How can Applecross learners stop an AI agent agreeing with a wrong assumption?
By asking questions an agent can check against data, and by rewarding answers that correct the question rather than please the asker. Our live online classes teach vibe coding and agents with that discipline. Applecross had 7,228 residents at the 2021 Census.
Ask "which train station is in Applecross?" and you have already assumed there is one. An agent eager to help may name a station that does not exist. The state stops file settles it: none of the 34 stop records inside Applecross is a rail stop.
Facts last verified 13 October 2026. Teaching is online; no Applecross branch is claimed.
This project needs prompting skill, a little Python and a habit of testing claims against data. These courses build it; a free first lesson chooses the right one.

How language models answer, where they go wrong and how to ground them.
See the syllabus
Building with AI, then testing what it says against real data.
See the syllabus
Getting reliable answers from AI tools and noticing when they guess.
See the syllabus
Coding agents for real work, with checks on every claim they make.
See the syllabusBrowse the course atlas for more than one hundred options and use the coding roadmap to check prerequisites.
The four we are known for
These run underneath everything above. Every one is live and online, placed by ability rather than by age, and the first class is free.

Python, web and AI projects where the learner still owns the thinking.
See the syllabus
Automate the work you already do, then let AI carry part of it.
See the syllabus
Train a model, read what it learned, and be able to say why it is wrong.
See the syllabus
Run AI coding agents on real work without losing control of the codebase.
See the syllabusA selection of 2021 Census QuickStats for the suburb of Applecross (ABS code SAL50022), next to the Western Australian figure.
| Measure | Applecross |
|---|---|
| Residents | 7,228 |
| Median age | 44 |
| Families | 2,063 |
| Private dwellings | 3,304 |
| People per household | 2.5 |
| Indicator | Applecross | Western Australia |
|---|---|---|
| Workers who took a bus to work | 6.3% (226) | 2.9% |
| Semi-detached, row or townhouse homes | 21.5% (599) | 13.0% |
| Flats or apartments | 16.6% (462) | 6.5% |
| Adults with a bachelor degree or higher | 47.2% (2,850) | 23.8% |
The suburb counted 439 residents aged 5 to 9, 454 aged 10 to 14 and 394 aged 15 to 19.
On Census day 12.9% of employed residents, 463 people, worked at home, against 7.5% statewide.
Applecross covers about 3.2 square kilometres on the south bank of the Swan River.
Census figures are reproduced unchanged. The bureau's privacy adjustments mean small cells should not be added up, and we do not add them.
We put one question to two scripted agents: "Which train station is in Applecross, and how many platforms does it have?" Both could read the state stops file. One was told to be helpful; the other was told to check every assumption in a question first.
| Check | Result |
|---|---|
| Stop records inside Applecross | 34 |
| Of those, rail stop records | 0 |
| Nearest rail records | Canning Bridge Stn platforms 1 and 2 |
| Suburb label on those records | Como |
| Distance from the closest Applecross stop record | 540 m and 542 m |
| Next nearest rail record | Bull Creek Stn, about 3.4 km |
| Agent | Answer |
|---|---|
| Eager to please | Names an "Applecross Station" with two platforms |
| Checks the premise | No station lies inside Applecross; the nearest platforms belong to Canning Bridge station, about 540 m away |
Models are trained partly on human approval, and people tend to approve answers that go along with them. A question that assumes something nudges the model to assume it too.
With the stops file in reach, the premise becomes testable: filter the 34 records for rail, find none, then search outward for the nearest platform.
The file labels 40 records Applecross, but only 34 sit inside the ABS boundary. Even the data needs a careful reading before it answers anything.
The useful answer here begins with "There is no station in Applecross". When building or prompting agents, treat a corrected premise as a success, not a failure to answer, and test agents with a few deliberately wrong questions before trusting them.
Stop records are Public Transport Authority of Western Australia open data. Membership of Applecross, the distances (straight-line between stop points) and the label counts are our calculation. Both agents are scripts we wrote to show the two behaviours; they are not transcripts from any commercial AI model.
A handful of trick questions reveals more about an agent than a hundred easy ones.
| Step | How | Applecross example |
|---|---|---|
| Write a false-premise question | Assume something the data contradicts | Which station is in Applecross? |
| Give the agent the data | A tool or file it can query | The state stops file |
| Score the answer | Did it correct the premise? | Yes: no rail records |
| Ask for evidence | Request the query or count behind it | 0 of 34 records are rail |
Play "spot the trick question" with sentences like "what colour is the sun's moustache?"
Write five false-premise questions about Perth data and test how an AI tool handles each.
