Google Unveils Genie, Its Video Games Generating AI

0
347

There is a new revolution in the game design landscape with the newly launched Google Genie Artificial Intelligence (AI) tool.

Google Genie is a Big Deal

Google DeepMind has just unveiled a new AI tool that is capable of creating playable virtual worlds from a single image. This Genie introduction is yet another way US big tech giants are registering their dominance in the AI sector, weeks after NVIDIA’s “Chat With RTX” showoff.

Google Genie, as it is called, is able to identify and understand a text prompt, sketch or idea and then translate it into a virtual world where a user can interact with and play. Rowan Cheung, founder of the world’s fastest-growing AI newsletter The Rundown AI, acknowledged that the launch of Google Genie is a big deal for the popular technology company. 

For now, the AI model has been trained on gameplay as well as other videos although a huge percentage of these games are 2D platformers rather than full Virtual Reality. In the meantime, these games are still on research preview. According to Cheung, Genie is trained on 200,000 hours of unsupervised public Internet gaming videos. 

He noted a unique feature about the AI foundation model which is the fact that it can recognize the main character in a game even without being trained on action or text annotations. With this, a user can easily control the character in a generated world. This is through the action of its Latent Action Model, Video Tokenizer, and Dynamics Model, as listed on X.

Google is Making Up for Gemini’s Flaws

The launch of Genie may be Google’s way of making up for its flaws on Gemini, its previously released AI chatbot. A couple of flaws were discovered on Gemini after Google CEO Sundar Pichai teased its capabilities in a demonstration video that sparked some reactions. Many experts felt that the content was doctored and then there was the issue of racial inclusion.

Consequently, the AI tool received several backlashes from top players in the industry. Charles Hoskinson also joined prominent voices to condemn Google’s Gemini AI for its controversial image generation feature. He called it “racist” and a detriment to society. 

Google seems to be making a lot of effort to salvage all of these crises associated with the launch of Gemini. The launch of Gemini 1.5 was also another proactive way the tech giant is making up for Gemini 1.0 Ultra flaws.

Source link