James Cameron discussed using artificial intelligence (AI) in producing films. The award-winning director admitted that while he loves the technology, it also leaves him scared.
James Cameron on Using Artificial Intelligence, Technology in Producing Films
Cameron sat down with CNN's Fareed Zakarita, and they talked about the use of AI in the film industry and how it impacts society. The "Avatar" director admitted he is "a little more scared than excited" about it.
Cameron said they used small AI tools in developing "Avatar" to speed up their workflow and increase the accuracy of their facial pipeline in taking the actors' performance and translating it into computer-generated imagery (CGI).
However, he warned the people that "we can't trust what we see" due to deepfakes and the like. Deepfake AI is a technology that produces credible audio, video, and image frauds. The phrase combines deep learning and fake, covering both the technology and the phony information that results from it.
Deepfakes frequently alter existing source material by switching out one individual for another. Additionally, they produce wholly unique content in which individuals are depicted doing or saying things they did not do or say.
Deepfakes' capacity to disseminate misleading information that looks to come from reliable sources poses the biggest threat. For instance, a deepfake film portraying Ukrainian President Volodymyr Zelenskyy pleading with his soldiers to surrender was made public in 2022.
According to Cameron, it's more difficult to trust our sources due to technology, and it will be harder as we go along because the "looks are compelling." Thus, it might be more challenging to determine the sources in the future unless one is physically present.
He quoted Socrates and the latter's idea of reality in Plato's "Allegory of the Cave." Cameron said that we won't be able to tell if our feeds are accurate in the future.
How Deefakes Work?
Deepfake uses two algorithms to produce and hone fake content - a generator and a discriminator. The discriminator evaluates how realistic or false the original version of the material is while the generator creates the initial fake digital content using a training data set depending on the desired outcome. Repeating this procedure enables the discriminator and the generator to identify better defects that the generator should fix.
A generative adversarial network is produced when the generator and discriminator algorithms are combined. In order to construct the fake images, a GAN first employs deep learning to identify patterns in real photographs.
A GAN algorithm views photos of the target from various viewpoints to gather all the details and perspectives while constructing a deepfake image. The GAN analyses the video from several perspectives while examining behavior, movement, and speech patterns to create a deepfake. The final image or video's realism is adjusted by repeatedly passing this information through the discriminator.
There are two approaches to making deepfake videos. They can use the target's actual video source to make them say and do things they would never normally do, or they can face-swap the target's face onto the face of another person in a video.
RELATED ARTICLE: Teenage Hacker In Control of Over 20 Tesla Vehicles in Several Countries Using Software to Operate the Cars Even Without Keys
Check out more news and information on Technology in Science Times.
