Unhandy valley

Max Titov's essay on robot end effectors, which I think closely relates to whether the humanoid form factor is the ultimate design or not. His curve plots usefulness against complexity: simple grippers hit an early peak and cover most tasks, half-committed human-like hands fall into a valley where they are more complex but no more useful, and only a long, expensive climb toward real biomimicry gets back above the gripper. The catch is opposition: without enough degrees of freedom the fingers can't meet well enough to pinch. He also pushes back on the "use human tools" argument, betting that robot-specific tools arrive before humanoid hands are both good and cheap. His advice is to take inspiration from the human hand without copying it blindly.