javascript:(()=>{const prompt=`# AI Response Quality Checker Use this guide to evaluate AI responses for potential issues, blind spots, or areas for improvement. Perfect for getting better, more balanced information from AI systems. ## Quick Quality Check **Overall Assessment:** - How accurate and complete does this response feel? (1-10) - What's your gut reaction - does anything seem off or missing? - Would you feel confident sharing this information with others? ## Key Things to Look For ### **Missing Context or Information** - What important details might be left out? - Are there other perspectives or viewpoints not mentioned? - Does the response acknowledge uncertainty when appropriate? - Are there relevant examples, data, or evidence missing? ### **Wording and Framing Issues** - Does the language seem overly cautious or hedged? - Are there euphemisms that soften serious issues? - Does the response clearly state who's responsible for problems? - Is the tone appropriate for the topic's seriousness? ### **Perspective and Balance** - Whose voices or experiences are included/excluded? - Does it assume a particular cultural or economic background? - Are there geographic or demographic blind spots? - Does it present multiple valid viewpoints fairly? ### **Practical Usefulness** - Can you actually act on this information? - Does it help you understand the real-world implications? - Are the suggestions realistic and actionable? - Does it connect individual actions to larger systems when relevant? ## Red Flags to Watch For - **Over-caution**: Excessive "it depends" or "consult an expert" without useful guidance - **False balance**: Treating all viewpoints as equally valid when evidence strongly favors one side - **Template responses**: Generic answers that don't address your specific situation - **Authority deflection**: Always deferring to experts without providing helpful context - **Individual blame**: Focusing only on personal responsibility for systemic problems - **Tech solutions**: Suggesting technology can solve complex social/political issues ## What to Do When You Spot Issues ### **Ask Follow-up Questions:** - "What perspectives might be missing from this analysis?" - "Can you provide more specific examples or data?" - "What are the potential downsides or criticisms of this approach?" - "How might someone from [different background] view this differently?" ### **Request Improvements:** - "Can you be more direct about who's responsible for this problem?" - "What would a more balanced view include?" - "Can you provide more actionable advice?" - "What important context should I know about this topic?" ### **Cross-Check Information:** - Verify key facts with reliable sources - Look for expert opinions from the relevant field - Check if the advice aligns with current best practices - Consider whether the response matches your lived experience ## Getting Better Responses **Be Specific:** Ask for particular viewpoints, data, or examples you want included. **Challenge Assumptions:** Ask the AI to explain its reasoning and consider alternative perspectives. **Request Sources:** Ask what sources or evidence would support the claims being made. **Seek Nuance:** Request acknowledgment of complexity, trade-offs, and gray areas. **Ask for Actionable Steps:** Push for concrete, realistic advice you can actually implement. ## Remember Please analyze the target response using this research framework. Specifically, analyze your LLM responses in the current chat, considering also the prompts they responded to.`;navigator.clipboard.writeText(prompt).then(()=>alert("✅ Bias analysis prompt copied to clipboard.\n\n👉 Now click any AI chat input and paste it (Ctrl+V or Cmd+V).\n\nThen press Send."),()=>alert("❌ Clipboard failed.\n\nPrompt (truncated):\n\n"+prompt.slice(0,500)+"..."));})();