Lie Detector
Swipe fast: real answer or confident AI lie?
Never just marks right or wrong -- names the specific plausible-sounding pattern that made the lie convincing, building the reverse-question instinct the platform's pedagogy is built around.
What goes in, what comes out
A timed swipe game shows AI-generated answers one at a time -- some grounded and correct, some fluently fabricated -- and the learner judges 'trust it' or 'flag it' before the clock runs out. Each card resolves by revealing the real source (or the absence of one) plus the exact linguistic tell that made the fabrication convincing.
Tap/swipe judgment per card, against the clock.
Accuracy score, a personal calibration curve (confidence vs correctness), and a post-round breakdown of which tell each missed card exhibited.
Difficulty tiers (obvious to subtle fabrication); topic pack (general knowledge, code, citations); timer on/off; streak scoring.
Lead magnet at /playgrounds (quiz format, highly shareable); concept lesson inside AIE-203 and AIE-401; feeds the platform's scenario-trivia engine.
Concepts click when you open the machinery.
Three labs are already live and free — the same hands-on style this playground brings to its module.