What is Sora? New OpenAI product stuns users with its ultra-realistic videos

The simulation hypothesis states that our reality is not what it seems. Rather than living in a real, physical universe, it suggests that we are actually living in an advanced computer simulation. Think of it like the movie The Matrix, in which the world that people perceive is just a hyper-realistic digital construction.

The theory goes: if technology reaches a level where we can create lifelike simulations, wouldn’t a more advanced civilization be able to do this already? Well, we may have just reached that point.

OpenAI has unveiled a new artificial intelligence system called Sora that can create stunningly realistic videos from text descriptions. It’s driving the internet crazy, with people claiming they can’t tell the fake results from the real ones.

If you’re getting a sense of déjà vu, it’s because the wild reaction reflects the renewed sense of wonder generated by ChatGPT, the AI ​​chatbot also from OpenAI. The digital helper is now used by more than 100 million users per month, including more than half of UK students.

Will Sora grow up to be just as big? And when can you use it to amaze your friends? Here’s what you need to know about the new AI product everyone is talking about.

What is Sora?

Sora is a new AI diffusion model from OpenAI, the company that has become an example for the modern field of artificial intelligence.

These types of software can create new content that is very similar to the data they were trained on. In the case of Sora, the AI ​​can produce hyper-realistic videos of up to one minute based on the instructions you type. It can also expand videos to make them longer.

The story continues

AI diffusion models are another form of generative AI. They work by ‘destroying’ an image by turning it into indistinguishable pixels, then reversing the process to create a coherent image. One way to look at it is that the AI ​​crumples and smooths out a piece of paper with a scribble on it.

Once trained, a diffusion model can guide users in creating highly detailed images from large amounts of data.

What can Sora do?

In the words of OpenAI: “Sora can generate complex scenes with multiple characters, specific types of movements, and precise details of the subject and background.

“The model understands not only what the user asked for in the prompt, but also how those things exist in the physical world.”

The results are astonishing, to say the least. To show what Sora is capable of, OpenAI shared a series of realistic demo videos, including scenes of moving people and animals, a tracking shot of a vehicle on a dirt road, historical footage of California during the gold rush, CGI red pandas in a petri dish and an interior of an art gallery.

As with other AI graphics, look closely and you might spot some flaws, like a cat with an extra leg or a floating chair that makes you feel like you’re looking at a failed video game.

OpenAI says Sora is trained on data with variable durations, resolutions and aspect ratios, which it claims can result in a more accurate representation of reality.

Meanwhile, Al boffs think it may be connected to Unreal Engine 5, a video game engine designed to build interactive 3D experiences. The massive toolbox provides the core elements, from physics to lifelike lighting and sound design, that help developers create more realistic-looking games.

More broadly, OpenAI believes video generation models like Sora “provide a path to building general-purpose simulators of the physical world.”

We’re waiting for a call from Morpheus at any moment.

Is Sora gone now?

Sora is not yet available to the public. Maybe because we’re in an election year and OpenAI doesn’t want to open a can of worms with a realistic video generator that can be leveraged to create malicious deepfakes.

As such, the AI ​​model undergoes security checks and testing by selected users and experts, including creatives such as visual artists, designers and filmmakers. The goal is to collect feedback and ensure that it cannot be used to create malicious content and sow discord.

As with its other AI tools, OpenAI says it will ban clips that violate its content guidelines, including videos that contain extreme violence, sexual content, hateful images, celebrity likenesses, or the intellectual property of others.

For now, OpenAI is previewing its research to give the public a taste of what’s to come. Or it wants to regain some of the hype it recently lost to rival AI products, including ChatGPT rival Google Gemini and fellow video generator Pika.

When Sora is eventually released, we assume that access may be limited, just as it is for OpenAI’s image generator Dall-E. So you can make a few videos for free and then pay for additional clips.

But since the power-hungry AI comes with a huge energy bill, OpenAI might choose to make Sora a premium feature for ChatGPT Plus subscribers.