New text-to-video AI model Sora will unleash creative potential but requires 'extreme accountability'

The new Sora artificial intelligence (AI) program has obvious pitfalls, but experts highlighted the benefits of the new tech while acknowledging the need for greater safeguards and public education on its use. 

"While the rapid advancement of Al video generation is a remarkable technological breakthrough worthy of recognition, we must also acknowledge the profound challenges this innovation presents to the social media ecosystem and information environment as a whole," Jake Denton, Research Associate at the Heritage Foundation’s Tech Policy Center, told Fox News Digital. 

"Consumer-facing platforms are currently unprepared to handle the inevitable influx of hyperrealistic synthetic video content enabled by tools of this caliber," Denton added. 

OpenAI recently announced its new Sora text-to-video model, presenting a new and radical step forward for content creation. The company proudly touted the model’s ability to "generate complex scenes with multiple characters, specific types of motion and accurate details of the subject and background." 

STATE PASSES NEW LAWS RESTRICTING AI-PRODUCED DEEPFAKE CAMPAIGN MATERIALS

The announcement included over a dozen examples that range from a hyperrealistic octopus to a sprouting flower and even a video of a person rolling over in bed — the latter showing the person’s hand detaching from their arm in an example of the current strains the system suffers. 

"The current model has weaknesses," OpenAI wrote on a blog post. "It may struggle with accurately simulating the physics of a complex scene, and may not understand specific instances of cause and effect." 

Critics quickly latched onto the obvious pitfalls of the new technology, such as a more powerful and even harder to distinguish form of deepfakes, but experts tried to focus on the benefits that this technology can provide.

WHAT IS ARTIFICIAL INTELLIGENCE (AI)?

"This is an undeniably important step forward for graphics related generative AI, which has incredible potential," Christopher Alexander, Chief Analytics Officer of Pioneer Development Group, told Fox News Digital. 

"However, the actual value will become more evident in a few months as prompt engineers try to develop specific products," Alexander explained. "Midjourney is incredible for creating images when you first use it, but when you start to get really specific for a set creative project, there is quite a bit of friction."

"I expect to see the same from Sora," he added. "The release of Sora is incredible as a next step in generative AI. I think it will prove helpful for creatives, but will struggle to match the hype."

OpenAI pledged to take important safety steps ahead of making Sora available to the public, such as "adversarially testing the model" and tools to "help detect misleading content" including a detection classifier to flag videos generated by Sora. 

NEW COMPANY COULD AIM TO REPORTEDLY DETHRONE GOOGLE AS THE SEARCH KING

Phil Siegel, founder of the Center for Advanced Preparedness and Threat Response Simulation (CAPTRS), asserted that Sora will create "huge productivity gains" when "used positively," such as developing demonstrations of products or complex health procedures — effectively "how-to" guides on processes and products that might require greater study or preparation and cutting down time.

"That’s why you need guidelines or regulations," Siegel added, arguing that if AI models dominate the tech space they will require "extreme accountability."

"I think there's a lot of positive uses, but I think, you know, with all of these, there's also a lot of negative uses," Siegel warned. "Anything that makes it easier for a good guy or a good person to do something better is going to make it easier for a bad person to do something better."

CLICK HERE TO GET THE FOX NEWS APP

"If they’re going to dominate this, they’re going to have to live with a lot of accountability," he continued. "I guess the question is — clearly people are going to use this stuff to do cool things, and a lot of people will become more proficient at their jobs. They’re going to get a whole new class of creative people that work in business." 

"They will hallucinate," he stressed. "I’m not sure they’re ever going to be able to stop that from happening," at least until predictive text models move away from only working one word at a time and thus limiting the potential for bizarre interpretation.