The first batch of the Brightroom Library is live as of this morning. Five lessons in Number Properties — divisibility, factors and primes, parity, remainders, and the strategy playbook that ties them together. The modular-arithmetic floor that every later topic in Quant rests on. Each lesson is roughly twenty slides. Each slide does one of three things: ask the candidate to predict, reveal a widget that shows why the prediction did or did not hold, or synthesise what the last five slides built.
That grammar — predict, reveal, synthesise — is the interface and the pedagogical contract. The Library does not have videos. It does not have walls of text. It does not have a "watch the lecture, then take the quiz" mode. A candidate finishes a Library lesson having made a small bet on every page and seen it played out.
What's in v0
Five lessons, in order:
NP-1 — Divisibility, the cleanest split. The modular-arithmetic basis behind every "divisible by N" shortcut on the test. The Library teaches the rules as one rule, parameterised by 10 mod d, instead of as a list of party tricks. NP-2 — Factors and primes, the atoms of integers. Prime factorisation as a unique-decomposition identity and the count-of-factors formula derived rather than asserted. NP-3 — Even, odd, and the parity trap. Closure properties walked the way they actually appear on Data Sufficiency. NP-4 — Remainders and cycles. Modular arithmetic as a tool, not a topic; remainders treated as cycles the candidate can predict. NP-5 — The number-properties playbook. A capstone that ties the four prior lessons into a sequence of moves the candidate can run on any number-properties item at test pacing.
How a lesson actually runs
Three moves repeat across roughly twenty slides per lesson.
Predict. A slide asks a question with choices, and the candidate must commit before the reveal arrives. The commit is not a guess against a timer; it is the requirement that the candidate has a position before the rule is shown. People learn from prediction error, not from passive exposure. The whole Library is built on that one finding.
Reveal. The slide that follows uses an interactive widget — sometimes a dot grid, sometimes a sliding numeral, sometimes a small table the candidate manipulates — to play out why the prediction was right or wrong. The widget does the explanation; the text carries only what the widget cannot.
Synthesise. Every fourth or fifth slide rolls the prior moves into a one-sentence claim the candidate is asked to either restate, derive, or apply on a new instance. Synthesis slides are the ones the engine reads from for the candidate's mastery update.
What the Library is not
Three things we get asked, often.
It is not a textbook. No long-form prose, no chapters, no glossary in the back. A lesson is a sequence of interactive slides; a topic is a sequence of lessons; a section is a sequence of topics. The unit of reading is the slide.
It is not a video course. We do not ship video. The reason is not aesthetic — it is that the prediction step does not survive video. A candidate watching a video cannot commit before the reveal; the reveal is already on the screen.
It is not a replacement for the engine. The Library teaches the underlying ideas. The engine routes the candidate through items that probe whether the ideas have been internalised. The two surfaces do different work; a candidate who runs the Library and never sits an engine session is missing half the product, and so is the inverse.
What is not in v0
Three things explicitly held for later batches.
The other Quant topics. Number Properties is the foundation; fractions and percents, ratios and proportions, algebra, and the rest of the Quant topic tree are in authoring. Two further Quant batches are scheduled across the autumn. The first Verbal batch — Critical Reasoning, Assumption family — ships in September.
Adaptive within a lesson. v0 lessons run in a fixed slide order. The engine routes candidates between lessons based on their mastery vector, but inside a lesson the sequence is the same for every candidate. Adaptive-within-lesson is on the v1 roadmap; it depends on the per-slide response data v0 is now collecting.
Author-facing tooling. v0 lessons are authored in a custom internal toolchain that is not yet good enough to expose. The Library scales to its full size only when authoring is fast for non-engineer curriculum staff. That is the work of the autumn.
Who gets it
The v0 Library is available today to every Pro 4-month, Pro 6-month, and Ultra subscriber. Pro 1-month does not include the Library in this release — a 30-day window is too narrow to get materially through five lessons alongside engine sessions and mocks, and we would rather not ship a surface a tier cannot use well.
Free diagnostic candidates can preview one lesson — NP-1, Divisibility — without subscribing. The preview is the same lesson Pro candidates run; it is not a teaser.
The next note will be the August percentile refresh when GMAC publishes it. The lesson after that will be the Verbal Critical Reasoning batch in September.
— Brightroom Curriculum