There is something unsettling, and more than a little creepy, about a future in which our fancy technology becomes so sophisticated that it can think faster and smarter than the humans who created it. Well, the future is here, and it’s been here for a while.
Researchers at Penn State and the U.S. Air Force have created materials that can — believe it or not — “think” autonomously.
It seems like just yesterday when Jeopardy grandmasters Ken Jennings and Brad Rutter were pitted against Watson, a room-sized supercomputer created by IBM. The battle took place in 2011, as millions of us were riveted to our televisions for the 3-day tournament. At the end of day 1, Rutter and Watson were tied for the lead. On day 2 Watson pulled ahead by hitting its (his?) buzzer in a few hundredths of a second, faster than mere men. Jennings pulled ahead on day 3, but when Watson hit a Daily Double, it (he?) sealed the tournament win. It was Watson $77,147, Jennings $24,000, Rutter $21,600.
Perhaps it’s not entirely surprising that a so-called “supercomputer” could outwit humans. But how could this futuristic material engineered to make decisions independently be possible?
To start, researchers developed a novel alternative to traditional integrated circuits. Integrated circuits are typically composed of multiple electronic components housed on a single semiconductor material, usually silicon. They run a myriad of modern electronics, such as phones, cars, and robots. According to principal investigator Ryan Harne, professor of mechanical engineering at Penn State, integrated circuits are essential to scalable computing of signals and information, but the composition has been limited to silicon semiconductors.
“We have created the first example of an engineering material that can simultaneously sense, think, and act upon mechanical stress without requiring additional circuits to process such signals,” Harne says, in a statement. “The soft polymer material acts like a brain that can receive digital strings of information that are then processed, resulting in new sequences of digital information that can control reactions.
Harne says that the material uses a “thinking” process similar to that of humans. Its potential applications are envisioned for search-and-rescue systems, infrastructure repairs, and in hybrid materials that can destroy airborne pathogens.
“What makes humans smart is our means to observe and think about information we receive through our senses, reflecting on the relationship between that information and how we can react,” Harne explains.
Nerves digitize sensory information into electrical signals which travel to the brain. The brain assesses the signals and tells the body how to react.
When the researchers subjected their engineered material to mechanical information — applied force that deforms the material — it digitizes the information to signals that its electrical network can advance and assess.
The researchers were stuck, until they rediscovered a 1938 paper published by Claude E. Shannon, who later became known as the “father of information theory.” Shannon described a way to create an integrated circuit by constructing mechanical-electrical switching networks, the same binary logic gates Harne used previously.
“Ultimately, the semi-conductor industry did not adopt this method of making integrated circuits in the 1960s, opting instead to use a direct-assembly approach,” Harne says. “Shannon’s mathematically grounded design philosophy was lost to the sands of time, so, when we read the paper, we were astounded that our preliminary work exactly realized Shannon’s vision.”
The researchers are now modifying the material to process visual information like it processes physical signals.
“We are currently translating this to a means of ‘seeing’ to augment the sense of ‘touching’ we have presently created,” Harne said. “Our goal is to develop a material that demonstrates autonomous navigation through an environment by seeing signs, following them, and maneuvering out of the way of adverse mechanical force, such as something stepping on it.”
The research is published in the journal Nature.