How to Create the Viral Talking Animal AI Podcast Videos on Autopilot (No-Code n8n Tutorial)

Key Concepts

• AI-generated podcast-style videos
• NADN (low-code/no-code automation platform)
• API keys
• Prompt engineering
• 11 Labs (text-to-speech)
• HRA.io (video generation platform)
• OpenAI GPT (image generation)
• Workflow automation
• Dynamic variables
• Social media automation

Generating Viral Podcast-Style Videos with AI

The video demonstrates how to automatically generate viral podcast-style videos for various niches (e.g., baby podcasts, dog podcasts, historical figures) using AI automation. The process involves selecting a category and topic, and the AI generates the video content.

Examples:

• AI-generated podcast featuring babies.
• AI-generated podcast featuring animals (squirrels).

Success Case:

• The video highlights a content creator who shifted from music and comedic videos to AI-generated podcasts with babies and animals, achieving significant views (e.g., 4.2 million, 11.2 million).

Using NADN for Automation

The core of the process relies on NADN, a low-code/no-code platform.

Accessing NADN:

• A link to NADN with a 14-day trial is provided in the video description.

Importing Automation Blueprints:

• Lazy Method (Easiest): Importing a pre-built blueprint from the community.
  • Download the blueprint file from the community resources.
  • In NADN, click the three dots and select "Import from file."
  • Select the downloaded blueprint file.
• Step-by-Step Method: Building the automation workflow from scratch (covered in another video).

Customization:

• The primary customization involves changing API keys and adjusting prompts for different niches.

Workflow Demonstration: Animal Podcast Generator

The video showcases an animal podcast generator as an example.

Form Submission:

• A form is used to collect input parameters:
  • Animal type (e.g., dog, squirrel)
  • Mood (e.g., whiny)
  • Problem topics (e.g., trees)

AI Processing:

1. Prompt Generation: Based on the form inputs, a prompt is generated for the audio and image.
2. Audio Generation: 11 Labs is used to generate the audio from the text prompt.
3. Image Generation: OpenAI GPT image model generates an image based on the prompt.
4. Video Generation: HRA combines the audio and image to create the video.
5. YouTube Upload: The generated video is automatically uploaded to a YouTube channel.

Example Output:

• A video featuring a squirrel complaining about trees.

Step-by-Step Breakdown of the Automation Workflow

The video provides a detailed look inside each node of the NADN workflow.

1. Form Submission: Triggers the workflow based on user input.
2. Generate Prompt: Uses GPT-4 to create a podcast-style rant prompt based on the selected animal, mood, and topic.
   • Example prompt: "Generate a podcast style rant for the animal type [animal type] who's feeling [mood] and it's complaining about a particular topic [topic]."
3. 11 Labs: Generates audio from the generated text prompt.
   • Requires a voice ID.
4. HRA (Create Audio Asset): Creates an audio asset in HRA using the audio generated by 11 Labs.
5. Merge Audio Data: Combines the ID from HRA and the file from 11 Labs.
6. HRA (Upload Audio): Uploads the audio asset to HRA using the API key, file, and ID.
7. OpenAI (Image Generation): Generates an image based on the dynamic variables from the form submission.
   • Example prompt: "Generate a high-resolution image of the animal type [animal type] appearing in this tone asking or speaking into a professional podcast microphone with the background and all that stuff."
8. Convert to File: Converts the B64 JSON string from OpenAI into an image file.
9. HRA (Create Image Asset): Creates an image asset in HRA.
10. Merge Image Data: Combines the ID and image file.
11. HRA (Upload Image): Uploads the image asset to HRA.
12. HRA (Call Model): Calls the HRA model to combine the audio and image.
    • Uses the native HRA 3 model.
    • Requires the AI model ID, image ID, and audio ID.
13. Wait Node: Waits for 5-7 minutes for the video to be generated.
14. HRA (Get Video): Grabs the generated video from HRA.
15. Wait Node: Waits for 10 seconds to ensure processing is complete.
16. Download Video: Downloads the video.
17. Google Drive (Upload & Download): Uploads the video to Google Drive and then downloads it again.
18. YouTube Upload: Uploads the video to YouTube.
    • Privacy status can be set to "unlisted" or "public."

Customizing for Different Niches

The video explains how to adapt the workflow for different niches.

Form Modification:

• Add a new node (Form) to generate a new trigger.
• Modify the form elements to include relevant categories (e.g., historical figures).
• Change the title and add form elements as needed.

Prompt Engineering:

• Use ChatGPT or Claude to generate new prompts for different niches.
• Copy and paste the generated prompts into the OpenAI node.
• Adjust the prompts to include dynamic variables from the form.

Example:

• For historical figures, use a prompt like: "Generate a podcast-style rant for the historical figure [historical figure] who's feeling [mood] and is complaining about [topic]."

Social Media Automation (Future Topic)

The video briefly mentions future content on automating social media posting.

Botto API:

• The Botto API can be used to post to multiple social media platforms (Instagram, LinkedIn, YouTube, TikTok, Blue Sky, Threads, Twitter) simultaneously.

Planned Tutorial:

• A future tutorial will cover generating podcast-style videos and automatically posting them to various social media platforms.

Conclusion

The video provides a practical guide to generating AI-powered podcast-style videos for various niches using NADN, 11 Labs, HRA, and OpenAI. It emphasizes the ease of use through pre-built blueprints and the flexibility to customize the workflow for different content categories. The video also teases future content on automating social media distribution for these AI-generated videos.