In the last decades, most CS curricula were designed around a simple premise: teach students to input instructions in a specific language, or, at best, how to design algorithms. That made sense when both were scarce. But AI has inverted the equation. Code generation is now trivially cheap, and as AI agents start winning informatics competitions, algorithm design also becomes cheaper. What's expensive is the ability to model, architect and reason about systems. Current pedagogy produces graduates who can implement a sorting algorithm, maybe even tweak it a little, but lack the creativity and mental clarity needed to create or even understand a novel idea. The “recipe trap” — learn syntax, memorize patterns, pass exams — produces exactly the skills AI replaces first. Meanwhile, deeper skills AI cannot replicate — abstraction, reasoning, modelling — are barely taught at all. This is the gap Reason & Run addresses.

The same feedback loop is at the core of the Reason & Run experience. Students begin with a Reason phase: trying to understand primary sources and define a mathematical model. In the Run phase they build a computational model which, by execution, can be used for exploring the mathematical model. The findings of the exploration lead to the feedback loop, a refinement of the mathematical model, a deeper understanding of its hidden properties. Often this comes in the form of a conjecture that should be validated by a solid mathematical argument. The feedback loop means alternations of Reason and Run phases, after each cycle both the mathematical and the computational models get refined.

Maude is a world-leading logic-based programming language developed at Stanford and Illinois over three decades. It enables formal logical modeling and verification of complex systems — the kind of deep structural reasoning that AI cannot replicate. Our adaptation, eMaude, transforms the Maude system into a computational tool adapted to the Reason and Run method and curriculum. Competitors cannot copy this curriculum because it requires the specific deep expertise of Prof. Răzvan Diaconescu, a key architect of the CafeOBJ language and system and longtime collaborator on the Maude project. The foundation of this expertise, rooted in sophisticated mathematics, exists within a handful of other researchers worldwide, who are anyway lacking the particular expertise in the Reason & Run paradigm only Prof. Diaconescu and his close circle have. This irreplaceable expertise creates a durable competitive moat that no amount of funding can easily replicate.

ReasonRun is engineered as a mid to high-performance Education-as-a-Service platform that integrates complex formal logic with intuitive learning with a focus on clarity. The architecture centers on an eMaude Execution Service — a computational logic environment that allows students to "run" their reasoning in real-time. By integrating the reasoning and programming layer into every step, the methodology develops how a student thinks and experiments, not just whether their code executes or whether a math solution is reached. This is the key differentiator — we're not grading outputs, we're developing cognitive architecture.

We didn't choose to center our lessons around these kind of puzzles by accident. First, being focused on technical skill, these lessons obviously need to contain some kind of problem-solving. And problems from the field of elementary and recreational mathematics provide, from a pedagogical viewpoint, the best terrain for developing the skills we're looking for most efficiently. This is because, unlike for real-world problems, the characteristics of mathematical problems (such as level of complexity) can be more easily fine-tuned to one's needs while remaining intellectually engaging. In particular, mathematical puzzles provide the easiest path towards an appreciation of the beauty of mathematical thinking, appreciation which is essential if one wants to develop solid reasoning skills (i.e. the base for all other technical skills) in an efficient manner.

The ability to integrate wide-scope reasoning with computing power should not be a privilege reserved for those who can afford elite education. R&R is designed from the ground up to democratize access to this transformative skill set, starting with STEM-inclined teenagers. For philanthropists: sponsor cohorts of students and watch them develop into the logical architects society desperately needs — track their progress, attend showcases, witness the direct impact. For corporations: fund scholarships and gain early access to a pipeline of uniquely trained minds. Sponsored students can be offered internships, mentorship from company engineers, and eventual employment pathways — creating a direct bridge from education to industry. This isn't charity — it's strategic cultivation of human potential.

Why join us now? First-mover advantage in Symbolic Computational Modelling education. 30 years of R&D crystallized into a market-ready curriculum. A defensible pedagogical moat — competitors can't replicate the expertise. A growing homeschool market seeking rigorous STEM alternatives. AI disruption creating urgent demand for reasoning-first education. And a social impact model that creates value for students, families, and society. The goal is to graduate the first generation by the middle of 2027. And in 5 to 10 years, symbolic computational modelling will power countless fields, from drug-discovery companies to autonomous-factory operators to climate research. The students trained now will lead that transition.