can resist answering. The advisory panel exists to assess those things directly, with people qualified to judge them, and to keep the judging honest by publishing what would prove the method wrong.
Status label. This is a design document for a pilot, not a report on an operating school. No panel has yet convened and no candidate has been auditioned. The physiological measures described here are working instruments with known limits, and each is labelled with those limits below.
Brief
Why the selection standard has to change
What we are not hiring for
Subject-matter coverage and technology fluency. Both are trainable in weeks. Neither predicts whether an adult can hand a decision back to a child instead of making it for them.
Why a panel rather than a principal
A single hiring judgement compounds its own blind spots across a whole cohort. Independent raters from different disciplines disagree with each other, and the disagreement is the point — it is where a bad hire gets caught.
Composition
Who sits on the panel
Social workers and family therapists
They assess relational capacity, trauma-informed listening, and baseline empathy — whether a candidate can sit with a distressed child without needing the distress to stop for their own comfort.
Developmental psychologists and neuroscientists
They assess the candidate's own emotion-regulation strategies and cognitive reserve. The working requirement is a steady autonomic state: an adult whose calm is real enough to be borrowed by an adolescent who has lost theirs.
Sociolinguists and Socratic scholars
They assess linguistic and narrative range — wordplay, metaphorical inquiry, and the specific discipline of active listening that gets a child to explain, justify, and defend a conclusion instead of guessing at the expected answer.
Practising guides
At least one certified guide sits on every review. Panels made only of specialists drift toward theory; a working guide keeps the criteria tied to what actually happens in a room at three in the afternoon.
Recruitment
Four pipelines, two generations
Three of these pipelines draw on adults whose experience is currently underused. The fourth is not an afterthought: a desk-free, outdoor, movement-heavy school cannot be staffed by experience alone. It needs young guides with the stamina to be on their feet all day, and it needs to pay them like a career.
Former teachers and academic retirees
Educators who still care about learning but were worn down by high-stakes testing and compliance paperwork, not by children. Retraining focuses on qualitative observation: tracking comprehension through explanation and transfer rather than through scores.
Retired creatives — artists, musicians, writers
In the early years the Academy treats music as structure: rhythm, ratio, and pattern taught as a language for mathematical and emotional intelligence. Working and retired creatives can lead that plus movement and kinesthetic sessions without turning them into decoration.
Retired professionals and senior leaders
Judges, engineers, clinicians, administrators, scientists — people who spent decades exercising judgement under real consequence. The offer is part-time, purposeful, and specific: help a young person find what they are actually good at, and step back once they have it.
Early-career guides — young, energetic, physically capable
Recent graduates, coaches, outdoor instructors, working musicians and dancers, and career-changers in their twenties and thirties. A desk-free school runs on legs: trail sessions, movement studios, instrument salons, and long days on foot. Experience alone cannot staff that, and pretending otherwise would break the model within a term.
Staffing model
The paired corps — elder mentors and early-career guides
The elder-mentor pipeline is the distinctive part of this model and the part most likely to be misread as the whole of it. It is not. A room that runs on trails, instruments, movement, and long unbroken attention needs both corps, held to one certification standard and paid on two different structures.
Two roles, not a hierarchy
Early-career guides are not assistants to elder mentors. They hold their own sessions, carry their own certification, and are rated on the same mirror audition. The pairing is a division of labour, not a rank order — and an early-career guide can lead a session an elder mentor observes.
What the younger corps carries
Physical range and stamina: outdoor and trail sessions, movement and dance, instrument work, sports and building projects, and the long unbroken presence a full day of desk-free learning actually costs. Also the fluency to sit at an AI station and recognise drift in real time, because they grew up inside these tools.
What the elder corps carries
Pattern recognition across decades, regulated composure under a student's distress, and the patience to leave a silence alone. This is the part that cannot be shortcut by training, and it is why the two-corps model exists rather than a single younger staff.
Deliberate cross-exposure
Every student cohort works with both. A young person who only ever learns from adults their parents' age loses one register; one who only ever learns from someone twenty-five loses another. The mixture is part of the curriculum, not a staffing accident.
The younger pipeline needs a career, not a stipend
A retiree can take $3,500 a month as purposeful part-time work. Someone twenty-eight with rent and student debt cannot. Early-career guides need full-time salaries, benefits, and a visible path — lead guide, curriculum work, training new guides — or the Academy will train excellent people and lose them annually.
