Speaking during a panel hosted by the BFI London Film Festival looking at the impact of Generative AI on the creative industries, Framestore chief creative officer Timothy Webber discussed how the technology is already changing the “whole workflow”.
While the technology works well with still images, Webber says using it to create video is “much, much harder,” adding that the biggest issue is temporal consistency.
Framestore is using Generative AI in the “early stages” of content creation, such as exploratory concept art to help creatives get feedback on their ideas, but “there’s still a huge amount of human input,” added Webber.
“Generative AI can help speed things up, and during the ‘ideation’ phase it can be very, very useful. But in the final image phase, very challenging,” he continued.
“We’re sort of using it in a small way in the very early days of the process, and it will gradually work its way through to the end. Beyond that, I am absolutely loathe to predict anything very far in the future to do with AI due to the speed of change.”
However, Webber did say he thinks it will be “more than five years” before Generative AI creates images viewers see on their screens.
“My feeling having worked with Generative AI for a while is it is neither quite as revelatory and amazing as we all thought it might be in the early days, nor is it quite as scary and threatening,” he added.