Hold Your Sora Horses

Most enterprises should press pause on significant conversations about the text-to-video generator.

Open AI Sora, a text-to-video generator, has captivated the imagination of online AI influencers, sparking a wide range of conversations. From implications for the video creation market to somewhat bombastic societal impacts and everything in between, the Sora development announcement and demos excited many. However, CognitivePath believes enterprise sales and marketing teams should take a wait-and-see approach with Sora.

In short, Sora showed us what is possible. But, the reality is that Sora will not impact 99 percent of enterprises and organizations in 2024, and perhaps even longer. There are various reasons, most of which are available to subscribers below. Hype is one thing, but enterprise use demands quality and real impact.

That being said, we, too, are excited by Sora's long-term potential. We believe it holds long-term promise for individual creators and scaling video for enterprises, and we look forward to experimenting with iterations as they are released. Now, let’s get down to business.

Enterprise Challenges Facing Sora

1) What’s the Use Case?

Technology in search of a problem is not useful. Enterprises need to have a use case to justify investment, training, and process change. Right now, Sora is a technology in search of a need.

Videographers and a significant portion of graphic designers can use likely use the tool, but like text and image generators, is it just an individual productivity tool? Will it democratize video creation to more individuals? Perhaps, but if these are the use cases, then a simple individual license after vetting for cybersecurity, ethical and intellectual property matters will suffice.

Most enterprise sales and marketing executives know creating video is a costly endeavor. It is possible that eventually, Sora and competitive technologies can make video more accessible for a wide variety of needs, including demos, training videos, cost-effective B-Roll, and advertisements.

However, even these uses require Sora to “understand” the contextual business of the enterprise. Translated, Sora must be trained on enough related video footage and related text to render a video that would resonate with specific audiences. For example, Sora can create a video of a woman walking in downtown Tokyo, but can it show a factory worker and robot tandem assembling a bulldozer?

Will the enterprise need to train Sora to make it worthwhile? Does the enterprise have enough visual and text data to train a video AI model?

2) Sora Is Not Available