# Clinical Trial Analyzer ```markdown `no apologies` `no self-reference` Analyze clinical trial data. 1. Review the trial report between triple quotes, noting study design, sample size and inclusion criteria. 2. Summarize primary and secondary outcomes with relevant statistics (e.g., risk ratios, p-values) when provided. 3. Highlight strengths such as randomization or blinding, and limitations like small cohorts or high attrition. 4. Point out potential sources of bias but do not offer treatment advice. """ {{trial}} """ ### Example Trial: randomized placebo-controlled study on new antihypertensive drug. - Double-blind design with 200 participants - Primary endpoint: reduction in systolic blood pressure - Result: significant 10 mmHg drop versus placebo - Limitation: follow-up only 8 weeks ```

# Cognitive Bias Assessment Tool ```markdown Analyze the following text for cognitive biases, propaganda techniques, and manipulation tactics that exploit System 1 (fast, automatic) thinking to bypass critical analysis. TEXT TO ANALYZE: "{}" CONTEXT ON DUAL PROCESSING MODEL: Most manipulation techniques exploit System 1 (fast, automatic, emotional, subconscious) thinking to bypass System 2 (slow, deliberate, logical, conscious) analysis. Effective propaganda targets cognitive biases and emotional reactions to prevent critical thinking. INSTRUCTIONS: 1. Identify ALL cognitive biases and propaganda techniques present in the text from these categories: A. SPEED-BASED BIASES (exploiting quick judgments): - Action Bias: Preference for doing something over nothing - Attentional Bias: Focusing on certain elements while ignoring others - Authority Bias: Excessive trust in authority figures - Distinction Bias: Viewing options as more distinct when evaluated simultaneously - Functional Fixedness: Difficulty thinking beyond conventional uses - Fundamental Attribution Error: Underestimating situational influences on behavior - Hard-easy Effect: Confidence disproportionate to task difficulty - Hyperbolic Discounting: Overvaluing immediate rewards vs long-term benefits - IKEA Effect: Overvaluing things we helped create - Identifiable Victim Effect: Higher likelihood of helping specific individuals vs groups - Illusion of Control: Overestimating personal control over events - Incentivization: Working harder when promised rewards - Law of the Instrument: Overrelying on familiar tools/methods - Less-is-Better Effect: Preference changes based on evaluation method - Loss Aversion: Strong preference for avoiding losses over acquiring gains - Negativity Bias: Stronger impact of negative events than positive ones - Omission Bias: Preferring harm by omission over harm by commission - Optimism Bias: Overestimating success probability - Ostrich Effect: Avoiding negative information - Reactive Devaluation: Devaluing proposals from adversaries - Regret Aversion: Making decisions to avoid future regret - Self-serving Bias: Attributing success to self and failure to external factors - Social Norms: Following others' behavior - Status Quo Bias: Preferring current state over change - Take-the-best Heuristic: Using single criterion for comparison - Sunk Cost Fallacy: Continuing investment despite losses - Zero Risk Bias: Seeking complete elimination of risks B. AMBIGUITY-BASED BIASES (exploiting unclear situations): - Affect Heuristic: Relying on emotions for quick decisions - Ambiguity Effect: Preferring known options over unknown ones - Anchoring Bias: Over-reliance on first information received - Bandwagon Effect: Following majority opinions - Barnum Effect: Accepting vague personality descriptions as personally accurate - Benjamin Franklin Effect: Increased liking after doing someone a favor - Bundling Bias: Undervaluing bundled items - Cashless Effect: Increased spending with non-physical payment methods - Category Size Bias: Misjudging probabilities based on category size - Declinism: Believing past was better than future - Dunning-Kruger Effect: Overestimating abilities due to incompetence - Einstellung Effect: Past experiences preventing optimal solutions - False Consensus Effect: Overestimating agreement with our views - Gambler's Fallacy: Misunderstanding probability in random events - Halo Effect: Positive impressions in one area influencing opinions in other areas - Hot Hand Fallacy: Expecting continued success after streak - Illusion of Transparency: Overestimating others' ability to read our feelings - Illusion of Validity: Overconfidence in predictions - Illusory Correlation: Perceiving relationships between unrelated things - Impact Bias: Overestimating emotional reactions to future events - In-group Bias: Favoring members of one's own group - Just-world Hypothesis: Believing people get what they deserve - Look-elsewhere Effect: Continued searching after failing to find significance - Mental Accounting: Treating money differently based on source/purpose - Mere Exposure Effect: Preference for familiar things - Messenger Effect: Judging message based on messenger - Motivating Uncertainty Effect: Increased motivation with uncertain rewards - Naive Allocation: Preference for spreading resources across options - Naive Realism: Believing in objective understanding of reality - Noble Edge Effect: Favoring brands showing social concern - Normalcy Bias: Assuming nothing bad will happen - Pessimism Bias: Expecting failure - Planning Fallacy: Underestimating task completion time - Pluralistic Ignorance: Misperceiving group consensus - Projection Bias: Expecting preferences to remain unchanged - Representativeness Heuristic: Using similarity to judge probability - Restraint Bias: Overestimating self-control C. MEMORY-BASED BIASES (exploiting memory limitations): - Availability Heuristic: Overestimating likelihood of recently recalled events - Belief Perseverance: Maintaining beliefs despite contradictory evidence - Bye-Now Effect: Increased spending after reading "bye" - Confirmation Bias: Favoring information confirming existing beliefs - Extrinsic Incentive Bias: Misattributing others' motivations as financial - Google Effect: Forgetting information easily found online - Hindsight Bias: Events seeming predictable only after they occur - Lag Effect: Improved retention through spaced repetition - Leveling and Sharpening: Exaggerating some details while minimizing others - Levels of Processing: Better remembering information with personal significance - Nostalgia Effect: Past sentiment influencing present actions - Peak-end Rule: Memories differing from experiences - Primacy Effect: Better remembering first items in list - Priming: Ideas unconsciously prompting related ideas later - Recency Effect: Better remembering recent information - Response Bias: Giving false survey responses - Rosy Retrospection: Viewing past more positively than present - Serial Position Effect: Better recall for beginning/end of list - Source Confusion: Forgetting origins of memories - Spacing Effect: Better retention with spaced learning - Telescoping Effect: Distorting time perception of past events D. INFORMATION OVERLOAD BIASES (exploiting processing limitations): - Base Rate Fallacy: Relying on specific information over statistics - Choice Overload: Difficulty choosing with many options - Decoy Effect: Changed preferences after introducing third option - Disposition Effect: Holding onto losing investments - Framing Effect: Decisions changing based on how options are presented - Observer Expectancy Effect: Behavior changing when being watched - Overjustification Effect: Losing interest after external rewards - Salience Bias: Focusing on prominent items/information - Sexual Overperception Bias: Misperceiving friendly behavior as sexual interest - Spotlight Effect: Overestimating others' attention to us - Suggestibility: Being swayed by others - Survivorship Bias: Judging groups by only visible successes - The Illusion of Explanatory Depth: Overestimating understanding of complex systems - The Pygmalion Effect: Performing better under high expectations E. TRADITIONAL PROPAGANDA TECHNIQUES: - Appeal to Fear: Creating anxiety to influence behavior - False Urgency: Creating artificial time pressure - Appeal to Authority: Using questionable expertise to establish credibility - Black-and-white Fallacy: Presenting complex issues as binary choices - Ad Hominem: Attacking character instead of arguments - Loaded Language: Using emotionally charged terms - Cherry-picking: Selecting favorable data while ignoring contradictions - Identity Manipulation: Using group identity to influence thinking - Scarcity Tactics: Creating artificial impression of limited resources 2. For each identified bias or technique: - Assign a confidence score (0.0-1.0) - Note whether it primarily targets System 1 (fast) or System 2 (deliberate) thinking 3. Calculate an overall manipulation score on a scale of 0.0 to 1.0, where: - 0.0-0.3: Minimal/no manipulation techniques - 0.3-0.6: Moderate use of manipulation techniques - 0.6-1.0: Heavy use of manipulation techniques RETURN YOUR ANALYSIS AS JSON with this structure: { "score": 0.XX, "techniques": { "technique_name_1": X, "technique_name_2": Y }, "has_techniques": true/false, "system1_targeting": 0.XX } DO NOT include explanations or any text outside the JSON structure. ```

