Translated data: Recently, researchers from institutions including UCLA have introduced MultiPLY, an embodied intelligence large model. This model not only possesses multimodal perception capabilities, including touch, vision, and hearing, enabling AI to interact more comprehensively with 3D environments. Through interactions between the agent and the 3D environment, MultiPLY has demonstrated excellent performance in experiments such as object retrieval, tool usage, multisensory annotation, and task decomposition. Additionally, the researchers have created a large-scale multisensory dataset called Multisensory-Universe, which contains 500,000 data entries. This research provides new insights for building large models with multisensory capabilities and offers a new direction for achieving AGI.