To put real robotics into the hands of students whose talent already exceeds their access.

The Initiative begins from a position consistent with the evidence: aptitude for engineering, mathematics, and computational thinking is not concentrated in any one zip code, income bracket, school district, or family. The structures that develop that aptitude into capability are concentrated, and the concentration is severe.

A student who can keep a robot on their desk for a year is having a different educational experience than a student who used one for forty-five minutes in a school club. The same student, in the same school, with the same teachers, learns at a different rate and arrives at a different place. The hardware is not the entire story. It is the precondition for most of the story.

The Initiative makes a deliberate trade. We award fewer scholarships than the scale of the need would justify so that each scholar gets sustained mentorship rather than a kit and a goodbye email. The unit of intervention is one student supported for one year, not one kit shipped.

The components of that unit are:

• Hardware they own. A Reachy Mini, not on loan, not to be returned. Ownership matters. It changes the relationship between the student and the work.
• Mentorship at frequency. Monthly virtual sessions, plus access to engineers and educators between sessions when something is stuck. The frequency is the point. A single workshop is not mentorship.
• A documented record. Scholars publish their work in the Showcase. The act of writing about a project changes what the project teaches. The artifact also becomes part of the scholar's record for future applications.
• A cohort. Students meet other students from other schools and other regions doing comparable work. Isolation is one of the things the program exists to interrupt.

We will report what happened. Not what we hoped would happen. The Initiative tracks a small number of outcomes that are observable rather than rhetorical:

1. The number of scholars who completed the program year, defined as documented projects plus participation in the majority of monthly sessions.
2. The number of scholars who continued working with their hardware after the program year ended.
3. The number of scholars who pursued further STEM coursework, competitions, internships, or programs of study they would not have applied to without the scholarship year.
4. The number of scholars who returned in subsequent years as peer mentors.

These metrics are reported on the Showcase page each year. We are explicit about what the data can and cannot tell us. A small cohort cannot prove causation. It can demonstrate that the program worked as designed, for the students it served, in a given cycle.

The Initiative is not a curriculum vendor, a competition organizer, or a workforce development pipeline for any company. We do not place students into jobs. We do not sell hardware. We do not collect data on scholars beyond what is needed to run the program and report outcomes.

The Initiative is independent. Dr. Finley's work as a learning consultant in industrial automation is conducted separately from the Initiative and confers no obligations on scholars.

Every scholar in the Initiative interacts daily with a large language model through their robot. That interaction is itself an educational object. How a young person learns to engage with AI now, what to verify, when to doubt, how to reason in the absence of certainty, will shape how they think for the rest of their lives.

The program's approach is informed by AI-Epistemic Resilience: A Framework for Knowledge Integrity in the Age of Artificial Intelligence (Finley, 2026), Dr. Finley's recent book on the cognitive and educational practices that allow people to engage with AI critically rather than uncritically. The framework's four elements inform how the scholarship's mentorship is conducted:

• Verification practices — treating AI output as a claim to be checked, not a conclusion to be accepted.
• Epistemic humility — calibrating confidence to evidence, including one's own.
• Adaptive reasoning — adjusting approach when context or information changes.
• Metacognitive monitoring — noticing one's own reasoning while doing it.

These are not a curriculum the program imposes. They are dispositions the mentorship is meant to cultivate alongside whatever technical project a scholar pursues.