When the original Sora generative AI video model was released last year, it kind of set off an existential crisis in the film and video world that we’re pretty sure no one has recovered from. And while AI hasn’t taken over just yet, OpenAI’s announcement of Sora 2 isn’t going to calm any fears—whether they’re rational ones or not.

Sora 2 is the latest video generation model developed by OpenAI, and it promises to be more physically accurate, realistic, and controllable than prior systems. Sora 2 will also feature synchronized dialogue and sound effects and a host of upgrades to its performance and controllability that, indeed, will make it more appealing to aspiring AI filmmakers or other artificial artists.

Again, love it or hate it, generative AI video is happening, and it’s worth covering so that you know what’s going on in this ever-changing industry. So, without further ado, here’s everything you need to know about Sora 2 that will absolutely wreck your day.

OpenAI Introduces Sora 2

If you follow AI at all, this might hit (if not, maybe scroll a bit lower), but OpenAI is liking this upgrade from the original Sora to Sora 2 to potentially be as impactful as the jump from GPT-1 to GPT-3.5. Basically, OpenAI is trying to say that this is a huge leap forward not only in performance, but more so in overall intelligence.

Sora 2 promises to be able to do things that the original Sora, or any other generative AI video models, could dream of in terms of difficult video generation requests. OpenAI lists some complex maneuvers, for example, such as “Olympic gymnastics routines, backflips on a paddleboard that accurately model the dynamics of buoyancy and rigidity, and triple axels while a cat holds on for dear life.”

Prior AI video models will seem oversimplistic compared to Sora 2’s reasoning capabilities, which promise to be able to understand more of the greater nuances of the physical world, which ultimately will lead to more realistic results that can look cinematic and, above all, more real.

Sora 2 Video-Audio Generation

The other big news here is that Sora 2 will be capable of creating sophisticated background soundscapes, speech, and sound effects with a higher degree of realism as well. Sound and audio have been one of the key characteristics that AI video has lacked so far, and have honestly been a much easier way for many to spot AI in the wild.

All that might change now with Sora 2 as OpenAI reports that the model can insert audio into any Sora-generated environment with an accurate portrayal of appearance and voice. However, we’ll have to see how well this actually performs in the coming weeks and months before we crown Sora 2 as the scary full package it promises to be.

Sora 2’s Availability

Sora 2 is out and available now, but that’s not all. OpenAI is also launching a new social iOS app called “Sora” that will be powered by Sora 2, which will allow AI creators to create, remix, and share each other’s generations in a new customizable Sora feed.

We’ll have a lot more to explore about Sora’s social app, as well as much more to explore about the models' AI filmmaking possibilities here soon. But for now, if you’re curious, you can try out Sora 2 yourself on OpenAI’s website here.

From Your Site Articles
Related Articles Around the Web
openaisoraopenai sora 2ai videoai video modelgenerative aisora 2
AITopics