Inception is Easier than Extraction

I finally watched inception last night. If you're interested in the mechanics of the subconscious, it's a must-see, both expository and provocative. (Keep your thumb on the pause button, though, so you can remind yourself what's going on in each dream level.) But it's based on the premise that discreetly planting ideas in people's heads ("inception") is much harder than discreetly extracting ideas. It seems to me that they got this reversed.

The closest thing we have to extraction as illustrated in the movie is fMRI. In a 2007 paper in Nature, researchers reported that they created a computational model of the visual cortex and trained a computer to decode fMRI data well enough to guess which of 1000 images a subject was looking at with about 80% accuracy. The researchers say that eventually a similar method could potentially reconstruct--not simply identify--pictures that people have seen, or even read dreams and mental imagery. (Privacy concerns may crop up in 50 years or so when TSA employees start poking around in your carry-on emotional baggage.) But the dream-reading potential is still hypothetical.

For years, psychiatrists have used both dream-analysis and Rorschach inkblots to deduce what lies deep in people's heads, but in both cases, people can just lie about what they're thinking. And even when they tell the truth, the art of interpretation has not yet reached the level of science and can be more misleading than informative.

More recently, research psychologists have used the Implicit Association Test to look at the strength of people's conceptual associations by measuring how quickly, for example, people rate "white people" as "good" versus "black people" as "good." The test can get past conscious defenses--it can reveal subconscious racial associations even if you don't think you're prejudiced--but its scope is limited. Good luck using it to extract complex intellectual property, or even a woman's phone number.

Various lie detection techniques have potential but are still relatively indirect, unreliable, and indiscreet.

Inception, on the other hand, is not nearly as hard as the movie makes it out to be. No fancy dream-machines required. Hundreds of studies demonstrate the power of subtle primes in changing behavior. In a classic study by John Bargh and collaborators, after test subjects unscrambled sentences with words related to the elderly in them, they walked more slowly down the hallway. And holding a warm cup of coffee makes you judge people more warmly. (Primes might take root more strongly in a hypnotized individual, but it's hard to hypnotize people without their awareness.)

A highly-cited review paper by Richard Nisbett and Timothy Wilson highlights the opaqueness of our internal processes even to ourselves. We generally don't know why we feel a certain way or where an idea came from. Sometimes we think we do, but we're often wrong. In Inception, the characters argue that we're always aware of the sources of our ideas and will recognize intruders trying to plant them in us, which is clearly untrue. There's a classic 1931 experiment in which the solution to a puzzle was to use a dangling cord as a pendulum. When the experimenter subtly brushed against the cord and made it sway, subjects quickly solved the puzzle but did not acknowledge the role of the hint when asked to explain their insight.

Our inability to properly monitor the sources of our ideas leads to the common phenomenon of cryptomnesia--thinking that what is really a memory is actually a new idea--and in some cases inadvertent plagiarism. In brainstorming tasks, people will often repeat others' ideas without realizing it. To counter cryptomnesia while researching my book, I make note of when I come across an idea that surprises me so that I can accurately attribute it later. It's easy to absorb an idea and then believe honestly that it was generated by yourself. (Or, more subtly, to remember where one first heard an idea but later find it no longer surprising and in fact so obvious and intuitive that it doesn't deserve explicit attribution--a type of hindsight bias.)

So inception is a simple matter of persuasion plus cryptomnesia.

When, during the movie, I pointed out to my roommate that inception was not such a big deal, her reaction was, "Exactly. I'm in sales. My whole job is convincing people of things and making them think they thought of it first."