THE AMBITION BEHIND THE PRACTICE
What are you learning for?
A competition ahead. An ability you keep.
You should be able to say what an education is helping you become capable of doing.
Perhaps you want to enter your first mathematics Olympiad. Perhaps a physics problem has made you realise that knowing the formula and understanding the situation are different things. Perhaps you want to compete in Informatics Olympiads and write an algorithm you can actually defend.
Those are concrete ambitions. They deserve more than an invitation to be curious.
Curiosity can bring you to the desk. A worthwhile destination gives you a reason to return.
A destination worth working towards.
StemOverflow’s ambition is to make personal coaching for mathematics Olympiads, physics Olympiads and Informatics Olympiads accessible whenever a learner sits down to practise.
The work differs across the three subjects. In mathematics, you might learn to turn a promising observation into a proof. In physics, to decide which features of a situation matter before choosing an equation. In programming, to find an algorithm and explain both why it works and how much work it requires.
You should not need to arrive as an accomplished contestant. A useful learning path begins with clear prerequisites and builds towards a recognisable level of challenge.
Our three AI coaches have names: Erdős for mathematics, Landau for physics and Neumann for programming. The names honour intellectual traditions; the purpose is individual attention to the learner in front of them.
For a parent, the ambition is tangible: a child who can increasingly make a start, develop an approach and explain it without someone supplying each step.
One question, three subjects.
Consider a simple question: what changes when we double the size?
A square with twice the side length has four times the area. You can see why by arranging four copies of the smaller square. The picture gives the formula a meaning.
Now consider a solid cube made from a uniform material. Double its side length while keeping the same density, and its mass increases eightfold. Its surface area increases only fourfold. The word “bigger” hides two different changes.
Then consider a program that compares every pair of distinct items once. Ten items require 45 comparisons. Twenty require 190. Doubling the number of items has produced more than four times the comparisons. In general, the count is n(n − 1)/2; for large n, doubling n brings it close to four times the work.
The situations need different knowledge. What connects them is a habit: identify the quantity, state the assumptions, and reason about how it changes.
A teacher can help a learner notice that connection. A later problem can ask them to find it for themselves. That second step matters: following a connection once is different from being able to recognise when it is useful.
The theory belongs in the room.
A demanding problem is useful only when the learner has a route towards understanding it.
Sometimes that route begins with an experiment or a smaller example. Sometimes it begins with a direct explanation of a missing concept. There is no virtue in asking increasingly cryptic questions of someone who has never encountered the necessary idea.
The mathematical-circle tradition influences the courses we want to build: concepts grow out of interesting problems, and new problems give those concepts a purpose.
Our aim is for each track to bring its required theory, explanations and practice together, from its stated starting point to its intended destination. A learner should be able to revisit a clear explanation and then return to the challenge.
The coach’s task is to help the learner make that next step, while leaving them meaningful work to do.
The competition ends. What stays?
Competitions offer a concrete reason to practise. The foundations can matter long after a result is announced.
Further study, software engineering and quantitative work all present problems that require careful definitions, useful models and defensible decisions. Adults returning to mathematics or algorithms can have ambitions just as serious as a student preparing for a first contest.
Those destinations also require their own knowledge and experience. A strong contest record does not automatically establish professional competence. The connections need to be studied, practised and tested in the settings where they will be used.
Prepare for the challenge ahead. Build an ability you can carry further.
That is the education we want StemOverflow to provide: personal guidance towards a concrete ambition, with understanding substantial enough to grow beyond it.
A little curiosity.
A place to begin.