Google never ceases to amaze with its innovations in the field ofartificial intelligence (TO THE). The latest news is called Geniean AI model created by the department DeepMind which has the ability to transform any uploaded image into a video game.
A tool that could revolutionize the world of video games, making it possible create virtual worlds starting from a simple image, such as a photograph or a sketch. These worlds are then playable and interactive, and offer a variety of scenarios and challenges.
How Genie works, Google’s AI that “creates” video games
The operation is relatively simple: the user uploads an image (be it a photo or a drawing) and the AI ”invents” a virtual world with which it is possible to interact.
Tim Rocktäschel, leader of the team that developed Genie explained on X/Twitter that the model is based on a huge dataset, composed of more than 200 thousand hours of 2D platformer videos. Genie learned autonomously, without supervision, to generate movements, controls and actions for the characters in the virtual worlds it created.
I am really excited to reveal what @GoogleDeepMind‘s Open Endedness Team has been up to 🚀. We introduce Genie 🧞, a foundation world model trained exclusively from Internet videos that can generate an endless variety of action-controllable 2D worlds given image prompts. pic.twitter.com/TnQ8uv81wc
— Tim Rocktäschel (@_rockt) February 26, 2024
An AI model that therefore arises from intense work by gameplay training and other videos found online. However at the moment Genie is not yet accessible to the public, but only to researchers in the beta phase. However, there is an official website, where it is possible to take a look at the video games created by the model. Of course, don’t expect ultra-realistic games from triple-A studios: at the moment Genie can only create simple 2D platformers but, as the AI slogan states, enough “one image to create a new world”.
Stay updated by following us on Google News!
Follow!