Beware the Brain Police

Tin foil cat. Image by jeffhall2069 via Flickr

Apparently, in the foreseeable future, cops will be able to carry around devices to read your mind and find out if you’re thinking nasty thoughts about them. At least, that’s according to this somewhat panicky column in the International Herald Tribune.

The reality is a little less sensational — but, in many ways, much cooler.

The problem in any “cops with mind readers” scenario is that serious scanning requires fMRI machines the size and weight of recreational aquatic vehicles. Further, these machines require subjects to sit inside for long periods of time while remaining absolutely motionless. (Not to mention what happens if there’s any metal around when one of these babies is running.)

But, let’s take for granted that technology gets smaller and lighter and more portable… eventually. Being able to “listen in” on a person’s internal monologue is simply not something scientists are anywhere close to being able to do, even under the most controlled of circumstances.

There are, however, some things they are getting close to being able to do. And they’re pretty cool:

I wrote a bit about fMRI lie detection the other day. It’s a long way from being usable, but we are able to see some things related to deception on brain scans; we’re likely to figure out a lot more in the future.

Eight trials (top) and an average (bottom). copyright Elsevier

What’s cooler is that scientists in Japan have recently used fMRI to crudely reconstruct an image a subject views in a brain scanner. Pictured at left are a series of images that the researchers created by reading data from the visual cortexes of their subjects, who were viewing 10 X 10 pixel black-and-white images of various letters. (They spell out Neuron. The name of the journal where the research was published. Cute.)

It’s impressive, but not surprising, that scientists were able to do this. As John Timmer writes at Ars Technica, previous research has shown that it’s possible to figure out which image a person is looking at, when given a limited set of images. And it’s been shown that the visual cortex contains something like a rough map of the eye’s retina; if you can learn to read that map with a scanner, you can reconstruct the image. But having people look at a novel image and then recreate it — pretty amazing. (Even if this is all still miles away from fine-grained, color, 3D, moving images, etc.)

Another cool thing scientists are working on is determining what word a person is thinking of. Researchers at Carnegie-Mellon University are making rather amazing progress in being able to tell what word (for instance, “knife” versus “barn”) a person is thinking of in a brain scanner (link to Science here; PDF of study here; 13 minute CBS News segment embedded here). Their technique, which I won’t pretend to fully understand, combines brain scanning along with super-computing to predict what brain activation will look like for various words. So far, they’ve had more luck with concrete words, like nouns describing objects, than with more abstract ones, like verbs and concepts. But, their work continues apace.

So, we’re a long way from a world of dystopian mind-reading and government thought control. Still, perhaps (contra Mind Hacks), such arm-chair theorizing helps us begin sorting through the real issues of cognitive liberty that will emerge — possibly much sooner than we think — when various mind-reading technologies achieve a discomfiting maturity.