Hollywood has propagated the idea of autonomous, humanoid robots for decades. In Metropolis, the highly influential film from the early 20th century, the inventor Rowtang builds a robot in the likeness of his deceased lover. 2014’s Ex Machina explored the inherent challenges of building sentient, humanoid machines. Why is popular culture so obsessed with robots? Their presence is apparent in every aspect of life. Perhaps this obsession stems from the knowledge that convincingly human machines are inevitable. Technology has progressed to the point where some experts say that these robots could be a reality before 2050. Which begs the question, are people ready for human-like robots?
Robots Have Been Around for Centuries
The concept of artificial servants has been around since ancient Greece. Hephaestus, the Greek god of craftsmen, created the bronze man Talos to defend Crete. In India, mechanical robots protected the relics of the Buddha. Extant Christian legends even note creation of artificial servants by Bishop Albertus Magnus, much to the annoyance of his student, Thomas Aquinas.
Earliest plans for robot-like machines date as far back as the fourth century BC. Many innovations in automata came about from work done by Hero of Alexandria. Notably, legends say that he had created an automaton that could speak.
It wasn’t until Leonardo da Vinci that there were practical designs for humanoid robots. Though based in large part on his iconic Vitruvian Man, no one knows for sure if he ever attempted to build the machine. Since discovery of the blueprints in the mid 20th century, experts have faithfully constructed Leonardo’s robot and found that it is fully functional.
Automata using geared mechanisms rose to prominence during the late 19th century. Largely used to advertise businesses and show off engineering prowess, they persisted in some form until the middle of the 20th century. They soon fell out of favor with the development of computers and applications of practical technology.
The Uncanny Valley
The servants and robots of history were humble, mechanized machines easily distinguishable from a human being. They did not possess autonomous thought or move about of their own volition. You would never mistake the Tin Man from Frank L. Baum’s the Wonderful Wizard of Oz for a person on the street.
Modern robotics technology has changed that. Technology today promotes the use of human-like features as a way for people to trust technology. Getting medical attention from a cold, calculating computer in a box does not compare to getting the same attention from one able to display signs of empathy using an actual face.
Scientists and engineers struggle with one issue, though: the uncanny valley. The uncanny valley is the phenomenon whereby a humanoid object appears almost, but not exactly, like a real person. The “uncanny” resemblance is familiar yet elicits a revulsion or eeriness in those who see it.
Until recently, it is the uncanny valley that has kept robotics companies from trying to make their products look too much like humans. The technology just hasn’t been there. Instead, corporations rely on novelty and simplicity for their designs. Projects such as Honda’s ASIMO and Boston Dynamics’ PETMAN are impressive in their own right. However, they don’t resemble people in anything other than anatomy. If you saw one in the street, you would never mistake it for another person.
Software and artificial intelligence continue to improve at record speeds. It is the hardware side of robotics that needs further development before it can overcome the challenges of the uncanny valley.
What Does the Future Hold?
Promoting emotions in machines starts the human race down an interesting path, one laden with moral conundrums. Even artificial empathy relies on something like emotion. Does that mean that the artificial intelligences and robots of the future must come equipped with all manner of emotions and feelings?
Emotion is contingent upon one very important point. The minds of increasingly human-like robots rely on a skill pivotal to all species: learning. Machine learning involves programming artificial intelligences to observe, process, analyze, and utilize information from massive amounts of real-world data. The more input, the better equipped an intelligence becomes at learning. This idea is central in the development of machine emotions. They analyze thousands of speech characteristics, including speed, volume, tone, and pitch, to learn to recognize what certain mental states look like.
Machine learning is key when developing the kinds of robots now taken to market. Though they may not look enough like humans to creep people out, companion robots must display more human traits than a simple anatomical resemblance. Robots such as Aldebaran Robotics’ Pepper need a degree of social awareness and perception in order to function properly. Yet, for all their human characteristics, these robots do not always take initiative when it comes to interactions. Though they encourage and welcome them, these companions are largely responsive rather than active.
Ultimately, the truest form of humanoid robotics comes with the marriage of companionship and functionality. Current robots are either one or the other. You can choose your Roomba or your Pepper. Right now, you can’t have both. Once perfected though, the convergence of these two characteristics can bring about the first convincingly human-like robots.
What are your thoughts on this hotly debated issue? Do you think that scientists and engineers should stop pursuing humanoid robots? Do you believe the future holds dangerous possibilities if technology passes through that uncanny valley? Maybe you think that this is just the natural progression of society. Perhaps you think that, regardless of how they look, robots exist to make our lives easier. No matter what side of the issue you fall on, the future of robotics stands to change the world for everyone.