Repurposing video content is usually a massive headache that requires hours of tedious transcribing, editing, and design work. It often feels like a slow, painful grind that discourages people from sharing the valuable insights they consume daily. I just saw this incredible post from an AI professional that breaks down a workflow to turn a 20-minute video into a sleek carousel in literally two minutes. This isn’t just about speed; it’s about finally unlocking the massive amount of value trapped in your “watch later” playlist.
The Core Mechanism: The AI Relay Race
The genius of this method lies in how it treats the content creation pipeline as a relay race between two specialized tools. The creator of this workflow realized that trying to do everything in a single interface usually leads to mediocre results. Instead, this strategy leverages Google’s Gemini for the heavy cognitive lifting and a tool called Gamma for the visual execution.
Gemini, particularly with its ability to process YouTube links directly, acts as the “brain.” It watches the video, digests the audio, and structures the information based on a very specific set of instructions. Once the logic is solidified, the baton is passed to Gamma. Gamma is an AI-powered design engine that takes structured text and instantly formats it into slides. By separating the strategy from the design, the original poster has found a way to maintain high editorial standards while eliminating the manual labor of slide layout.
💡 Insight 1: The “Senior Editor” Prompt Strategy
The real magic isn’t just asking an AI to “summarize a video.” The expert points out that generic requests lead to generic, fluffy outputs that no one wants to read. To combat this, the provided prompt uses a sophisticated role-playing technique. It commands the AI to act as an “expert Content Strategist and Senior Editor.”
This is a crucial distinction. By assigning a specific persona, the user forces the Large Language Model (LLM) to adopt a critical, analytical tone rather than a passive, descriptive one. The prompt explicitly demands the extraction of the “Core Thesis” and “Golden Quotes.” This ensures that the output focuses on authority-building material—like data points and hard facts—rather than just listing what happened in the video chronologically. It changes the output from a book report into a strategic asset.
💡 Insight 2: The Power of Narrow Angles
One of the most valuable takeaways from this post is the strict prohibition of generic titles. The author includes a constraint in the prompt that forbids titles like “Summary of [Video Name].” Instead, the AI is instructed to pitch “5 narrow, specific article angles.”
This solves a major problem in content marketing: the lack of a specific hook. A 20-minute video might cover five different topics. If you try to jam them all into one carousel, the message gets diluted. By asking for narrow angles, the workflow identifies specific slices of content—like a specific negotiation technique or a unique mental model—and expands on them. This means a single source video can actually generate multiple distinct pieces of content, drastically increasing the ROI on the time you spent watching the video.
💡 Insight 3: Automated Visual Formatting
The final hurdle in content creation is usually design. Even with great text, staring at a blank PowerPoint slide is daunting. The workflow handles this by utilizing Gamma’s “Social” mode. The savvy professional who shared this guide specifies selecting “Portrait” orientation and “Minimal text.”
These settings are vital. Portrait mode ensures the content occupies the maximum amount of screen real estate on mobile devices, which is where most LinkedIn and Instagram consumption happens. Furthermore, selecting “Minimal text” forces the design AI to be concise, preventing the common mistake of creating “wall of text” slides that kill engagement. Gamma then generates relevant AI imagery to accompany the text, ensuring the final product looks polished and professional without you needing to open Photoshop.
📌 The Workflow and Prompt
Here is the exact step-by-step method the innovator laid out, including the specific prompt you need to copy to make this work.
The Process:
- Select Source: Find a high-value viral video on YouTube.
- Extract: Open Gemini and paste the video link along with the prompt below.
- Generate Design: Copy Gemini’s output, go to Gamma.app, and select “Create with AI” -> “Studio mode.”
- Configure: Choose “Social” -> “10 cards” -> “Portrait” -> “Minimal text.”
- Finalize: Paste the text and hit generate.
The Prompt to Copy:
“Act like an expert Content Strategist and Senior Editor for a leading digital publication. Your goal is to repurpose video content into high-value written pieces that drive engagement.
Analyze the following YouTube video: [YOUTUBE LINK].
Part 1: The Extraction
First, provide a comprehensive breakdown of the video content. Do not just summarize; extract the “meat” of the content using the following structure:
Core Thesis: What is the single most important argument or lesson in one sentence?
Key Data/Facts: List specific numbers, case studies, or hard facts mentioned.
Golden Quotes: Extract 3-5 verbatim quotes that are punchy or profound.
The Framework: If the speaker uses a specific step-by-step process or mental model, outline it clearly.
Part 2: The Angles
Based only on the extraction above, pitch 5 narrow, specific article angles.
Constraint: Avoid generic titles like “Summary of [Video Name].”
Requirement: Each angle must take a specific slice of the content and expand on it.
Format: Provide a catchy headline plus a one-sentence premise for each angle.
Example of a “Narrow Angle”
Bad: How to do Sales.
Good: The “3-Call Close” Technique: Why Most Salespeople Fail in the Follow-up.”
Potential Nuances
While this tool stack is powerful, there are a few things to keep in mind. First, always review the AI-generated images from Gamma. Sometimes they can look a bit too abstract or generic, so swapping them out for real screenshots from the video can add authenticity. Second, verify the quotes extracted by Gemini. While usually accurate, AI can occasionally hallucinate or slightly alter wording, so a quick fact-check against the video source is a good habit to build!
Check the full post for the visual walkthrough.