Google’s Genie: A New AI Model That Can Generate Video Games with Simple Prompts

Feb 27, 2024 | 0 comments

Google's Genie

Google DeepMind has launched another AI model known as Genie. Google’s Genie helps users generate interactive, playable video games with simple images and text input prompts. Users can use hand-drawn sketches as well as real-world photos as input prompts.

Google’s Genie seems very cool, as it can take our imaginations to the next level with video games. The model was mainly trained on videos of 2D platformer games and robotics. Google’s Genie is trained on 2,00,000 hours of unsupervised public internet videos. So the model trained itself without any human supervision.

Google’s Genie can also determine who the main character is and enable users to control it in the generated virtual environment. 

How Google’s Genie Works

The model can generate interactive environments with simple images, photographs, and sketches. Genie has a size of 11 billion parameters. It utilizes a video tokenizer, a latent action model, and a dynamic model to generate interactive video games. 

Genie utilizes a novel video tokenizer that processes huge amounts of video data into small tokens. It then uses a causal action mode, which analyzes transitions between the tokenized video data to extract latent actions that the user can take in the game. These actions can be anything from moving up and down to jumping or running with objects in the virtual environment. It is then finally passed to an autoregressive dynamics model, which predicts the next frame of the game based on the user’s actions and video sequence. 

Ethical Considerations with Google’s Genie

With every AI model, security is always a concern, and Genie can also be used for bad practices as it can turn anyone’s imagination into an interactive environment. So Google is continuously taking steps to ensure responsible use of this new AI model.

Google Genie Availability

Genie can currently generate games at a 1FPS frame rate. It is a research-only model for now and is not available for public use.

Conclusion

Every day a new AI model is showcased, and it really makes us think about how much more AI can do. With Google’s Genie, you can even generate video games without doing much of the hard work that goes into generating games. Although it is very new and isn’t able to generate more extensive games with good visual quality, it is certainly a big step in that direction and in game development with AI.

Read More

Stable Diffusion 3 is here to beat DALL-E 3 and Midjourney

Humane’s AI Pin Shipping Gets Delayed: Early Adopters Get 3 Months Free Subscription

RECENT POSTS

Elon Musk’s xAI Announces Grok 1.5 with Great Capabilities

Elon Musk’s xAI Announces Grok 1.5 with Great Capabilities

Image Credits: xAI Elon Musk's xAI launched Grok last year in November to compete with chatbots from big tech giants like Google, Microsoft, and OpenAI. Elon Musk's xAI is soon launching the next version of their chatbot, Grok 1.5, which performs really well as...

Meta’s Ray-Ban Smart Glasses Are Getting New AI Features

Meta’s Ray-Ban Smart Glasses Are Getting New AI Features

Image Credits: Meta Meta’s $300 smart glasses, made in collaboration with Ray-Ban, allow users to take pictures, record videos, make calls, hear music, and do much more. Now, new AI features are being added to Meta's Ray-Ban smart glasses.  New AI Features in Meta’s...

Claude 3 beats GPT-4 for the First Time on LMSYS Leaderboard

Claude 3 beats GPT-4 for the First Time on LMSYS Leaderboard

Anthropic released the Claude 3 model family earlier this month, and they have become highly popular since their release. Now Anthropic's Claude 3 Opus Model beats OpenAI's GPT-4 model for the first time on the LMSYS Chatbot Arena Leaderboard. LMSYS Chatbot Arena is a...

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *