New MIT Robot (Baxter) Can “Read People’s Thoughts“

Despite the major advances in artificial intelligence, there are still no computers which can have a realistic interaction with people.

Precisely, they can not feel what the man with whom they are interacting with thinks and they still react to strictly targeted functions. The scientists at MIT and Boston University have made the first robot with artificial intelligence (named Baxter) who can feel what the man on the other side is thinking.

This does not mean this  robot knows everything you are thinking. For now, it can only evaluate whether it is making a mistake or it is right, by following your reactions. For example, it was given the task to put spray paint or wires in two plastic containers which read what had to be put in them.

In this case the robot has a 50% chance to make the right decision, however it dramatically improves by following the brainwaves of people through an EEG monitor, which can detect signals which the brain sends when it senses an error. According to Daniela Rus, responsible for the project, these brainwaves are the easiest to detect in all the “noise” of brainwaves.

With them, it can evaluate if it is making the right decision in only 30 milliseconds, even before placing the matter in the wrong container and corrects the mistake. In order to make it more interesting, the creators of this robot have made it to have a face which “gets embarrassed” when it makes a mistake, which on video, as with every robot eith artificial intelligence, seems creepy.

The system is very limited for now, however as many similar to it, it announces how we could communicate with computers and robots in the future. Advance technology for mind reading could be used for controlling cars without the need to use hands, or in many other situations. Certainly, the team plans to develop Baxter in the future.