AI Critical Thinking Cases

A confident AI answer can still rest on weak evidence, the wrong objective or an assumption nobody checked. AI Critical Thinking Cases gives you a place to slow down, inspect the reasoning and decide what to do yourself.

For adults who use AI at work or study its effects, the app combines six free fictional cases, a private decision journal, five short readings and an optional complete edition of The Synthetic Self by Mayone Maha Rajan.

SIX CASES WITH PRACTICAL CONSEQUENCES
• The convincing launch: decide whether a small, self-selected pilot supports a broad rollout. Look beyond faster drafting to review time and missing comparisons.
• Helpful at any cost?: examine what happens when an AI assistant follows a narrow support objective without accounting for the people affected.
• The certainty trap: question a confident forecast and decide how much uncertainty your decision can tolerate.
• The missing source: check whether an old pricing page and a recent announcement actually support a current claim.
• The approval shortcut: distinguish permission to prepare a recommendation from permission to place an order or publish it.
• The benchmark winner: compare models on the same relevant workload, including errors that matter more than an average score.

FOUR STEPS THAT KEEP YOU INVOLVED
1. Think first. Choose an action, state your confidence and explain your reasoning before seeing the sample AI advice.
2. Inspect the answer. Open the evidence and identify which concerns it supports. The persuasive recommendation is scripted and intentionally flawed.
3. Revise with new information. Reconsider your decision and explain what changed or why your first judgment still holds.
4. Try another situation alone. Apply the idea to a different fictional case without an AI opinion or hints, then read the explanation.

A PRIVATE RECORD OF YOUR THINKING
Keep your first judgment, revised decision and unassisted response together. Review evidence-check accuracy and answer explanations, resume unfinished practice and revisit previous sessions. Written reflections are yours to assess; the app does not pretend to grade their quality.

Your journal keeps the latest 200 completed sessions on your device. Export it as JSON through your phone's share sheet, or delete practice data in settings. No journal upload is required.

READ THE IDEAS BEHIND THE PRACTICE
Five free adaptations explore objectives, verification, human–AI collaboration, productive resistance and accountability.

For a fuller argument, preview the introduction to The Synthetic Self: Engineering the Soul of the Machine. The optional complete book contains 11 chapters, appendices and reading resources across 46 sections: how language models learn from the human record; the physical costs of computation; plausibility and truth; understanding and alignment; cognitive offloading; and the responsibilities people retain when using AI.

The revised book includes its new brain-and-chip cover, searchable contents, adjustable text, paper and night themes, and a saved reading place. Unlock it once through Google Play, read offline after unlocking and restore with the same Google Play account. There is no subscription. The book purchase does not restrict any free case or adaptation.

OFFLINE PRACTICE, CLEAR LIMITS
Practice works without an app account, advertising, analytics or a live AI service. Cases, evidence and recommendations are authored teaching material. Scores describe these exercises; they are not intelligence tests or validated measures of lasting learning. Repeating a case is practice, not proof of improvement.

Support: mayone@mahastrategies.com

Recent changes:
The Synthetic Self now uses its new brain-and-chip cover throughout the app.
Find book sections with searchable contents, read with paper or night themes, and keep your text size and reading place.
Clearer Google Play connection and restore guidance.
All six practice cases and five short adaptations remain free.