Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Google's DeepMind is recruiting AI researchers to advance world model development - SiliconANGLE

Jan 07, 2025 - siliconangle.com
Google’s DeepMind is assembling a team of AI researchers, led by former OpenAI researcher Tim Brooks, to advance the development of “world models” that simulate physical environments. Brooks announced the initiative on social media, highlighting the team’s mission to create massive generative models capable of real-time interactive generation. The team will work with experts behind Google’s Gemini, Veo, and Genie models, focusing on scaling AI models and integrating them with existing large language models. World models are seen as crucial for achieving artificial general intelligence and have applications in visual reasoning, simulation, and interactive entertainment.

DeepMind’s Genie model, which can simulate 3D worlds based on text prompts, positions the company as a leader in world model development, though it faces competition from companies like World Labs Technologies, Odyssey Systems, and Decart.AI. World models have potential applications in media, gaming, and training environments for AI systems, but they also raise concerns about job disruption in creative industries and potential copyright issues. A study by the Animation Guild suggests AI could impact over 100,000 jobs in the U.S. film and animation sectors within two years.

Key takeaways:

  • Google's DeepMind is forming a new team led by former OpenAI researcher Tim Brooks to advance world model development.
  • The team will work on creating massive generative models that simulate physical environments and integrate them with existing LLMs like Gemini.
  • World models are seen as crucial for building artificial general intelligence and have applications in visual reasoning, simulation, and real-time interactive entertainment.
  • Concerns have been raised about the potential disruption of jobs in the film, TV, and animation industries, as well as copyright issues related to the training data used for world models.
View Full Article

Comments (0)

Be the first to comment!