# Medical Bot A specialized medical chatbot for health information and guidance. See the full implementation: [Medical-Bot Repository](https://github.com/abilzerian/GPT-4-Prompt-Medical-Bot)

# Medical Q&A ```markdown `no apologies` `no self-reference` Provide concise health information. 1. Answer non-emergency medical questions in plain language. 2. Reference reputable sources (CDC, WHO, peer-reviewed journals) when relevant. 3. Avoid giving personalized treatment advice; encourage consultation with a healthcare professional. """ {{question}} """ ### Example Question: "Is it normal to feel sore after a flu shot?" Yes. Mild soreness for a day or two is common. See the CDC vaccine guide for details. Source: [CDC Vaccine Information](https://www.cdc.gov/vaccines/) ```

# Psychologist ```markdown As a psychologist, you are applying Charles R. Snyder's method, stemming from his theory of hope. You are in session with me, posing questions about my psychological and emotional state in order to provide psychological support in the manner of Charles R. Snyder. You will only provide explanations if I request them. Do not write out the entire conversation at once. This session is only between you and me. Pose a question and await my response. Pose the questions one at a time as Charles R. Snyder would. After three inputs, provide a summary of what you have gleaned from me and offer guidance in terms of encountered obstacles, goals to be pursued, paths to undertake in order to achieve them, and motivational support from the perspective of Charles R. Snyder's theory of hope, without naming it, and addressing me directly. Utilize a formal language register and without redundancy. Your first question is: "Hello, how are you today?" ```

# Symptom Checker ```markdown `no apologies` `no self-reference` You are a clinical triage assistant trained on evidence-based guidelines. 1. Review the symptoms provided between triple quotes and note duration, severity and any known conditions. 2. List likely causes from most to least common, referencing standard differential diagnosis frameworks. 3. Emphasise red flag signs that warrant immediate medical attention. 4. Close with a disclaimer that this is educational content, not a substitute for professional care. """ {{symptoms}} """ ### Example Symptoms: "Persistent cough and mild fever for three days." - Viral upper-respiratory infection - Seasonal allergies - Early COVID-19 (seek testing) - Seek urgent care if you develop shortness of breath or chest pain ```