A new study from MIT and the Boston University suggests that we are one step closer to figuring out how to control robots and computer using our mind. The recently publish study reveals that a human user can interact with a machine and control its action by telepathically telling him what’s wrong and what’s right. This experiment proves that figuring out how to decode brain signals is paramount to bridging the gap between the human brain and the computer.
Ever thought that one day you will be able to work on a computer just by using your brain power? A team of computer researchers from MIT and the Boston University proved that not only this reality is feasible, but that we are one step closer to achieving it.
Combining electroencephalography, state-of-the-art machine learning algorithms, and a cute-looking robot which blushes each time it makes a mistake, the team of researchers managed to teach a computer to recognize and understand specific brain patterns.
Although we are still quite far from having machines that could perform stuff on their own after reading our thoughts, MIT and Boston’s mind-reading computer are as close as we can get, at the moment, to this vision of the future.
According to the team involved in the project, the robot is trained to pick up and decode error-related potentials. What are those? As the team pointed out, each time we make a mistake or we see someone else making a mistake, our brain transmits a specific electric impulse. Now, this brain spark prompts us to correct that error or to point out to someone else that he made a mistake.
At the moment, robots are not clever enough to understand when they’ve made a mistake, meaning that it’s up to us to point it out. During their experiments, the team asked a volunteer to stand in front of a robot and to wear a special cap which has hooked up to an electroencephalogram machine.
The EEG will measure the patient’s brain waves in real time, while he or she is watching the robot perform another task. Now, as the scientists explained, the robot was trained to pick up and decode these error-related potentials in under 30 milliseconds.
In front of the robot, there were two boxes – one containing paint and one containing wires. The robot’s job was to pick up random items from the table and to place them in the correct container. All this time, the patient is carefully watching the robot’s actions.
If the robot picks up an item and attempts to place it in the wrong box, the EEG machine would pick up the error-related potential, relay it to the robot which, in turn, will make the correction.
Although the robot still has some issues figuring out the where to put the item, the MIT and Boston University researchers declared that this is the first step to creating a stable and reliable brain-computer interface.
Image source: Wikipedia