I also must somewhat disagree with Seth.
Currently much of the Brain-computer interface (BCI) research has been focused on re-enabling discrete sensory impairments, and it is argued that much of the BCI research is rooted in an old-fashioned and increasingly outdated cognitivist perspective. This cognitive perspective fails to account for much of what happens when humans interact in the world.
A growing body of research from fields such as robotics, artificial intelligence, psychology, learning, and human-computer interaction are finding that humans understand, learn, and interact with the world in a much more sophisticated way than purely cognitivist theories can explain. Theories of Embodiment are providing a much more complete understanding for these scholars. For example, many of the advances in robotics in recent years have come from letting go of purely cognitive theories and applying modern theories and philosophies of embodiment.
Therefore, in order to truly complement the ways in which humans fundamentally interact and understand their world the "perfect" human-computer interface needs to be situated outside the human body in order to interact as part of a social and embodied world.