OpenAI has been showing off Sora, its artificial intelligence video generation model, to media industry executives in recent weeks, building enthusiasm and allaying concerns about the technology's potential to be disruptive to certain sectors.
The Financial Times wanted to test Sora and its rivals.
We asked advertising, animation, and real estate executives to create prompts for generating videos they might use in their work. We then asked them for their views on how such technology might change their jobs in the future.
Since Sora is not yet publicly available, OpenAI adjusted some of the prompts before submitting the resulting clip. For Runway and Pika, initial and fine-tuned prompts were populated using state-of-the-art models from both companies. The results are as follows.
Charlotte Bunyan, Arq Co-Founder, Brand Advertising Consultant
OpenAI's revised version of Bunyan's prompt to create a campaign for a “famous high street supermarket”:
Pike and Runway video based on Bunyan's original prompt:
“Although the representation of Sora's people was consistent, the actual visualization of the fantasy playground was faithfully rendered in terms of illustrating various elements that others could not produce.
“It’s interesting that OpenAI changed “children” to “people,” but I’d love to know why. Is that a safety question? Is it difficult to represent the children because they don't have as much training? They chose “the people” instead of white men with beards and brown hair. This is what Sora actually produced, and it raises questions about bias.
“Pika felt surreal, like you were in a trippy moment in a movie. The children's version is much better than that. league of gentlemen Although it has the surreality of an adult iteration, the rest of the environments lack prompt detail.I have a certain preference for its liveliness [Pika’s children’s] This is the version that conveys a stronger sense of joy and happiness than any other version. ”
“Runway was pretty much in the middle. Sure, there were fewer glitches in the adult version, but there was a lack of representation of the playground elements.
“The Sora video could be used as a test drive for what can be achieved with a virtual experience. there is.
“These tools speed up the way we communicate creative ideas and make them more tangible. For example, in the early stages of presenting a concept to a brand, this allows the client to understand what it could look like. It will be easier to understand what it is and how it works.
“My prompts involve abstract creative concepts that are difficult to use with these tools. Often in the world of creativity, we try to create something that has never existed before. I know there are a lot of concerns and naysayers about taking away all our jobs, but I think we should think about how AI can make our jobs easier and lessen our burdens. think.”
Alex Williams, animator, whose credits include “Who Painted Roger Rabbit?” And “The Lion King”
Video generated by OpenAI's revised prompts:
“AI-generated work has a slightly morphic quality that I don’t think is client-ready yet, but that will improve over time.
“Each feature is great, but each [has] Obvious mistake. . . The head shape changes and the flamingo blends into other flamingos, but it's not working yet. ”
“We weren't able to make a short film with a beginning, middle, and end, and it didn't turn out to be what we expected. On the other hand, it's very impressive in terms of animation.
“Since I started animating in the 80s, very important technological advances have changed the medium of animation. Without a doubt, this is the biggest change in my career.
“I compare it to the switch from 2D to 3D animation that happened in the late 90s. toy story It came out. Initially, there was a lot of resistance to these changes among the hand-drawn animation community, myself included.
“It took me a few years to realize that I had to embrace this change. For a while, we all came together to fight this issue, and it became a huge driver of box office success. Never Wrong We need to embrace technology as an industry because we don’t want to take sides.”
Ashley Shakibai, Production Manager at commercial real estate agency OBI Property
Video generated by Manchester Commercial Building Promotional Video Prompt:
“Sora did a decent job at first. Transitions are always difficult and we struggled with it. But I think the photorealism at the end of the shot was very pleasing and surprising.
“Technically, there was a prompt for people walking through the building, but it wasn't showing and there were a lot of other elements that weren't being accomplished.
“All Pika gleaned from my prompt was 'a clear day'. I could see flares and some buildings, but no people.”
“I had to laugh when I saw this runway. There's a little more photorealism, but it's certainly not a believable scene because people walk forward and sometimes they walk backwards.
“As an industry expert, what I expect is perfection. I'm looking for realistic quality video, and AI will probably never quite get there.
“At the end of Sola's video, the couple are happily conversing in a coffee shop. It's a shot they use to sell nearby commercial real estate space as an amenity.
“Eventually, we will reach a point where this becomes such a powerful tool for creators that we will inevitably no longer need to use other tools. I will seriously take on the challenge of taking on this role.
“You have to be very careful when adding computer-generated images. If that's not your goal or you don't believe it, it can be too distracting. It's really just a testing phase.”
Additional reporting by Madhumita Murgia
