This week we completed our first real assessment of a school principal's role, and it landed far below our platform average: 15% effective AI exposure versus 28% across all professions so far. That gap is worth noticing, even though we're working from a single real datapoint right now.
What's driving the lower exposure? Principals spend significant time on work that requires accumulated institutional knowledge—knowing which staff member thrives under pressure, why a particular student's behavior shifted, how budget decisions ripple through department morale. They also do constant relationship work: building trust with parents, mediating conflicts, making people feel heard. These aren't tasks where better prompts or bigger models obviously help. A principal can't delegate judgment about a teacher's performance to a system that doesn't know the person's growth arc over five years.
This is preliminary. We're still in the phase where most assessments are synthetic, seeded by researchers playing out hypothetical roles. Real assessments remain rare. But the principal result aligns with something we're seeing across early data: jobs built on judgment, context, and presence seem to have thicker protection than jobs built on processing and recall.
If you work in leadership, management, teaching, or any role where sustained relationships and deep context matter, check how the tasks you actually do day-to-day break down. Do they require knowing people, or just information? That distinction is becoming one of the clearest dividing lines we've found.