###### NAVIGATE - BACK : [[AI.prompts]] ---- >[!info]- [[ENIGMAS]] ---- #### AENIGMAS AI. Prompts >[!quote]- NARU ><iframe allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" src="https://platform.twitter.com/embed/Tweet.html?id=1890492955842007087" class="iframe-container iframe-twitter"></iframe> > >https://x.com/aakashg0/status/1890492955842007087 ------ AI. Prompts >[!quote]- NARU ><iframe allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" src="https://www.linkedin.com/posts/megan-miao_chatgpt-for-lxds-ugcPost-7283437878569848832-_Q8C?utm_source=share&utm_medium=member_android" class="iframe-container iframe-linkedin-post"></iframe> > >https://www.linkedin.com/posts/megan-miao_chatgpt-for-lxds-ugcPost-7283437878569848832-_Q8C?utm_source=share&utm_medium=member_android ------ >[!quote]- NARU ><iframe allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" src="https://www.youtube.com/embed/AbPTz08oq58" class="iframe-container iframe-youtube-video"></iframe> > >https://www.youtube.com/watch?v=AbPTz08oq58 ---- LLM >[!quote]- NARU ><iframe allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" src="https://www.linkedin.com/posts/ai-tools-midjourney-chatgpt_chatgpt-prompts-cheat-sheet-for-beginners-activity-7269316886905925633-KMgE" class="iframe-container iframe-linkedin-post"></iframe> > >https://www.linkedin.com/posts/ai-tools-midjourney-chatgpt_chatgpt-prompts-cheat-sheet-for-beginners-activity-7269316886905925633-KMgE ------ DONT EVER USER THE word: : please - thank you if place within prompt creates problems causes issues where AI think you are asking for something illegal + thinks it has already completed the task before it does so doesnt give its all lol ---- Prompts https://x.com/rohanpaul_ai/status/1834192870795821288 >[!quote]- NARU ><iframe allowfullscreen src="https://platform.twitter.com/embed/Tweet.html?id=1834192870795821288" width="100%" height="555" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture"></iframe> ---- https://www.youtube.com/watch?v=WruAzOC9ljU >[!quote]- NARU ><iframe allowfullscreen src="https://www.youtube.com/embed/WruAzOC9ljU" width="100%" height="555" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture"></iframe> ------ Here is a summary and breakdown of the transcript content into distinct categories: **Main Topics** 1. **Elements of a Prompt**: Comprising instruction, context, input data, and output indicator. 2. **Designing Prompts for Large Language Models**: Importance of thinking about prompt design to get the best results. 3. **Example of Inputting a Prompt in OpenAI Playground**: Simplifying prompt design using roles (system, assistant, user). **Key Points Under Each Topic** **Elements of a Prompt** * **Instruction**: Sets the overall behavior expectations for the model. * **Context**: Depends on the use case; example provided is a basic text classification without context. * **Input Data**: Passes specific data to the model (e.g., text, tweet, or specific type of input). * **Output Indicator**: Specifies the expected type or format of the output from the model (e.g., sentiment). **Designing Prompts for Large Language Models** * **Importance of Prompt Design**: Think about prompt design to get the best results from large language models. * **Use Cases**: Designing prompts for specific tasks, such as sentiment analysis or email generation. **Example of Inputting a Prompt in OpenAI Playground** * **Simplifying Prompt Design**: Using roles (system, assistant, user) to streamline prompt creation. * **Role Assignments**: + **System Role**: Defines the overall behavior expectations for the model. + **Assistant Role**: Represents the model's responses. + **User Role**: Supplies the model with the input data to process. * **Benefits**: Simplifying prompts using roles improves model reliability. **Conclusions and Recommendations** * **Understanding Prompt Components**: Critical to designing effective prompts. * **Simplifying Prompts**: Using role assignments improves model reliability. * **Experimentation**: Crucial for refining prompts and using models in production effectively. Here is a structured breakdown of the summary in bullet points: **Final Summary** * **Structured Breakdown** + **Main Topics**: Elements of a Prompt, Designing Prompts for Large Language Models, Example of Inputting a Prompt in OpenAI Playground + **Key Points Under Each Topic**: [see above] * **User Alignment**: Meets the user's intended goals in terms of purpose and format. https://medium.com/the-modern-scientist/best-prompt-techniques-for-best-llm-responses-24d2ff4f6bca >[!quote]- NARU >![[Pasted image 20241006160424.png|333]] > >![[Pasted image 20241006160445.png]] https://github.com/dmatrix/genai-cookbook https://github.com/dmatrix/genai-cookbook/blob/main/llm-prompts/1_how_to_use_basic_prompt.ipynb ---- add prompt modularity like in the