We're exploring the potential of the latest video-focused launch from Open AI; Sora.
In the ever-evolving world of video advertising, the rise of AI has the potential to be a game-changer. But up until now, AI tools for the production of video clips have been full of promise for the future - but actually pretty underwhelming right now.
Sure, AI can generate single images quite well, as long as you’ve got plenty of spare time to tinker with the input prompts. We’re already seeing it used in video productions to generate concept visuals, mood-boards, style frames and storyboards, and even for generating finished background images for green-screen shoots. But beyond that? It’s still pretty “Meh” for actual video clips: plagued by jittery glitchy motion, weird artefacts, and things randomly popping in and out of the scene.
At the moment it’s almost always easier and less time consuming to find an existing piece of stock from a library than it is to try and find the miracle prompt that generates the correct image or video that you need.
However, hold onto your hats… because the future is looking seriously cool, especially with the emergence of a new kid on the block: Sora, from Open AI.
Traditionally, creating killer video content involved a lot of logistics with crew, kit, time and money. Not to mention endless cups of coffee. But AI might just change this faster than you can say "lights, camera, action!" - especially for certain kinds of video productions.
What is Sora?
The demo version from Open AI (the makers of Chat GPT and DALL-E) was released to much fanfare in Feb 2024. It’s a new kind of AI model that can create realistic and imaginative scenes from text instructions. It can also create clips up to 60 seconds long. Previous AI video generators have struggled to get smooth results with clips any more than a few seconds long.
The examples released over on the Sora website really do speak for themselves:
With Sora, it looks like the lines between reality and fiction will blur, as it can create lifelike footage with ease, even with flowing camera moves, realistic lighting and reflections. It can also take an existing image, and make a video from it. Something I can see many useful applications for, such as:
Lower budget or quick turnaround video productions
E-learning
Corporate videos
As a replacement for stock video in social media videos
For creating motion-graphics backgrounds
It’s going to be a real game changer!
As to whether it can go so far as to replace the need for higher end bespoke shoots with real human talent remains to be seen, and is probably unlikely at this stage.
But while, right now, AI might feel a bit like your grandma trying to use Snapchat – well-meaning but a tad clumsy -with Sora leading the charge, the imminent future of AI in video production is now looking more promising than ever.
Get in touch with us to discuss how video can help your business.
Producer and Head of Broadcast
