""SIMA 2 is a step change and improvement in capabilities over SIMA 1," Joe Marino, senior research scientist at DeepMind, said in a press briefing. "It's a more general agent. It can complete complex tasks in previously unseen environments. And it's a self-improving agent. So it can actually self-improve based on its own experience, which is a step towards more general-purpose robots and AGI systems more generally.""
""Working with so-called "embodied agents" is crucial to generalized intelligence, DeepMind's researchers say. Marino explained that an embodied agent interacts with a physical or virtual world via a body - observing inputs and taking actions much like a robot or human would - whereas a non-embodied agent might interact with your calendar, take notes, or execute code.""
""SIMA 2 is powered by the Gemini 2.5 flash-lite model, and AGI refers to artificial general intelligence, which DeepMind defines as a system capable of a wide range of intellectual tasks with the ability to learn new skills and generalize knowledge across different areas.""
SIMA 2 integrates a Gemini 2.5 flash-lite model with embodied-agent training to produce a more general, self-improving agent capable of handling complex tasks in unfamiliar virtual environments. Prior training on hundreds of hours of video-game data enabled earlier generalization across multiple 3D games. SIMA 1 could follow basic instructions but achieved only a 31% success rate on complex tasks versus 71% for humans. Embodied agents perceive and act via a body, enabling richer interactions than non-embodied agents that manipulate digital tools. SIMA 2 advances toward more general-purpose robotics and broader AGI capabilities.
Read at TechCrunch
Unable to calculate read time
Collection
[
|
...
]