Manifesto
A Frontier Lab for Human Learning
We work to ensure that as machines get smarter, humans do too.
Human Capability Is a Frontier
GlowingStar is a frontier lab for human learning. We build and study AI learning systems with a long-term ambition: to expand what people can understand, create, and do.
As machines become more capable, we believe human learning deserves the same depth of ambition. Progress should mean more people able to ask good questions, reason through uncertainty, and contribute ideas of their own.
Build for the Person Who Is Learning
An answer can resolve a question. A learning experience should help someone handle the next question with greater understanding. We design around that ambition, creating opportunities to practice, explain, receive feedback, and try again.
Our work begins with AI tutoring and assessment. We want these systems to adapt to learners while preserving the effort, curiosity, and reflection through which capability develops.
Bring Research Into Practice
We connect building with studying how people learn. Classroom deployments bring our systems into real learning settings, where teachers, students, and researchers can help identify what works, what breaks, and what deserves a closer look.
Our standard is to test our assumptions, make limitations clear, and let the evidence shape the next iteration. A promising interaction is a starting point for investigation. Durable learning remains something to demonstrate.
Protect Independent Judgment
We want people to collaborate with AI while remaining able to question it. That means practicing how to check a claim, explain a choice, recognize uncertainty, and decide when an answer needs more scrutiny.
Assessment should help us understand those abilities. Alongside answers and scores, we study the learning process: the questions people ask, the reasoning they share, and how their confidence relates to their understanding.
Keep Learning Across a Lifetime
Our horizon extends beyond a course or a credential. We want learning systems to support people as they enter new fields, take on harder problems, and find new directions throughout their lives.
That requires attention to motivation, confidence, and belonging alongside intellectual rigor. Our goal is to make meaningful challenge more accessible and help people build the capacity to keep learning.
Keep People in Control
Learners should have a meaningful role in setting their goals and choosing how to pursue them. Educators should retain responsibility for standards, interpretation, and the human relationships at the center of teaching.
We believe useful personalization depends on trust. People should understand what information a learning system uses and why, and have clear ways to question its guidance. Agency is a design principle we intend to carry through every stage of our work.
Build This Future Together
Human learning grows through conversation, mentorship, collaboration, and shared work. We want AI to strengthen those relationships and create more room for people to learn with and from one another.
This work calls for researchers, educators, builders, and learners to shape it together. We invite people who share our ambition to help us explore what learning can become, and what it will take to make that future useful to more people.