Google announces Veo, its AI for generating videos

This Tuesday (14) the Google announced with great enthusiasm new tools for Artificial Intelligence (AI) that have the potential to revolutionize the way we create and consume multimedia content. These tools are designed to generate videos and photos in impressive resolutions, providing visual quality that comes closer and closer to reality.

One of the highlights of this announcement was the I see, a powerful tool that allows you to create high definition videos. Veo is directly competing with Sister yes OpenAI, which shows how committed Google is to establishing itself as a leader in this segment. Additionally, the company has also brought significant improvements to its imaging AI with the launch of Image 3expanding the possibilities for generating detailed and realistic images.

Read too
Google I/O: Project Astra is artificial intelligence that understands the environment through cell phone cameras
OpenAI reveals music clip created using just Sora

More details about Veo

These advanced AI models are initially only available to a select group of creators, allowing them to explore and experiment with new features before they are released to the general public. Although we do not yet have a specific date for this broader access, it is certain that many are eagerly awaiting the opportunity to use these incredible tools.

Regarding Veo, the tool demonstrated the ability to create videos lasting more than one minute in 1080p resolution. Furthermore, Veo displayed versatility by being able to imitate different cinematic styles, which suggests a notable advance in the personalization of audiovisual content created by AI.

According to Google, one of Veo’s biggest advantages is its ability to accurately understand both natural language and visual content. This translates into the ability to execute long, complex commands, enabling the creation of videos that specifically meet the creators’ demands and artistic vision.

The company ensured that Veo is capable of generating incredibly realistic scenes, where people, animals and objects exhibit natural and convincing movements, expanding the range of possibilities for producing dynamic and engaging content. The tool also understands technical terms, including “timelapse” and “aerial recording,” which gives users greater creative control over projects.

Filmmakers and content creators had the chance to interact with Veo, testing its features and providing feedback for future improvements. Google has announced plans to integrate some of Veo’s features into other platforms, such as YouTube Shorts, although it has not yet set a clear timeline for these implementations.

For those who want to see what Veo can do first, it is now available to select creators. They have access through VideoFX in a private preview on the website aitestkitchen.withgoogle.com. You can join a waiting list. So, when you have space, you will be notified to test this new technology.

New model for creating images

Furthermore, at Google I/O 2024, people saw news about the Imagen system, which is now in the Image 3. This new version of image AI has improved quality, better understands what we say and details images more accurately.

With Imagen 3, you can now add text to images with unprecedented ease. This opens up a new range of creative possibilities, such as creating personalized birthday cards, presentation slides with inventive titles, posts for social networks, among other applications suggested by the company itself.

Google has high expectations for producing images that not only look extremely real, but also have a wealth of detail with fewer flaws than in previous versions of the model. The company indicated that Imagen 3 will soon be available on Vertex AI, which is its specialized AI services platform.

Meanwhile, the new edition of Imagen can now be explored by some creators through ImageFX. Just like with Veo, there is the possibility of signing up to participate in the tests and have the opportunity to try this new feature from Google.