What that costs, honestly
The current pilot budget is modelled on part-time stipends and does not fund a full-time early-career corps. Adding four salaried guides at a defensible wage plus benefits is a material increase on a $662,900 pilot year. That number belongs in the sponsor conversation now, not after the first cohort walks.
Open question, stated as one. The right ratio between the two corps is unknown. Our working assumption for the pilot is roughly one early-career guide for every two elder mentors, weighted toward the younger corps in the physical and outdoor sessions. That is a guess to be corrected by the first year, not a finding.
Protocol
Three-tier audition, not a written exam
Tier one
Autonomic regulation screen
Non-invasive monitoring — heart-rate variability and electrodermal activity — while the candidate works through a simulated high-tension session. The panel is looking for deliberate top-down control and the practical use of SBNRR: stop, breathe, notice, reflect, respond.
Status: instrumented but unvalidated. HRV is a noisy proxy for composure. It is recorded as one signal among several and never used alone to reject a candidate.
Tier two
Socratic mirror audition
The candidate is paired with a student on an open-ended problem. The panel watches for the default: does the candidate lecture, or do they use patient diagnostic questioning that lets the student map their own path? Wait time is scored explicitly.
Status: the primary decision signal. Two panel members rate independently, and split ratings return the candidate to a second audition rather than being averaged away.
Tier three
Socratic mind sandbox
The candidate runs a high-compute oral-assessment session with a student. Guide status requires keeping the tool from becoming a crutch, and teaching the student to notice algorithmic drift and agreement-for-its-own-sake in the machine's answers.
Status: newest tier, least evidence. It is the tier most likely to be redesigned after the pilot year, because we are still learning what a good catch looks like.
A candidate who clears all three tiers is certified for unsupervised guiding with quarterly re-review. A candidate who does not clear a tier is offered coaching and a second audition. The pipeline is deliberately more generous than a hiring process and stricter than a credential.
Sponsors
How a sponsor measures the result
Sponsorship here is an impact investment with published measures, not a donation with a plaque. Each line below carries the horizon on which it can honestly be read.
Operational cost recovered
Sponsors employing working parents can measure reduced absenteeism, shorter vacancy times, and retention of key personnel against the cost of participation. This is the nearest-term and most auditable line.
Measurable within one year using the sponsor's own HR data.
The macro figure, handled carefully
The widely cited $122 billion annual cost of inadequate early-childhood systems in the United States is a national aggregate from third-party analysis. It frames the problem; it is not a per-sponsor return and should never be presented as one.
Third-party estimate. Cited as context, excluded from any sponsor-specific ROI calculation.
Pipeline quality
The long-horizon claim: workers who did not learn to hand judgement to a tool are better at open-ended work. Measured by whether graduates can hold a position, defend it, and revise it under pressure — not by grades.
Multi-year. No sponsor should fund on this line alone until the pilot produces its own data.
Collaborative capacity
Rooms built around shared tables and argued positions train people who can disagree without withdrawing. Sponsors can track cross-team trust and collaborative problem-solving in their own onboarding cohorts.
Proxy measures only. Honest label: promising, unproven.
The fuller financial argument, including the case against compliance-driven school metrics, is on Corporate Strategy, and the twelve-month cost model is on Pilot Operations.
Falsifier ledger
What would retire this selection model
An interdisciplinary panel selects better guides than a written certification exam.
Falsifier — If panel-selected guides and exam-qualified guides produce indistinguishable student explanation quality after two terms, the panel is expensive theatre and should be reduced to a single working guide plus a written screen.
Experienced adults in later career stages make the strongest guide pipeline.
Falsifier — If early-career candidates match or beat the retiree pipeline on the mirror audition and on retention across the pilot year, the pipeline thesis is wrong and recruitment opens on aptitude alone.
A paired corps of elder mentors and salaried early-career guides outperforms either alone.
Falsifier — If cohorts guided by a single-generation staff show equal engagement, physical participation, and explanation quality, the pairing is added cost without added effect and one corps should be dropped. The likelier failure is simpler: if early-career guides leave at more than roughly a quarter per year, the compensation structure has already falsified itself and must be fixed before anything else in this model is defended.
The sandbox tier predicts who can keep an AI station from becoming a crutch.
Falsifier — If sandbox scores show no relationship to observed student dependence six months into guiding, the tier is not predictive and is cut rather than kept for appearances.
Next