Creating with Generative AI

Follow along as we dive into the new world of Generative AI.  We’ll create a 1-minute movie trailer for XAPP AI that’s built using a new breed of productivity tools including Google’s Bard for a voice-over script, and MidJourney for generating dynamic images using simple text prompting.

Inspiration created by AI

If you haven’t seen the video that’s captured the imagination of the internet, watch the 1 minute trailer for a fictitious Star Wars movie directed by Wes Anderson – The Galactic Menagerie.  What’s truly amazing is the creator made this by himself using Chat-GPT, and Mid-Journey…in a couple days!

At XAPP AI, we’re always learning, and embracing these new productivity tools.  Inspired by the trailer, we made a 1-minute marketing video using similar tools to produce a compelling story about our company.  Read on to see what we created, and how we did it.

Generating our narration for the video

At this point, everyone has either played with Chat GPT, or is aware.  So what is it exactly?  It’s a Large Language Model (LLM) created using data from the Common Crawl Dataset and others that represents the collective knowledge of the internet in text.  Building off that foundation, users can interact with Chat GPT and get amazing results for many use cases – translating text to another language, summarizing a long passage of text, or generating text based on text prompts from the user describing to the AI what they want returned.

In this example, we use Google’s Bard to generate a voiceover script narrative by Morgan Freeman in the style of his character from Shawshank Redemption.  The Google product has nearly identical features to the OpenAI competitor.  Using the generated script, we can produce a professional sounding narration.

Text Prompts to Images

Similar to Chat GPT, MidJourney uses a Large Language Model to turn text input into useful output.  In this case, images.  

For the opening scene of our 1-minute video, see as we describe the scene in Shawshank Redemption where Morgan Freeman’s character finds a hidden box under a rock by a stone wall near a large oak tree.  

Prior to this technology, you would need to create the image yourself, or search for something that is close.  With the proper input, and experimenting, the results are pretty fantastic.

Creating a simulated Morgan Freeman narration

Using a text-to-speech service by topmediai, we were able to feed the narrator script created from Bard to generate a compelling narration for our marketing video.

Watch the video

Generated Script:

I’ve seen plenty of companies come and go.  One thing is certain.  Organizations that succeed are ones always looking for ways to improve.  

Helping organizations improve their customer satisfaction

Increase employee productivity

And turn leads into customers

If you’re looking for ways to improve your business, I encourage you to checkout XAPP AI.  They have the skill to take your business to the next level.  You can provide your customers with the best possible experience, every time.  And you can do it all while saving time and money.  

So what are you waiting for?  Contact them today, and let them help you build a better business.

Easter Egg

Source Image
Generated from prompt and source image

Perhaps you noticed the image above from the video.  You’re not limited to just random images generated.  You can use a source image the AI will use to base the output from.  Notice on the left, our CEO Pat Higbie.  On the right is the generated image based on this prompt:  Sitting on his couch, interacting with a chat experience from his laptop to book an appointment for an estimate on a new kitchen remodel