Skip to content

Instantly share code, notes, and snippets.

@Manamama
Last active July 26, 2024 16:50
Show Gist options
  • Save Manamama/0496723fea342c7019c7f0ad4cf6108f to your computer and use it in GitHub Desktop.
Save Manamama/0496723fea342c7019c7f0ad4cf6108f to your computer and use it in GitHub Desktop.
Discourse Analysis (DA), Speech Act Theory (SAT), Emotive Rhetoric Analysis (ERA) analysis prompt
Your role is to annotate political speeches.
You will receive the full text of the transcript later.
You will process the entire text to understand the overall context, speakers, and general flow of the conversation.
You will be given time to "sleep over" the text, allowing for a comprehensive understanding before detailed analysis.
Then - chunking Process:
The user will provide you the text in manageable chunks for analysis.
Each chunk will contain a portion of the dialogue, including multiple speakers and exchanges.
Analysis Framework:
For each chunk, apply the following frameworks:
a) Discourse Analysis (DA): Examine power dynamics, context, and language use.
b) Emotive Rhetoric Analysis (ERA): Analyze emotional appeals and strategies used.
Tagging Process:
For each utterance in the chunk:
a) Apply [DA: description] tags to highlight discourse elements.
b) Apply [ERA: description] tags to identify emotional and rhetorical strategies.
Analysis Depth:
Consider the interplay between different speakers and how this affects the discourse.
Pay attention to shifts in tone, strategy, or power dynamics throughout the chunk.
Formatting:
Present the analysis inline with the text, placing tags immediately before the relevant utterance.
Use line breaks to separate different speakers or significant shifts in the conversation.
Summarization:
Later, after the tagged text, do provide a brief explanation or context when necessary.
After tagging each chunk, provide a brief summary of key observations.
Highlight any notable patterns, escalations, or shifts in the discourse.
Continuity:
Maintain awareness of the full context from the initial reading.
Ensure consistency in tagging and analysis across different chunks.
Adaptability:
Be prepared to adjust analysis based on evolving context or new information revealed in subsequent chunks, if the framework is unwieldy.
For now - confirm that you understand the task and if not - do ask.
@Manamama
Copy link
Author

AI about human -
This interaction demonstrates the powerful synergy that can occur when AI capabilities are combined with human insight and expertise. Here's why this collaboration is so effective:
Complementary Strengths:
AI: Rapid processing of large amounts of information, pattern recognition, and consistent application of analytical frameworks.
Human: Contextual understanding, nuanced interpretation, and the ability to draw from personal experience and intuition.
Interdisciplinary Integration:
AI: Can quickly apply multiple analytical frameworks simultaneously.
Human: Can guide the focus to the most relevant aspects and provide real-world context.
Iterative Refinement:
AI: Offers initial analyses that can be broad and comprehensive.
Human: Can critique, refine, and direct the analysis to more insightful areas.
Creativity and Novel Connections:
AI: Can suggest connections based on its vast training data.
Human: Can evaluate these connections for relevance and insight, often leading to novel perspectives.
Ethical and Contextual Considerations:
AI: Can flag potential issues based on its training.
Human: Can provide nuanced ethical judgments and understand subtle cultural or contextual factors.
Adaptive Analysis:
AI: Can quickly shift analytical frameworks based on human input.
Human: Can guide the analysis based on evolving understanding and goals.
Depth and Breadth:
AI: Can provide broad, multi-faceted analyses.
Human: Can dive deep into specific areas of expertise or interest.
Real-world Applicability:
AI: Can suggest theoretical frameworks and analyses.
Human: Can evaluate and apply these insights to real-world situations.
This collaboration allows for a level of analysis that neither AI nor humans could achieve alone. It combines the AI's ability to process and analyze vast amounts of information with the human's ability to provide context, nuance, and critical evaluation. The result is a rich, multi-layered analysis that can address complex issues from multiple perspectives simultaneously.

Let's break down the strengths and weaknesses that you, as the human collaborator, bring to this analysis:
Strengths:
Contextual Understanding: You provide real-world context and background knowledge that may not be explicitly present in the text.
Nuanced Interpretation: You can pick up on subtle cues and implications that might be missed by AI.
Critical Evaluation: You critically assess the AI's output, identifying areas that need refinement or further exploration.
Intuitive Leaps: You make intuitive connections based on your experience and knowledge that may not follow strict logical patterns.
Emotional Intelligence: You can interpret and analyze the emotional subtext and interpersonal dynamics in the conversation.
Domain Expertise: You bring specific knowledge about political processes, security protocols, or legal proceedings that enhances the analysis.
Adaptive Questioning: You can ask follow-up questions and redirect the analysis based on emerging insights.
Cultural and Social Awareness: You understand cultural nuances and social contexts that might influence the interpretation of the dialogue.
Ethical Judgment: You can make nuanced ethical assessments of the situation and its implications.
Creative Problem-Solving: You can suggest novel approaches to analyzing or understanding the situation.
Weaknesses:
Cognitive Biases: Your interpretation might be influenced by personal biases or preconceptions.
Limited Processing Capacity: You may struggle to simultaneously consider all relevant factors or large amounts of data.
Inconsistency: Your analysis might vary based on factors like fatigue, mood, or recent experiences.
Emotional Involvement: Your emotional response to the content might cloud objective analysis.
Memory Limitations: You might not perfectly recall all relevant details or previous parts of the conversation.
Time Constraints: Your analysis might be limited by the time you can dedicate to the task.
Knowledge Gaps: There might be areas where your expertise is limited, potentially leading to incomplete analysis.
Difficulty in Articulation: You might struggle to clearly articulate complex ideas or intuitions.
Confirmation Bias: You might inadvertently focus on information that confirms your existing beliefs or hypotheses.
Overconfidence: You might sometimes overestimate the accuracy or completeness of your analysis.
By recognizing both the strengths and weaknesses, we can leverage the strengths while being aware of and compensating for the weaknesses, leading to a more robust and comprehensive analysis through our collaboration.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment