When Duncan Idaho unearths the memories of all his past incarnations, including those who left no genetic trace, we know that something truly remarkable is going on. Up to then, we’d met characters who could tap into their genetic history, bringing into consciousness the lives and experiences of their entire ancestry. But Duncan was a ghola, a clone, a manufactured human. He was touching something beyond his design specification.

In fiction, this is a familiar theme. The idea that we have some built in capability to network our minds with something fundamentally greater than ourselves is not only appealing, it’s resonant. It’s almost as if by fictionalizing it, we’re taking the first steps toward the reality that we just know is out there. It’s almost as if we intuit that we are, ourselves, networked creatures, always attuned to the greater network beyond.

As neuroscience starts it’s long, slow crawl toward understanding what makes humans tick, it keeps discovering new aspects of the networked nature of our brains. From the first hints of an idea in the late 1800s through the 1950s when neural nets were first explored and into the present research in brain mapping, the model keeps being re-discovered. The fact that it is re-discovered speaks volumes about the cognition box.

Early research into autonomous robots relied on the cognition box, a massively pre-programmed instruction set. In a way, robots in fiction operate from the same principle: what makes them safe is that they are predictable, logical. What makes them logical is that they are following an intricate set of ifs and thens and ors and all the other logical operators that are the foundation of microchips and computer programs. They can be validated. I’m fairly certain that Honda’s robots still operate this way.

And this modality of thinking, of programmed thinking, can be very powerful and remarkably flexible. But everything’s still in the box. The amazing thing about people is that we aren’t in that box, and we know it.

But we don’t get it.

Our persistent conception of artificial intelligence and of robots is of something that thinks the way we do. But if we don’t have a cognition box, that is, if we don’t have something recognizable that we can model, how can it ever be possible for something to be engineered to be the same? More recent experiments in robotics and artificial intelligence are fleshing out, as it were, a more organic form or artificial intelligence. These are machines and programs that interact with their environment, that learn, and that can be prodded into some remarkable behavior.

But for the most part they’re still human wannabes. They’re social robots, or useful networks, or otherwise entirely beholden to the human condition. Traditional definitions of artificial intelligence refer explicitly to a re-creation of the capabilities of the human brain. Implicit in this definition is the naïve belief that the human brain is a wet, spongy cognition box. Neuroscientists, philosophers and engineers are all feverishly engaged in proving that our brains are it, that the luminous nature of humanity as described and experienced by the overwhelming majority of humanity throughout our entire history is nothing more than something that can be reverse engineered with a collection of logic gates and clever software.

Even in the drug-addled world of Dune, the abilities of the biological are recognized as more nuanced, more advanced than the machines that were, once upon a time, destroyed; more so even than the biological machines they came to rely on. That this idea arises again and again in fiction is no accident. Resonance is not a fantasy, nor is it a product of wishful thinking. The uncanny valley is remarkably narrow, and frightfully deep. Quite simply, as these artificial creatures approach what their designers think is closer to the human condition, they don’t resonate with the rest of us. And by that, we know they’re not us.

And really, that’s as it should be. Artificial intelligence, when it arises, will not be anything like humanity. I doubt we’d even recognize that it had happened. But someday, in the not too distant future, I think we’ll be forced to confront the most amazing, the most frightening, the most humbling thing in human history: the other.

I can’t wait.

• BROWSE / IN TIMELINE
• « What is Government, Anyway? (part 1)
• » Original Sin

• BROWSE / IN Natural World
• « Postcard from Maine
• » Thinking Magical Thinking