Concurrently record the output of a brain scan A and standard input devices of keyboard and mouse B. Then neural network N evaluates A, forecasting B. After training with good data populating A and B, the intricacy of the commands and the statistical accuracy will enable a brain-only interface with a breadth of controllability and interactivity speed that will far exceed our current tools.

Initial systems will most likely track eye movement and visceral biofeedback that can be measured using electrodes and buttons.

6/15/2001 - I have received feedback that the brain thinks in words and concepts, rather than the transaltion to spelled and typed words, and would therefor render this invention useless. However, the input system described above would actually benefit from this characteristic rather in terms of input speed if the dataset A included a rolling period of time. (This is to allow for words or concepts to be interpreted in N over a series of typed inputs)

Labels: Tech