As we approach 2025, it feels like humanoid robots are closer than ever to becoming a part of our daily lives. But do we need AGI for this revolution to truly begin, or is the next step simply about refining and perfecting existing technologies like perception systems?<p>What do you think is missing to make humanoid robots a reality in our homes and workplaces?
I am curious if there is a coherent approach to modeling attention for long- or medium-term goal following. Currently, LLMs resemble a fish or a small child in their focus and behavior. Robots with such limitations would be strange to work with.