Slate's Bizbox
human nature: Science, technology, and life.

Full-Mental NudityThe arrival of mind-reading machines.

By William Saletan
Updated Tuesday, March 20, 2007, at 5:02 AM ET
Illustration by Rob Donnelly. Click image to expand.

Years ago, Woody Allen used to joke that he'd been thrown out of college as a freshman for cheating on his metaphysics final. "I looked within the soul of the boy sitting next to me," he confessed.

Today, the joke is on us. Cameras follow your car, GPS tracks your cell phone, software monitors your Web surfing, X-rays explore your purse, and airport scanners see through your clothes. Now comes the final indignity: machines that look into your soul.

With the aid of functional magnetic resonance imaging, neuroscientists have been hard at work on Allen's fantasy. Under controlled conditions, they can tell from a brain scan which of two images you're looking at. They can tell whether you're thinking of a face, an animal, or a scene. They can even tell which finger you're about to move.

But those feats barely scratch the brain's surface. Any animal can perceive objects and move limbs. To plumb the soul, you need a metaphysician. John-Dylan Haynes, a brilliant researcher at Germany's Bernstein Center for Computational Neuroscience, is leading the way. His mission, according to the center, is to predict thoughts and behavior from fMRI scans.

Haynes, a former philosophy student, is going for the soul's jugular. He's trying to clarify the physical basis of free will. "Why do we shape intentions in this way or another way?" he wonders. "Your wishes, your desires, your goals, your plans—that's the core of your identity." The best place to look for that core is in the brain's medial prefrontal cortex, which, he points out, is "especially involved in the initiation of willed movements and their protection against interference."

To get a clear snapshot of free will, Haynes designed an experiment that would isolate it from other mental functions. No objects to interpret; no physical movements to anticipate or execute; no reasoning to perform. Participants were put in an fMRI machine and were told they would soon be shown the word "select," followed a few seconds later by two numbers. Their job was to covertly decide, when they saw the "select" cue, whether to add or subtract the unseen numbers. Then, they were to perform the chosen calculation and punch a button corresponding to the correct answer. The snapshot was taken right after the "select" cue, when they had nothing to do but choose addition or subtraction.

Until this experiment, which was reported last month in Current Biology, nobody had ever tried to take a picture of free will. One reason is that fMRI is too crude to distinguish one abstract choice from another. It can only show which parts of the brain are demanding blood oxygen. That's too coarse to distinguish the configuration of cells that signifies addition from the configuration that signifies subtraction. So, Haynes used software to help the computer recognize complex patterns in the data. To dissect human thought, the computer had to emulate it.

Each participant took the test more than 250 times, choosing independently in each trial. The computer then looked at a sample of the scans, along with the final answers that revealed what choices had actually been made. It calculated a pattern and used this pattern to predict, from each participant's remaining scans, his or her decisions in the corresponding trials. Haynes checked the predictions—add or subtract—against the participants' answers. The computer got it right 71 percent of the time.

I know what you're thinking: Why would anyone want a machine to read his mind? But imagine being paralyzed, unable to walk, type, or speak. Imagine a helmet full of electrodes, or a chip implanted in your head, that lets your brain tell your computer which key to press. Those technologies are already here. And why endure the agony of mental hunt-and-peck? Why not design computers that, like a smart secretary, can discern and execute even abstract intentions? That's what Haynes has in mind. You want to open a folder or an e-mail, and your computer does it. Your wish is its command.

But if machines can read your mind when you want them to, they can also read it when you don't. And your will isn't necessarily the one they obey. Already, scans have been used to identify brain signatures of disgust, drug cravings, unconscious racism, and suppressed sexual arousal, not to mention psychopathy and propensity to kill.

