May 16, 2012
Computer Reads Your Thoughts, Does Your Multi-tasking Dirty Work
Researchers from MIT, Indiana University and Tufts University have developed “Brainput,” a system which detects when your brain is trying to multitask, then offloads some of that work to a computer. Erin Treacy Solovey led the new research in hopes of increasing productivity and focus by letting computers do some of our grunt work.
The novel idea of having computers do our dirty work isn´t a new one. After all, we did invent the calculator and computer, and without these there would be no internet and no world of knowledge one Google search away. Brainput will work in a different way, however.
The device uses functional near-infrared spectroscopy (fNIRS) to measure the activity of your brain. The data is then analyzed to determine what you´re busy thinking about. If Brainput determines your mind is busy multitasking, its software will kick in and begin to take over some of that work for you.
To test this system, Solovey created some robots and a maze. The goal was to have someone navigate both of these robots simultaneously via remote controls through a maze while wearing the Brainput device on their head. When the software determined the person´s brain was focused on one of the robots and not the other, it would tell the ignored robot to use its sensors to pilot itself through the maze. In the end, while Brainput was being used, the operators performance was improved. What´s more, they generally didn´t notice the robots were partially moving on their own, giving the user a sense of having controlled both simultaneously with their mind.
What´s interesting, however, is that when the robots became autonomous during a time the user wasn´t multitasking, user performance decreased, meaning the user was more inclined to let the robot take over and do the work rather than do it themselves.
Similar technology exists, such as cars and steering wheels which can kick into autopilot when they realize the driver is no longer paying attention or has fallen asleep. Google, for example, has built a car which drives itself.
Using cameras, lasers and radar, the Google car can safely navigate roads all by itself, without the need for human piloting.
Brainput could be used in such applications, allowing cars to take over the driving duties if we reach down to select a CD. Using Brainput is easy as well, and the entire device is lightweight and free of a lot of hardware. The fNIRS system is mostly composed of a headband and a few wires which attach to the computer.
According to Extreme Tech Solovey intends to move the research forward and investigate what other cognitive states can be reliably detected with fNIRS.
In the end, Solovey´s research could bridge the “communication” gap between what we want our tech and computers to do and what they actually do. Such an idea sounds very cool and very helpful, but some problems involving manners and matters of politeness could easily arise should our computers instantly give way to our every immediate notion and whim.