NSW syllabus · Year 7–12 maths
Ask the question you'd never ask in class.
Send a maths question you're stuck on. Get it explained on video in 60 seconds.
Free for the first 50 students · 36 spots left
$29/month when we launch. The 50 beta spots are free.
The ACED Loop
Ask. Watch. Then you do one.
Three steps the top maths students already use themselves. Here it runs tonight, without you asking anyone.
01 Ask
Type the question or photograph it. The one you are stuck on tonight, not a topic you should revise.
02 Watch
A short video that starts at the step where it breaks, in the method your course actually uses. Checked by a maths engine before it reaches you.
03 Practice
The same idea with different numbers, handed back so the only way forward is to do it yourself.
You get stuck. You don't know what to do about it. So you skip it and tell yourself you'll come back to it.You never come back to it.Enough of those and you decide you're just bad at maths.
The questions don't get harder. They get unfamiliar.
Not because you got worse. Because maths stacks, and one step you never quite got makes everything built on top of it look like something you've never seen. You're not bad at it. You're missing one thing, and it has a name.
Which is why none of this worked.
Moved on.
Gave you an answer. You needed a diagnosis.
Correct — and useless. It built on knowledge you don't have.
Saw you once a week, for an hour, on their schedule.
Taught ahead, when the hole was behind.
Go looking for the step where it actually broke.
The one you already use is costing you marks.
Roughly a thousand high-school maths students. Half were given an ordinary chatbot to practise with. It went well, right up until the exam.
The authors call it a crutch. To be clear: the problem isn't the technology. A version with the worked solution and the common mistakes built in removed the harm entirely. Unguarded is the problem.
Bastani et al., PNAS 2025 · Read the study ↗Same question. One of these answers is wrong.
A flock of 12 600 sheep, males to females 1 : 20. Male weights are normally distributed, μ = 76.2, σ = 6.8. Fifteen males weigh more than x kg. Find x.
Looked z up in a table. There is no z-table in the exam room. Right number, wrong method, no marks.
The empirical rule. The method the question was written for. 2σ above the mean is the upper 2.5%.
Where the marks go
Reaching for a table or a calculator function. This question is built so z = 2 falls straight out of the rule, and the marker is looking for the rule, not the number.
Solve it. Prove it. Aim at the break. Hand it back.
What it won't do.
When the checker can't confirm an answer, we show you nothing and say so. We'd rather be unhelpful than wrong.
Built for NSW Years 7–12 — strongest on Advanced and Extension 1 today. Extension 2 isn't ready, and we'd rather say that than half-do it.
Every explanation ends with the question changed and handed back. That's the design, and the study above is why.
Someone else paying?
No appointment to book, no session to cancel, nothing to argue about at the dinner table. It's there at 11pm because it isn't a person. A page written for parents is coming; until then the honest summary is below.
Ask the question you'd never ask in class.
Nobody has to know you were stuck.