A new wave of generative AI is about to revolutionise and possibly democratise media production, a conference heard.
Bots like Midjourney – which create pictures from text prompts – are being used by an increasing number of production companies.
But the Media Production and Technology Show was buzzing with discussion of the coming wave of text-to-video bots. While talk was centred on AI’s huge potential, there was agreement that human creativity and ingenuity will remain at the heart of media production – at least for now.
Eline Van Der Velden from Particle6 Productions (pictured) told the conference bluntly: “Get involved or the train will leave without you.”
She is already producing a children’s YouTube channel entirely with generative AI.
Companies like Dan Snow’s History Hit are using (openly and transparently) images created by Midjourney to reimagine scenes from the past.
My own company has used Midjourney images as an alternative to expensive stock footage. (See previous blog for the pitfalls!)
But it is clear that, with text-to-video bots like Sora, Vidu and Veo on the horizon, this is just the start. (The video below is a demo clip from Sora).
Creativity unleashed?
Van Der Velden said the potential for making creative and innovative content without the need for expensive filming and post production was enormous.
“This means any individual can make their creative vision come to life,” she said. It could lead to a huge democratisation of film-making.
Alex Hryniewicz from Little Dot Studios said: “There are opportunities for people to do far more than they can currently do.”
Guy Gadney of Charisma.ai thought AI could be used to explore whole new ways of storytelling – which go beyond simply generating video: “AI is a tool you can use in the telling of stories. What new things can we do that we could not do before?”
Dan Snow (pictured), the former BBC presenter and founder of History Hit, predicted that Generative AI could revolutionise CGI and graphics – some of the most expensive parts of the film-making process. He said it could have a similar effect to the availability of low cost, high performance cameras on other parts of the industry.
What will not change is the need for a creative vision. Writing prompts to create amazing content is at the heart of the process. And to get good content you will need to clearly articulate the output you are looking for.
There are potential limits to the technology. Hryniewicz predicted the sheer amount of computing power needed for text-to video would be a constraint on its availability. While text-based LLMs like Chat GPT can make new versions instantly available to all, text-to-video was a much hungrier beast. He feared such constraints could make the new technology expensive and limit its potentially democratising effects.
Gadney pointed out that the text to video bots had all only provided very brief clips, and they were a long way from telling stories. They are also technical demos, not necessarily things that will be easy for users to recreate. The days of creating Hollywood movies from text-to-video are a long way off.
He predicted that the immediate future of these bots was in the short form social media sphere, where the bots could create easy to use clips.
Clips like the one below could certainly be used in place of stock footage and, with skilled prompt drafting, clips could be much better tailored to the needs of a particular client needs.
To discuss how your company could use generative AI for low cost promotions get in touch.
Note: The cover image on this post was created by Microsoft Bing’s Co-Pilot AI. The prompt was: “Create an abstract image of AI and the world of neural networks.”
Rough Cut Media Ltd 
One thought on “Democratising Filmmaking: The Rise of Generative AI and Text-to-Video Bots”