We created Force Academy in response to a fundamental contradiction in contemporary education: students have unprecedented access to information, yet relatively few opportunities to participate in the creation and critical examination of knowledge.
Generative AI has intensified this problem. Producing a convincing answer is becoming easier; determining whether that answer is valid, understanding its assumptions, and accepting responsibility for its consequences remain profoundly human challenges.
Traditional classrooms often reward correct answers. Even project-based learning can become an exercise in producing attractive presentations rather than constructing defensible explanations.
Drawing on years of experience in mathematical modelling and STEM education, we established Force Academy in Chongqing, China, in 2026 to explore a different approach.
Our central question is: What if young people were treated not merely as consumers of knowledge, but as emerging contributors to it?
We believe authentic research should not be reserved for university students or elite laboratories. With appropriate guidance, adolescents can formulate meaningful questions, construct models, evaluate evidence, receive criticism, and revise their thinking.
Force Academy aims to make this intellectual process a repeatable, accessible educational experience.
Force Academy operates through a structured five-stage research cycle: Question, Model, Evidence, Critique, and Revision.
Students begin with authentic problems rather than predetermined answers. Working with mentors, they formulate researchable questions, develop mathematical or conceptual models, gather evidence, examine limitations, and revise their explanations through feedback.
The approach is implemented through research workshops, modelling sessions, reading discussions, and student-led investigations.
Three examples illustrate the model:
1. Mathematical Life Lab: Students investigate biological emergence, including collective movement and self-organising systems, by constructing computational simulations, experimenting with parameters, and examining reproducibility.
2. The Responsibility Layer of Generative AI: Students examine how AI changes human verification responsibilities, developing an occupational research model involving 774 occupations and refining their work through academic feedback.
3. Mathematics of Polo: Students connect geometry, motion, probability, and strategic decisions with real sporting observations.
Student outputs include research papers, simulations, datasets, model explanations, and documented revisions.
The essential outcome is not a competition result or polished presentation, but an evidence-based explanation that another person can inspect, challenge, and potentially reproduce.
Force Academy is at an early stage of development and dissemination. Its initial implementation has taken place in Chongqing, China, through student research activities within a school-based setting, while the initiative maintains a broader identity as an independent educational platform.
Our dissemination strategy prioritises sharing research methods and evidence of student learning rather than expanding participation numbers prematurely.
Student projects have created opportunities for intellectual exchange beyond the classroom. Research on generative AI and occupational responsibility has received feedback from academic researchers at international universities, introducing students to authentic scholarly criticism and methodological revision.
We have also begun connecting with international educational and scientific networks, including engagement with UNESCO's Global Engagement Network for Science (GENS), while exploring opportunities for cross-border youth research and science-club collaboration.
The next stage is to develop accessible research protocols, reproducible student examples, mentor guidance, and a six-to-eight-week implementation framework.
Our ambition is to enable educators in different countries to investigate locally meaningful questions through a shared research process, without requiring identical curricula, expensive equipment, or advanced technical infrastructure.
We do not yet claim large-scale replication; establishing credible evidence of tran
Force Academy has evolved from an emphasis on interdisciplinary exploration and mathematical modelling toward a more rigorous framework for reproducible research and intellectual accountability.
Early activities focused on helping students recognise patterns, formulate questions, and represent complex real-world phenomena mathematically.
As the programme developed, we recognised that an interesting model or compelling presentation was insufficient evidence of deep learning. Students needed to explain assumptions, test alternative explanations, identify weaknesses, and document how criticism changed their conclusions.
We therefore strengthened three elements:
First, reproducibility. The Mathematical Life Lab has undergone successive revisions, placing greater emphasis on executable simulations, parameter transparency, and replicable results.
Second, external critique. Student research increasingly incorporates feedback from mentors and relevant academic specialists, with greater attention to distinguishing expert suggestions from formal peer review.
Third, student ownership. We have developed a student research fellowship and leadership structure designed to give learners greater responsibility for inquiry, collaboration, and research communication.
These changes reflect our emerging principle: a student's capacity to revise an explanation responsibly matters more than the apparent sophistication of the first answer.
Start with one teacher, a small group of students, and one authentic question. No expensive equipment or advanced research experience is required.
Follow an adaptable six-to-eight-week cycle: **Question – Model – Evidence – Critique – Revision.**
**Weeks 1–2 | Question:** Students identify a real-world phenomenon that interests them and transform their curiosity into a clear, researchable question.
**Weeks 3–4 | Model and Evidence:** Students propose explanations, identify assumptions, and test their ideas through observation, data collection, mathematical modelling, or simulation. Tools may range from paper and spreadsheets to Python and AI.
**Weeks 5–6 | Critique:** Organise peer discussions or invite mentors to challenge students' methods, evidence, and conclusions. Record uncertainties and constructive criticism.
**Weeks 7–8 | Revision:** Students improve their models and produce a research artifact documenting their findings, limitations, and intellectual revisions.
Assess the quality of reasoning, evidence, and improvement rather than the sophistication of presentations. Students must verify AI-generated outputs and respect research ethics and privacy.
The goal is simple: help young people move from consuming answers to creating, testing, and responsibly revising knowledge.
