Take a look at the image above. Do you see a duck or a rabbit? Similarly, do you see the brain as being structured like a computer or the infinitely complex cosmos? Is memory more like RAM or an ever-changing symphony? Depending on how you look at it, it could be either.
Questions of perception are fundamental to how we investigate the world around us. This was the perspective of philosopher Thomas Kuhn, who coined the term "paradigm shift" as it relates to science. To illustrate the concept, he used the above example of the duck-rabbit illusion. The point is that our assumptions—the metaphors we employ to think about the brain, for example—determine how we investigate natural phenomena. And they can change over time.
Like the question of whether you see the duck and the rabbit, different people have different ideas of what the brain is and how it works—and certain ideas are more fashionable at one time or another. For example, the "computational theory of mind" holds that thinking is merely computing and is at the basis of much artificial intelligence research. But could other ideas of how the brain works help in the quest to reach AI, too?
"That could be cosmologists and astrophysicists, but it may very well be artists, as well"
In other words, if one idea of the brain, championed by a particular field of study, guides research down one path instead of another, will we lose out on what everybody else has to offer? Remi Quirion, a neuroscientist who was recently appointed as the chief scientist of the province of Quebec, thinks so.
Listen to Motherboard's podcast about brains:
"It is like, I speak French and you speak Chinese," Quirion told me at the IX Symposium in Montreal, before he participated in a panel discussion on interdisciplinary studies. "Once you start learning some key words, you can have an exchange. I hope that's what will happen with the human brain: learning from others. That could be cosmologists and astrophysicists, but it may very well be artists, as well. It could be artists who will see and think about the brain, human creation and activity, differently than a neuroscientist."
This kind of openness is essential as computers continue to creep towards attaining something that we can call "artificial intelligence." Already we have computer chips "inspired" by the brain, software that can "learn" using digitized simulations of groups of neurons, and powerful supercomputers than can calculate all the options in a game of poker. Scientists are even racing to build the first exascale computer—a machine with enough power to digitize every synaptic connection in the human brain.
Watch more from Motherboard: Dawn of the Killer Robots
But not everyone agrees that all this effort will culminate in something we can call true artificial intelligence. Duke neuroscientist Miguel Nicolelis has gone on record as saying that the brain is "not computable and no engineering can reproduce it," because it's dependent on unpredictable interactions between neurons. Computers, however, are based on math, rules, and logic.
Others have suggested that consciousness is embodied in the brain—the "hardware" and the "software" are one in the same—so any "intelligent" computer would have to exactly mimic our brain's structure. Doing so would be an extremely difficult, if not impossible, thing to achieve with a computer, since the brain has a unique property called neuroplasticity. Basically, it can rewire itself.
"Will computers take over and do better than the human brain? I'm not as confident or concerned about that," Quirion said. "Consciousness is not just brain wiring. We need experts from artificial intelligence, computer scientists, and astrophysics experts. If they talk to each other, I'm sure we can improve all our models."
"Merging these fields to some extent—well, maybe we'll get there"
An example of the kind of work that Quirion is describing might be that of physicist Dmitri Krioukov and his at the University of California San Diego. In 2012, they ran complex calculations on supercomputers and came up with a model for the growth of complex systems that could just as easily be applied to the universe, the internet, or the human brain.
Of course, this doesn't mean that the brain is the same as the universe. Rather, their work illustrates how a cross-disciplinary approach can help disparate fields to understand themselves, and each other, better. And, if artificial intelligence is the end goal, Quirion said, it could help us to develop it faster, if we are to develop it at all.
"The progress over the past 20 years, I'll say, has been quite amazing." Quirion said, when I asked a question about the role of interdisciplinary ideas of the brain in AI during the panel discussion after our interview. "Learning from what's done now in artificial intelligence, learning from data and cosmology, or astrophysics—merging these fields to some extent—well, maybe we'll get there."
Perhaps, after all, the question is not whether you are looking at a drawing of a duck or a rabbit. Instead, maybe all that's needed is to understand the perspective of someone who sees what you don't.