New generation technology artificial intelligence has recently been used and integrated into products by many companies in many sectors. In this context, the process that started with ChatGPT, which is known to almost everyone, raised the bar higher with Sora. Now, Google Genie, the new artificial intelligence offered by Google DeepMind, has been introduced. What is Genie that can turn an image into a 2D game?Let’s examine what its features are…
New Artificial Intelligence Google Genie Introduced
Developed by Google DeepMind Google Genie artificial intelligenceIt can transform any text, image or idea into a playable form. The research process of this artificial intelligence, which can produce a 2D game, has not been completed yet. According to the statement made by DeepMind, Genie is trained with games and videos available on the internet and is only a research study for now. Let’s explain what exactly is Google Genie…
Latest game news You can take a look here.
Google Genie, It is designed as a new and productive artificial intelligence. Genie, a technology trained on a dataset consisting of a total of 200 thousand hours of videos existing on the Internet, can turn any request such as drawing, text, visual and so on into a 2D game. In fact, this artificial intelligence is proof that game production can also be carried out by artificial intelligence. It turns out that; Artificial intelligence, which is used in many fields, seems to be able to do game development properly in the future.
Genie (Generative Interactive Environments) is a technology developed in collaboration with the University of British Columbia and Google. Genie can produce games similar to Super Mario Brothers and Contra with a single image. As Google introduces DeepMind Genie, “we are introducing generative interactive environments, a new paradigm for generative artificial intelligence: Genie.” He used the expression. In the statement, it was reported that Genie could do similar work in different fields other than producing the games in question.
Genie Can Also Create 3D Worlds
Video Tokenizer, Latent Action Model Genie, who can carry out the necessary production through the Dynamics Model, has learned the movements of the characters, their actions and controls through the 200 thousand videos we mentioned above. In fact, Genie is not trained in action and text, but despite this, she can detect the main character and control him.
Tim Rocktaschel (Google DeepMind developer) said in a statement that their focus is; He notes that the focus is not on adding inductive capabilities to Genie, but on scale. According to some researchers, Genie has the potential to teach 3D universes to other artificial intelligence models. Tim Rocktaschel’s social media platform His post on X is as follows:
While the usage areas of artificial intelligence models are increasing day by day Genie’s Rapid development is expected in this field. Genie’s current abilities are seen as just the beginning. It is obvious that he will be able to achieve interactive and unique works in the future.