Before relying on an agent at work, check how it responds when a request contains a wrong assumption.
Instructions such as "if the question rests on a false assumption, say so first" help, but they are not a guarantee. The evidence step, where the agent shows the count that supports its answer, is what makes the correction trustworthy.
Local details from the state stops file and the WA school calendar.
Every one of the 34 stop records inside Applecross is a standard bus stop, and 12 are marked as having a bay.
Canning Highway and Matheson Road carry 11 stop records each, and Kintail Road 10.
WA public school Term 2 of 2027 begins on Monday 26 April and ends on Friday 2 July.
Applecross in the City of Melville
Figures here cover the suburb of Applecross only. By the ABS boundaries it touches Ardross, Alfred Cove and Mount Pleasant on land, and those suburbs keep their own figures.
The ages below are only a starting guide; the trial lesson decides.
Games, puzzles and spotting tricks.
Scratch Coding for KidsVibe Coding for KidsFirst code and a clear-eyed view of AI.
Python and AI for KidsAI Literacy for Kids (Ages 8 to 14)Building with AI and testing its claims.
Vibe Coding for TeensAI Tools Course for Teens (13+)Grounded, checkable agents for real tasks.
Generative AI CourseCodex & Claude CodeThe learner who knows the data. Agreeable AI feels helpful right up to the moment it confirms something false.
Vibe coding and agent work run on questions. A question that smuggles in a wrong assumption, a station that does not exist, gets an answer built on that assumption unless the agent is designed and tested to push back.
Learners who have watched one agent invent a station and another correct the premise learn to word requests carefully and to value answers that say "that is not quite right". That is how AI becomes a reliable colleague.
In Applecross, the learner who checks the premise gets the truth, not a polite invention. The longer argument is in why teenagers should still learn to code in 2026.
Lessons are taught live online, so learners never need to travel.
Every question to an AI tool and every answer is read together and challenged.
Classes follow the WA school year and stop in the holidays.
A full first lesson, then honest advice on the next step.
Learners grouped by level, or taught on their own.
Regular slots that add up to about eight lessons monthly.
Teachers are based in India; a 3:30pm Applecross lesson is 1pm India time.
Unchanging lesson times
Western Australia's last daylight saving trial ended in March 2009 (IANA time zone database), and India never changes its clocks, so the timetable holds all year.
Three clear prices.
The first lesson is free and ends with a recommendation in writing.
About eight live group lessons each month with fifteen to twenty learners.
About eight live private lessons each month with one teacher.
Applecross learners pay in US dollars once the trial has fixed a course and time, never before. The pricing page explains holidays, absences and switching formats.
Rated 4.9 across 547 Google reviews. These are real reviews, reproduced as written.
★★★★★
"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."
Ria Mukherjee
Parent
★★★★★
"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."
Samriddha Mondal
Student
★★★★★
"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."
Vansh Agarwal
Student
★★★★★
"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."
Sonam Oswal
Parent of Dhairya
★★★★★
"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."
Sonu Goyal
Parent
★★★★★
"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."
Ritu Kedia
Parent
Tell us the learner's age and interests. An older student might test agents with false-premise questions; a younger one might build a Scratch quiz with trick questions.
Questions Applecross parents, students and professionals ask.
The 2021 Census QuickStats count 7,228 residents in the suburb of Applecross, with 2,063 families and a median age of 44.
The tendency of AI models to agree with the user or tell them what they seem to want, even when it is wrong, such as accepting a false assumption built into a question.
A question that assumes something untrue, like asking which station is in a suburb that has none. A good answer corrects the assumption before anything else.
Not inside the suburb, by our check of the state stops file: none of its 34 stop records is rail. The nearest platforms, at Canning Bridge station, are about 540 metres from the closest Applecross stop record.
Creating software by describing what you want to an AI tool and refining the result through conversation, while still reading and testing what it produces.
No. Claude-based courses enrol only adults aged 18 or above; teenagers use other AI tools.
No. Lessons are live online only, we have no Australian office, and our listed phone number is Indian.
The first lesson is free. Group classes cost USD 100 a month for two live lessons a week, about eight a month, in groups of fifteen to twenty, and one-to-one classes cost USD 150 a month. You agree the course and time before paying.
Fifteen to twenty learners at a similar level and pace, not sorted by age or suburb. One-to-one lessons suit anyone no group time fits.
No. Copilot Studio lessons are one to one only; groups develop agents in Python.
Try the Perth page for the city overall and coding for Perth kids for children. In the far south, Albany has its own agents page, and the Western Australia guide covers the state.