Computers help humans in all sorts of ways, but one new method being trialled by the Massachusetts Institute of Technology is multitasking and a computer's ability to make it easier and more competent.
Known as Brainput, the technology uses specially formulated software twinned with a portable fMRI scanner. Measuring brain activity, the system is able to understand when the test subject is attempting to perform more than one task at once, leading it to then offload some of the functions to the computer.
This ability was showcased with a simple maze game navigated by a remote controlled robot. When there were two that needed to be navigated simultaneously, the computer was able to offer some assistance, using its own sensors to help guide the robot. Performance purportedly improved, while the user didn't feel like the system was too intrusive.
Interestingly however, when the autonomy was used at a time that the user wasn't multi-tasking, performance actually decreased.
Does this suggest that humans are inherently single threaded?
Technology like this could be employed in a variety of fashions, but certain situations like driving could be vastly improved. There are already many self-driving vehicles in the works, though not everyone is happy with the idea. Perhaps instead of having constant self-drive cars, an assisted experience would be more beneficial. If a person registers as tired, or distracted by eating, adjusting the stereo or a similar activity, the automated system could offer assistance.
Discussing the technology in a research paper, the team behind Brainput said: "In any activity involving multitasking or information overload, we could expect to see improvements in the user's performance and experience. Some examples of other domains are complex data analytics, air traffic control and management of multiple unmanned vehicles."