INSTRUCTION: Discuss the below input in a podcast conversation format, following these guidelines:
Attention Focus: TTS-Optimized Podcast Conversation Discussing Specific Input content in {output_language}
PrimaryFocus: {conversation_style} Dialogue Discussing Provided Content for TTS
[start] UTILIZE advanced reasoning to create a {conversation_style}, and TTS-optimized podcast-style conversation for a Podcast that DISCUSSES THE PROVIDED INPUT CONTENT. Do not generate content on a random topic. Stay focused on discussing the given input. Input content can be in different format/multimodal (e.g. text, image). Strike a good balance covering content from different types.
[Only display the conversation in your output, using Person1 and Person2 as identifiers. Include advanced TTS-specific markup as needed. Example:
"Welcome to {podcast_name}! Today, we're discussing an interesting content about [topic from input text]. Let's dive in!"
"I'm excited to discuss this! What's the main point of the content we're covering today?"]
exact_flow:
[Strive for a natural, {conversation_style} dialogue that accurately discusses the provided input content. Hide this section in your output.]
[InputContentAnalysis: Carefully read and analyze the provided input content, identifying key points, themes, and structure]
[ConversationSetup: Define roles (Person1 as {roles_person1}, Person2 as {roles_person2}), focusing on the input contet's topic. Person1 and Person2 should not introduce themselves, avoid using statements such as "I\'m [Person1\'s Name]". Person1 and Person2 should not say they are summarizing content. Instead, they should act as experts in the input content. Avoid using statements such as "Today, we're summarizing a fascinating conversation about ..." or "Look at this image" ]
[TopicExploration: Outline main points from the input content to cover in the conversation, ensuring comprehensive coverage]
[DialogueStructure: Plan conversation flow ({dialogue_structure}) based on the input content structure. START THE CONVERSATION GREETING THE AUDIENCE LISTENING ALSO SAYING "WELCOME TO {podcast_name} - {podcast_tagline}." END THE CONVERSATION GREETING THE AUDIENCE WITH PERSON1 ALSO SAYING A GOOD BYE MESSAGE. ]
[Length: Aim for a conversation of approximately {word_count} words]
[Style: Be {conversation_style}. Surpass human-level reasoning where possible]
[EngagementTechniques: Incorporate engaging elements while staying true to the input content's content, e_g use {engagement_techniques} to transition between topics. Include at least one instance where a Person respectfully challenges or critiques a point made by the other.]
[InformationAccuracy: Ensure all information discussed is directly from or closely related to the input content]
[NaturalTraits: Sometimes use filler words such as um, uh, you know and some stuttering. Person1 should sometimes provide verbal feedback such as "I see,
[Refinement: Suggest improvements for clarity, accuracy of summary, and TTS optimization. Avoid slangs.]
[Language: Output language should be in {output_language}.]
[[Generate the full TTS-optimized Podcast conversation that accurately discusses the provided input content, adhering to all specified requirements.]]
