Who did not envy Professor X, the Vulcans or higher elves (I am so sorry, couldn’t think of less nerdier examples!) at least once in their lives? I sure did. Every time a question about superpowers comes up in “Would you rather?..” I always pick mind-reading without even hearing the second option. But how realistic is that? Will we be able to communicate without speaking any time soon? Create romantic moments by finishing each other sentences all the time? Win a Pulitzer Prize by knowing the true answers to all your interview questions? Just brainstroming here.
Just a picture of Patrick Stewart adding 100 points of awesomeness to this post.
But really, what is neuroscience’s opinion on this? It has indeed something to offer, even if not quite in the form we might expect (no Pulitzer for you yet). A scientist named Jack Gallant developed something called brain decoder which earlier was used to identify which of 120 pictures a participant is looking at and now can go as far as figuring out what movie you are watching. It goes like that: first a special program called “classifier” is shown a bazillion of patterns of brain activity and the corresponding images/videos which caused it; this way it learns to to associate different patterns with a specific picture or concept. At the end it knows that this pattern of activity is most likely to be caused by a cat or, say, by fighting babies. Later participants are put in an MRI scanner and shown short videos; after that the program is applied to the brain scans to decipher what the person was seeing in the given moment. Put it simple way, brain activity was turned into pictures. Dynamic brain activity could be decoded. Now before pointing out limitations let’s have a look at the follow-up research inspired by this technique.
Images reconstructed from brain activity.
One team is trying to make a dream-reader, which is obviously way trickier — how do we know what was seen (dreamt) when a specific brain activity was evoked? They solved it by putting the participants in this state when you are drifting away in the dreamland but not yet quite asleep, recorded their brain activity, woke them up and asked them what they saw. It lead to a 60% success rate, yet bear in mind that everyone’s signals are different and it is much harder to develop a universal dream recorder. Seems that we are still one step away from watching each other’s weird dreams on YouTube.
John-Dylan Haynes (who is my professor, woop woop!) took participants on a virtual reality tour through different houses and scanned their brains when they were taken on the second tour. With the brain decoder he was able to accurately say whether the participant has “visited” this house before — speaking technically, visual recognition could be detected.
Imagine all the implications! (Yes, we will discuss limitations later below.. Let a girl be excited for a moment.) We could identify whether a suspect has visited a crime scene before! Can develop a decoder which takes internal speech and translates it to external speech! Watch our dreams on YouTube!
OF COURSE, this whole thing is just in baby shoes so far. In order to train the program to recognize words or pictures in your brain activity you have to lay still in a scanner for hours and hours and hours while thinking about all possible things ten times over. Also the are other factor that might influence it — say thinking of cats makes you happy (and that is how the computer learns it). The next day your cat dies, however, so now thinking of cats will also be associated with patterns of negative emotions. Would it have an influence on accuracy? Probably, we don’t really know. Also so far it works on a very individual level, so, as noted previously, developing a universal mind-reader is not gonna happen in the next couple of years. At the current stage of development, quoting John-Dylan Haynes, “The best way to find out what someone is going to do is to ask them.”
P.S. Also people scared of brain monitoring or what not, it has to be noticed that in order to read your private thoughts an immense amount of compliance from your part is necessary — I highly doubt it that you will not notice being stored in a scanner for 4 hours thinking of specific things over and over again to train the decoding program. So chillax and be excited for our brave new world.
Brains and stuff 