What is Embodied Intelligence? Generative AI is when machines can understand human natural language and output text, images, sounds, codes, etc., while embodied AI means that machines can perceptually interact with the physical world and output various mechanical actions.
Tesla's breakthrough and potential in embodied AI has convinced the capital market that it will lead the field, just as OpenAI leads generative AI.
The most profound and technologically advanced in the field of robotics is Boston Dynamics, whose robot Atlas has been able to walk, jump, run, and even flip like humans.
In his speech at the 2023 ITF World Congress, Jensen Huang predicted that the next wave of AI will be embodied AI, including robots, self-driving cars and more.
Microsoft is recently planning to extend ChatGPT's capabilities to robotics, using words to control bots.
Google and a group of AI researchers at the Technical University of Berlin in Germany have published the Pathways Language Model with Embodied, PaLM-E.
From alpha dog to CHATGPT, the prototype of generative robots has basically been completed; Embodied robots trained using simulated environments and real environments have been widely used in autonomous driving and other fields; When generative robots and embodied robots become one, because they have a powerful neuronal system that can learn and judge themselves, silicon-based life will be born and humans will benefit from it.
When humanoid robots (silicon-based life) evolve to a certain extent, they will be able to replicate themselves and even prospect autonomously on Earth and alien systems, equipping themselves with mini-nuclear reactors to provide permanent power.
Scientists like Musk are justified in their fear of artificial intelligence, and such a powerful new species must be controlled by humans, with strict rules set for it, and not harmed by it. Once violated, the humanoid robot must be able to self-destruct.
