Does anyone know about puppy videos made by artificial intelligence? OpenAI’s Sora turns text into photorealistic video

The last two years have seen significant progress in artificial intelligence across the board, from pioneers like OpenAI to the steady progress of Google, which is establishing itself as an AI powerhouse. As we head into 2024, technology companies are ramping up their efforts to capitalize on this momentum. It is worth noting that Google has attracted attention by launching Gemini 1.5, and OpenAI has also launched its latest text-to-video AI model Sora.

As the name suggests, OpenAI’s Sora brings text-based prompts to life by crafting lifelike scenes. Imagine yourself writing a prompt for an imaginative scenario, such as imagining a cat dancing on a spaceship. As the company’s disclosure shows, Sora excels at translating these tips into detailed and authentic videos.

“We are teaching artificial intelligence to understand and simulate the physical world in motion, with the goal of training models to help people solve problems that require real-world interaction,” OpenAI said.

Sora can generate videos up to one minute long and maintain visual quality based on user prompts, it added.

If you visit Sora’s landing page, you can see countless results posted by it. One specific example of drone footage is that the waves can be seen crashing along Cape Garret Beach in Big Sur in a very realistic way. Personally I can’t tell the difference. If you handed me this video without telling me it was produced by artificial intelligence, I wouldn’t know.

OpenAI claims Sora can generate complex scenes with multiple characters, specific actions, and accurate details of themes and backgrounds.

The company also claims that “the model not only understands what users ask for in prompts, but also how those things exist in the physical world.”

A scary future?

As dystopian as it may seem, OpenAI says it’s taking significant steps to ensure the safety of its deployments. The company is working with experts to curb misinformation, hateful content and bias. Additionally, it works on detecting misleading content and can tell whether a video was generated by Sora. In addition, the security technology used in DALL-E 3 also applies to Sora.