Claims Library Entry
Ask AI to wash your car and it suggests you leave it in your garage...
A simple car wash question tripped up most of 53 AI models because they answered the words on the screen rather than the unstated context in the user's head. The fix is not fancier prompts or role-play but having the model interview you first to surface missing details. A structured interview prompt lifted pass rates to 85 percent on the same question.
Published June 17, 2026 by Kamil Banc
Lead claim
Only 5 of 53 AI models knew to drive a car 50 metres to the car wash — and four questions fix it.
Atomic Claims
What this article supports
Copy individual claims as needed.
Claim 1 · Source summary
Car Wash Test Failure
Only five of 53 AI models correctly answered that driving to a car wash 50 metres away.
Claim 2 · Source summary
Role Prompting Fails
Role definitions alone failed to improve scores, moving the model's pass rate from zero to zero.
Claim 3 · Source summary
Structure Beats Role-Play
A structure forcing the model to name the situation and task first raised pass rate to 85 percent.
Claim 4 · Kamil's interpretation
Interview-First Prompting
The interview prompt makes the model ask clarifying questions one at a time before answering.
Claim 5 · Kamil's interpretation
Pattern-Match Blind Spot
Models pattern-matched short distance to walking, never realizing the car itself must reach the car wash.
Evidence
Context behind the claims
Quote
"You can outsource your thinking, but you cannot outsource your understanding."
Key statistics
5 of 53
Only five of 53 tested AI models correctly answered the car wash question more than once in ten tries.
85 percent
Pass rate achieved when a structure forced the model to name the situation and task before answering, with no new information.
Zero to zero
Improvement in pass rate from adding role definitions alone, showing role-play prompting did not fix the blind spot.
Supporting context
The article's methodology centers on a simple benchmark question posed to 53 AI models, revealing that pattern-matching on 'short distance equals walk' caused near-universal failure because models never registered that the car itself needed to reach the car wash. Ablation testing isolated the mechanism: role definitions added nothing, while a structure requiring the model to state the situation and task before answering lifted pass rates to 85 percent without supplying new facts. For practitioners, the actionable technique is an 'interview me first' prompt that makes the model ask clarifying questions one at a time, surfacing tacit context the user knew but never wrote down. The author recommends testing this against a one-line prompt on a real task, then saving the clarified context for reuse in future sessions.
How to Cite
Use the claim-level citation when you need a precise statement. Use the article or claims-collection citation when you want the wider argument and source context.
Individual Claim
Best when you need to cite one atomic claim directly inside a memo, deck, research note, or AI output.
"[claim text]" (Banc, Kamil, 2026, https://kbanc.com/claims-library/ask-ai-to-wash-your-car-and-it-suggests-you-leave-it-in-your-garage)Original Article
Use this when you want to cite the full newsletter article at AI Adopters Club rather than the structured claims page.
Banc, Kamil (2026, June 17, 2026). Ask AI to wash your car and it suggests you leave it in your garage.... AI Adopters Club. https://aiadopters.club/p/ask-ai-to-wash-your-car-and-it-suggestsClaims Collection
Use this when you want to reference the full structured claims collection on this page.
Banc, Kamil (2026). Ask AI to wash your car and it suggests you leave it in your garage... [Structured Claims]. Retrieved from https://kbanc.com/claims-library/ask-ai-to-wash-your-car-and-it-suggests-you-leave-it-in-your-garageAttribution Requirements
- Include the author name: Kamil Banc.
- Include the source: AI Adopters Club or the structured claims page.
- Link to the original article or the claims page you used.
- Indicate any edits or transformations if you changed the wording.
Related Reading
More from the library
A structured prompt approach transforms performance reviews into actionable development plans by interviewing managers through six categories. The method prevents common AI pitfalls by collecting complete information before generating recommendations, producing budget-aligned plans in a single session.
5 claims
A detailed analysis of 30 days of ChatGPT and Claude conversations reveals 10 repeating prompt patterns that demonstrate systematic AI use. The author shares specific prompt structures for tasks like email triage, presentation assembly, and workflow documentation, showing how to treat AI as infrastructure rather than a casual tool.
5 claims
AI adoption fails because of habit problems, not training gaps. This practical guide shows how to build an AI reflex muscle in 20 minutes by automating one annoying task. The goal is developing automatic pattern recognition for AI opportunities.
5 claims