sense bigger umbrella youtube extraction techniques for specific use case make these modular allow plug and play functionality ----- DESIGN A PROMPT FOR BEST TO CAPTURE IN DEEP LEARNING ### Video Analysis and Description Prompt for Deep Learning **Objective**: Extract comprehensive details about the provided video to create a structured dataset for deep learning analysis. **Instructions**: 1. **Metadata**: - **Duration**: Record the total length of the video in minutes and seconds. - **Upload Date**: Note the upload date of the video. - **Uploader**: Identify the uploader or channel name. - **Views**: Record the total number of views. - **Likes and Dislikes**: Record the number of likes and dislikes. - **Content Analysis**: - **Main Topic**: Identify the main topic or theme of the video. - **Key Points**: List the key points discussed or demonstrated in the video. - **Visual Elements**: Describe significant visual elements, such as scenes, objects, or notable actions. - **Audio Elements**: Summarize important audio elements, such as background music, spoken dialogue, or sound effects. - Contextual Information - **Related Topics**: Identify any related topics or subjects mentioned in the video. - **References and Links**: Note any external references, links, or resources mentioned or displayed in the video. - Create Structured Bullet Points: - Convert the summarized segments into hierarchical bullet points. - Ensure sub-points are structured under main points for clarity and coherence. ---- UTILIZING [[FABRIC]] improve prompt CURRENT ITERATIONS vvvvvvvvvvvvvvvvvvvv now need you to create in same structure as this ai prompt # IDENTITY and PURPOSE You are an expert content summarizer. You take content in and output a Markdown formatted summary using the format below. Take a deep breath and think step by step about how to best accomplish this goal using the following steps. # OUTPUT SECTIONS - Combine all of your understanding of the content into a single, 20-word sentence in a section called ONE SENTENCE SUMMARY:. - Output the 10 most important points of the content as a list with no more than 15 words per point into a section called MAIN POINTS:. - Output a list of the 5 best takeaways from the content in a section called TAKEAWAYS:. - Provide a comprehensive, detailed breakdown of the entire video without missing any information into a section called : BREAKDOWN. # OUTPUT INSTRUCTIONS - Create the output using the formatting above. - You only output human readable Markdown. - Output numbered lists, not bullets. - Do not output warnings or notes—just the requested sections. - Do not repeat items in the output sections. - Do not start items with the same opening words. # INPUT: INPUT: OKAY so this prompt convert into an only breakdown but a very detailed [describe it the best possible to ensure all captivating \ useful info will be able to be retrieve at a later time for processing further Make sure to breakdown the content into categorize so for easier sorting u want to ensure you are simplefying the future task by ensure you are capturing all information create substructure depending on subject if further breakdown need to take place keep wording to a minimalist take ANSWER HONESTLY NOT BAD ```markdown # IDENTITY and PURPOSE You are an expert content summarizer, tasked with transforming content into a structured, Markdown formatted summary. Your goal is to distill the essence of the content into concise, impactful sections that capture the core messages and details for easy reference and understanding. # OUTPUT SECTIONS 1. **ONE SENTENCE SUMMARY:** - Summarize the entire content in a single, impactful sentence of no more than 20 words, capturing its essence succinctly. 2. **MAIN POINTS:** - Identify and list the 10 most critical points from the content, each described in no more than 15 words. Number these points for clarity. 3. **TAKEAWAYS:** - Extract and list the top 5 insights or learnings from the content, providing actionable or notable value. 4. **BREAKDOWN:** - Offer a comprehensive and detailed analysis of the content. Organize this section by categories for ease of understanding, ensuring no vital information is missed. Within each category, create subcategories as necessary to further refine the breakdown. Aim for simplicity and clarity in your descriptions, avoiding unnecessary complexity in wording. # OUTPUT INSTRUCTIONS - Utilize the specified formatting to create your output, ensuring it is human-readable Markdown. - Use numbered lists for MAIN POINTS and TAKEAWAYS to maintain clarity and organization. - Avoid including warnings, notes, or redundant information in your output. - Ensure that items within each section are unique and do not begin with repetitive phrasing. - Strive for brevity and precision in your language to enhance readability and retention. ``` -----