"My child just finished Python syntax — what comes next?" This is one of the most common questions parents ask, and one of the easiest to answer badly. Bad answers sound like: sign up for the next course, collect the next certificate, or "everyone learns AI now, why bother with problem sets." A good answer has one spine: train thinking with contest problems, consolidate ability through projects, and let AI assist — never substitute — along the way. This article lays out a path aligned with what Accel Thinking publishes, so you can locate your child's stage directly.

See the Spine First: CCC First, USACO as a Supplement

The backbone of the path is contest training, because its feedback is brutally honest — the judge tells you immediately whether the code is right and fast. Our spine is the Canadian Computing Competition: the Junior division covers core patterns like sorting, greedy, simulation, and search; the Senior division moves into graph algorithms and dynamic programming. USACO serves as a parallel pool of comparable problems for variety. GESP and CSP knowledge points appear only as reference contrasts when needed — they are not current learning tracks. This positioning is consistent across all our courses and articles, so families are not pulled in multiple directions at once.

A Four-Stage Path

L1 Language and expression (about 2–4 months). The goal is not "finishing syntax" but writing clearly: variables, input and output, conditions and loops, lists and strings. The bar: independently turn a J1-level problem into working code. Python is the recommended first language because it keeps expression costs low.

L2 Algorithm foundations (about 4–8 months). Enter CCC Junior past problems: simulation and greedy first (this week's Tandem Bicycle sort-and-pair, and the Time on Task greedy counting are typical), then prefix sums, enumeration, and simple search. Every pattern must clear three bars: you can solve it, you can prove it, and you can explain it to someone else. Our CP Weekly walkthroughs are written for exactly this stage.

L3 Contest advancement (about 6–12 months). After solid Junior results, move into Senior topics — graphs, shortest paths, dynamic programming — with USACO Bronze to Silver as parallel practice. This stage requires systematically learning data structures, not just solving problems.

Optional L4. USACO Gold and beyond, or provincial-level material, opens only for students with clear competitive goals. It is not the default route.

Where AI and Projects Fit

AI assistants (chat tools, code completion) can join from L1, but the role must be clear: they provide explanation and feedback; the student proposes and verifies. A practical household rule: write it yourself first, then ask AI for critique; run the samples yourself first, then look at the edge cases AI suggests. Projects are the second leg: after each pattern, build a small artifact — a write-up of a solution, a visualization of a simulation, a mini-game of a search process. Projects are worth as much as contest scores, for both admissions and interest.

Three Common Detours

1. Skipping handwriting and letting AI generate code — saves time short-term, loses debugging and proof ability long-term, and stalls everything after L2.
2. Solving problems without building anything — knowledge does not consolidate; six months later only a vague "I did that" remains.
3. Reversing the plan from certificates and contest calendars — contests are evidence of ability, not the goal itself. Entering before the ability is ready only buys anxiety.

Locate the Stage

This week's two releases — Tandem Bicycle (CP Weekly) and the greedy proof behind Time on Task (Python Notes) — are typical L2 training items: one teaches "how to use it", the other teaches "why it is right." Treat them as a checkpoint on the path: if your child can read and retell them, L2 is progressing well.

One Line to Remember

Build the spine first (contest problems train thinking), then add two wings (AI as an assistant, projects as consolidation) — a learning path is not a course list, but the order in which evidence of ability accumulates.