VIEW AS SINGLE PAGE
Print This ArticlePRINTDiscuss this in The FrayDISCUSSEmail to a FriendE-MAIL
Share on FacebookPost to MySpace!Share with MixxDigg ThisShare with RedditShare with del.icio.usShare with FurlShare with Ma.gnolia.comShare with SphereShare with Stumble Upon
William Saletan is Slate's national correspondent and author of Bearing Right: How Conservatives Won the Abortion War.
Illustration by Rob Donnelly.
Join the Fray: our reader discussion forum
What did you think of this article?
POST A MESSAGE | READ MESSAGES

Remarks from the Fray:

Well it seems that Ray Kurzweil's predictions of CPU and MRI technology are tracking pretty well.

Given the rate at which technology accelerates - it isn't all that far fetched to think that somewhere between 2025 and 2035, we will have the ability to scan the neuron and synaptic states of the brain and to replicate them within a computer using neural networks or something similar.

And the question will no longer be one of "mental nudity" or of epistemology but instead one of defining what is life. Is the scan of your cognitive state that then continues to run forwards INSIDE the computer any less cognitive than you?

Can a soul reside inside a computer?

--Degsme

(To reply, click here.)

Enough of this metaphysical nonsense from Saletan. fMRI (and neuroscience for that matter) is a materialist enterprise. I'm getting pretty sick of Saletan coming at us again and again with the "we've found the soul, we've scanned choices, natural selection is shaped by free will.. Did I blow your mind?"

No, dooood, you did not. The reason it sounds so freaky to scan "free will" or "the soul" is because you haven't...no one can and no one will. If these things exist at all they are nonphysical, so, for the purposes of science they can safely be ignored.

What you scan in an fMRI, or an EEG, or measure with a key press are the consequences of the physical events which are thought. Cognition is a material activity. Once you get this concept through your head then you can stop with the mysticism.

For instance, "thinking about faces" is a cognitive state instantiated in matter...thus, it has a chance to show up on an fMRI. The cognitive state that we experience as "the intention to add or subtract" something is also instantiated in matter. The fact that an fMRI picks this up, too, is not a transcendental experience unless you willfully misinterpret what's going on. Who's to say a cognitive "intentional state" is uncaused, separate from the material world, or otherwise special or supernatural? Why cannot an intentional state be caused just like anything else? The answer is "it can".

There's plenty of wonder to be had in the science itself...this is very cool stuff. We don't need Saletan's injecting inappropriate, outdated shorthands and concepts from Cartesian dualism and medieval philosophy muddying the waters in order to hook us.

--Mangar

(To reply, click here.)

The rights of the people to be secure in their persons, their houses, papers, and effects, against unreasonable searches and seizures, shall not be violated and no warrants shall issue but upon probable cause......

Our thoughts are our own, they are to our persons as personable as can be. Other than our very soul our thoughts are who we are. Indeed our real thoughts known only to us perhaps describe and identify our soul as no other means could hope to achieve.

The all intrusiveness of this technology flies in the very face of the freedoms we claim to be fighting for. This technology is in its infancy but given the rapid pace of development today this technology could possibly mature in a few short years. Strong, ethical safeguards must be put in place now along with strict limitations as to who may have access to this technology and for which purposes. Without such safe guards in place perhaps we should be as the throngs of people in the movie Frankenstein and surround this castle with torches and pitchforks, though I would hope we do a better job of preventing the monsters release.

--NickD

(To reply, click here.)

"Under controlled conditions, they can tell from a brain scan which of two images you're looking at. They can tell whether you're thinking of a face, an animal, or a scene."

The way to beat this is to be paranoid, and carry false images in your head, to deliberately fool the machines. Mutter constantly about all kinds of nonsense. That way, they can't "see" your mind. if you don't want them to "see" you thinking about a cow, then force yourself to imagine a chicken instead, while waving your arms about wildly. Oh, and possibly tinfoil hats help, too.

They can even tell which finger you're about to move."

Probably not, if I'm alternating with my mind: left, right, left, right, and then I make a split decision. Because even I didn't know which finger I was about to move until I did it.

--Caromer

(To reply, click here.)

(3/20)