OpenAI Introduces Its Text-to-Video AI Model – Sora!

The world of Generative Artificial Intelligence (GenAI) is blowing up!

What started with using simple text prompts to generate content, soon grew into generating images, music, videos, code and more.

Most recently, technology leader Google, which had initially launched its GenAI chatbot Bard, rebranded the platform to Gemini as it launched Gemini 1.0 with more capabilities. Recently, it launched the latest version, Gemini 1.5, to further its AI offerings.

Prior to this, Microsoft followed a similar pattern when it launched its chatbot Bing Chat and eventually renamed it Copilot.

OpenAI, the company that originally created all the buzz and headlines for GenAI’s capabilities with ChatGPT, recently conveyed its intention to further improve the AI world, this time by enhancing its hardware capabilities.

Having garnered 1 million users in just 5 days and 100 million users in just two months after launching ChatGPT, the company now has its sights set on new horizons.

So, what is OpenAI bringing now? Let’s explore!
 

What Did OpenAI Announce?

 

• In an announcement made through an X post (formerly a Twitter tweet), OpenAI introduced their new product, Sora, its text-to-video model.

• According to the post, “Sora can create videos of up to 60 seconds featuring highly detailed scenes, complex camera motion, and multiple characters with vibrant emotions.”
   

  Source

• The post even included a link to Sora’s webpage, which featured information about the model’s capabilities, safety measures and research techniques.

• Blaring across the webpage’s banner flashed a message boasting the model’s capabilities, which read, “All videos on this page were generated directly by Sora without modification.”

• The webpage featured almost 40 videos interspersed with the model’s information.

 

What Are The AI Model’s Capabilities?

 

• OpenAI mentioned Sora can generate complex scenes with multiple characters, motions and accurate details of the subject and background, as well as understand how the subject exists in the physical world.

• Generated videos can also feature multiple shots within single videos.

• Ahead of this, OpenAI did mention the model’s weaknesses, conveying it could have problems simulating complex scenes and may not understand specific instances of cause and effect, for example, “a person might take a bite out of a cookie, but afterward, the cookie may not have a bite mark.”

• Alternatively, the model could confuse spatial details of a prompt, i.e., mixing up left and right or following specific camera trajectories.

• While the model isn’t being made available to everyone as yet, OpenAI did mention it is essentially in its testing phase, as the company looks to gain “feedback on how to advance the model to be most helpful for creative professionals.”

• As such, the model is “becoming available to red teamers to assess critical areas for harms or risks” and a few “visual artists, designers, and filmmakers”.

• Additionally, the company said it’s sharing its “research progress early to start working with and getting feedback from people outside of OpenAI and to give the public a sense of what AI capabilities are on the horizon.”

 

What Kind Of Videos Did The Sora’s Webpage Feature?

 

• Using the prompt “A stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage. She wears a black leather jacket, a long red dress, and black boots, and carries a black purse. She wears sunglasses and red lipstick. She walks confidently and casually. The street is damp and reflective, creating a mirror effect of the colorful lights. Many pedestrians walk about,” Sora provided the below output:

 

• Another prompt “A movie trailer featuring the adventures of the 30 year old space man wearing a red wool knitted motorcycle helmet, blue sky, salt desert, cinematic style, shot on 35mm film, vivid colors” brought:

 

• Another simple prompt “Photorealistic closeup video of two pirate ships battling each other as they sail inside a cup of coffee” generated:

 

• Another simple prompt, “Reflections in the window of a train traveling through the Tokyo suburbs,” generated:

 

• Another video generated used the prompt “Tour of an art gallery with many beautiful works of art in different styles.”

 

• The webpage featured a range of other generated videos without modification, which displayed the model’s capabilities.

Do you think OpenAI will be able to garner the majority of market share for its video generator given the reputation and popularity it possesses for ChatGPT? Do you think OpenAI’s move will inspire other Gen AI companies to follow suit?
 
Let us know in the comments below!