AI / Project

Generative AI Video

How It Feels to Be a Cincinnatian - YouTube

Tap to unmute

How It Feels to Be a Cincinnatian Tronik XR

thumbnail-image

Tronik XR1 subscriber

Watch on

How It Feels to Be a Cincinnatian

Generative AI Short Film | Created for Cincy AI for Humans (2025)

Overview

This short was created the morning of my presentation at Cincy AI Week as a live demo of what’s possible using the open-source launcher Pinokio AI. My goal: make something from scratch in 2–3 hours, lean into humor, and reflect the everyday vibe of being a Cincinnati resident — all through the lens of generative video tools.

The result is a satirical, slightly absurd AI-generated film that had the audience laughing mid-presentation. It was also a personal experiment in dropping perfectionism: I finished the final edit 15 minutes before heading to the event.

Process + Tools

• Script: Written manually (morning of the presentation)

• Video Generation: Entirely created in WAN 2.1 via Pinokio

• Voiceover: AI TTS narration based on my script

• Face Refinement: Used Roop to deepfake my face onto my own face — correcting WAN’s render glitches and preserving consistency

• Editing: Finalized in Adobe Premiere Pro

Audience Response

One of the biggest reactions came from the line: “I like to come home and melt into my couch” — timed with a warped deflate effect on my AI-generated body. The imperfections actually added charm. The video resonated because it was relatable, local, and clearly made under pressure — which helped people connect with the creative process behind it.

Tools Used

WAN 2.1, Roop (via Pinokio AI), AI voice generation, Adobe Premiere Pro

Generative AI Video: "Rachel and Leah" - Production Workflow

Project Overview:

This project demonstrates a comprehensive application of generative AI tools to produce a narrative-driven video adaptation of the biblical story of Rachel and Leah. Every stage of the creative process, from script development to final rendering, leveraged AI capabilities.

Production Workflow:

1. Script Development and Scene Breakdown:

• Utilized ChatGPT to create a contemporary adaptation of the Rachel and Leah narrative.

• Instructed ChatGPT to delineate the script into distinct scenes, providing detailed visual prompts and corresponding narration for each.

2. Image Generation:

• Employed DALL-E 3 to generate high-fidelity images for each scene, based on the visual prompts provided by ChatGPT.

3. Video Synthesis:

• Utilized Minimax to transform the generated images into a cohesive video sequence.

4. Facial Consistency and Deepfake Refinement:

• Addressed potential inconsistencies in character facial features introduced during the video synthesis process.

• Implemented FaceFusion AI to apply deepfake techniques, ensuring consistent and accurate facial representations throughout the video.

4. Voiceover Narration:

• Leveraged ElevenLabs to generate realistic and nuanced voiceover narration based on the script provided by ChatGPT.

5. Ambient Audio Design:

• Integrated ambient sound effects (e.g., rain, applause) using MMAudio AI to enhance the atmospheric quality of specific scenes.

7. Subtitle Generation:

• Employed Whisper AI to generate precisely timed subtitles from the finalized voiceover narration, ensuring accessibility and clarity.

8. Final Editing and Rendering:

• Integrated all generated assets (video, audio, subtitles) within a video editing environment.

• Overlaid the generated subtitles onto the video timeline for final rendering, producing a polished and comprehensive video product.