OpenAI Releases Its AI Video Tool
Bi-weekly news commentary
Hey friend,
How are you? I hope you’re doing well and looking forward to closing the year.
As you plan your visual storytelling toolset for 2025, here is a new tool to add to your list.
The gist:
A few days ago, OpenAI publicly released its AI video tool, Sora. If you recall, I’ve recently covered The Landscape of AI Video Tools and included a story about how Toys R Us created the first brand film using Sora. The video generation model is designed to take text, image, and video inputs and generate a new video as an output. Yet, with the insane demand to try the tool, OpenAI has blocked access for a while. Checking it out early Sunday morning, I was able to access it!
Find out why it’s important, my first Sora video, and a few tips about how you could use Sora in your video storytelling strategy.
Why it matters:
As we enter the AI video age, where you can simply type a brief description and get super realistic video content, we’ve also entered a post-truth age that can work both ways for good and bad actors.
Based on people who tried Sora, there are still a few deformities in the outputs that would require you to uproot them, likely with other tools.
However, as AI video technology improves, I sense that there will be a higher demand for visual storytellers who stick to honest and relatable stories to cut through.
The truth will be a differentiating selling point. Who knew?
This is especially true these days with the rise of disinformation and deepfakes. If you look around, people already find it hard to figure out what is real and what isn’t.
If you’re a ChatGPT Plus subscriber, you can produce a 5-second video that can function as a scene and then stitch it with another 5-second video using the Storyboard tool.
But here is the rub: Sora still can’t handle consistent continuity from one 5-second video - say, a cat wearing a red hat moving left to right - and keep the same cat in subsequent scenes.
As such, you’re likely to be using Sora for standalone social content and b-roll video content in your explainer or training videos.
Sora, along with all other AI tools that are now available to creators, means the creative floor is quickly rising, and with it, I foresee new emerging aesthetics - we can’t even imagine now.
What you can do:
First off, the tool is available for ChatGPT Plus subscribers to produce up to 5-second video (50 monthly videos at 720p resolution) and for ChatGPT Pro - a 20-second video (500 monthly videos at 1080p resolution).
With the latter subscription, you can also download videos without the watermark.
My first Sora video
Prompt: “A pink horse with a rider dressed as a cowboy with a trench coat and hat, riding a snowy mountain.”
Once OpenAI finally opens access to the tool, I’d pick a specific use case you want to test.
You can start simply with a video meme to convey a single idea around your brand. And then scale to a longer video stitched with several 5-second video clips using the Remix, Blend and Storyboard tools.
You can control the style, and I’d even urge you to try prompting your HEX#s of your brand colors to see what kind of result you get.
As I mentioned earlier, the output may not be perfect. You may need to fix it by several prompt iterations or using other tools like Canva’s Magic Eraser.
Always check T&Cs for proper use of AI-generated content to ensure you comply.
Lastly, AI video tools should not replace your traditionally made videos. Each has its purpose. I’d say imaginative vs. authentic content.
Happy experimenting :)
See you next time!
Best,
- Shlomi
Shlomi Ron
Chief Storytelling Officer
Visual Storytelling Institute
story > visual > emotion > experience
