It's not that building a human robot would be impossible, it's whether or not we understand how robotics already fit into our lives.


Earlier this month DARPA showed off the winner of its Robotics Challenge competition, a humanoid robot called Atlas, which stood 5 feet 10 inches and weighed 240 pounds. The design was made by Boston Dynamics in an effort to create a robot that could be of use during emergencies, taking on many of the most hazardous tasks a human might otherwise have to do.

While still a simple prototype, Atlas is able to stand and move on two feet, and can turn on fire hoses, and is controlled by an on-board computer to help it identify and complete tasks.

The idea of a humanoid robot is a longstanding fixation in Western culture, one that often leads to paranoia about either human obsolescence or the infiltration of machines that can't be distinguished from humans. 


It's a form of self-flattery to think the inevitable function of robotics should mimic the human form. In 100 years, the practice of building human robots will likely seem like an archaic curiosity.


In truth, the future of robotics will probably have very little to do with replicating the human form, but will rely on the more primordial structure of swarms. Away from the promethean effort to create a metallic replacement self, the field of swarm robotics has been thriving in recent years.

Researchers at Harvard have released an insect-sized robot called Kilobot, which can be used in large groups to swarm an area, collecting geographic data about it, searching for survivors in emergencies, and building support structures to keep collapsed buildings from further damage.

A researcher at North Carolina State University has created a similar swarm structure for robotics by applying small battery backs to cockroaches, which can then be steered with small electrical pulses and used to collect data or search for people across wide space. Researchers at the Labratory of Intelligent Systems have built flying swarm robots, which communicate with one another through a local wireless signal and can be used to assist with mapping and environment monitoring in disaster zones. And earlier this year the FDA approved RP-VITA, the first robot allowed for use in hospitals, which functions in a data gathering and monitoring role similar to other swarm robots.

The idea that robot swarms will make up a bigger part of our future than humanoid ones is an extension of the human microbiome, the autonomous population of roughly 100 trillion bacterial cells living in the human body, which significantly outnumber human cells. While we think of ourselves as distinct, autonomous entities, we are dependent on external lifeforms cohabiting in our cavities to remain properly alive and functional.

In this light, the desire to create an autonomous, superpowered machine in human form seems like a Freudian daydream of self-negation, creating a world of superbeings that became so physically competent and functionally efficient that all of the discomforting emotional and intuitional paradoxes of their creators can be isolated and treated as dead weight.

Human-looking robots will inevitably become an artifact of our naive and unknowing past. It's not unlike all the strange fantasies of flight people had 100 years ago, which contorted present day machinery into flying variants. The zeppelin was a floating boat, and early airplane prototypes were just bicycles with flapping bird wings attached to them. All of these inventions missed the point of flight, which was not to put things that existed already into the air, but to imagine a kind of machine that didn't exist in any fashion.

Robotics has languished in this same trap for long periods, attempting to recreate with hydraulics and computer processors, the functioning of creatures that already exist in nature. The persistence of these fantasies about androids, replicants, and cyborgs masks the flourishing of robotics all around us, from the pocket-dwelling Siri to the unseen circuitboards that control 747s for the duration of most flights.

It's a form of self-flattery to think the inevitable function of robotics should mimic the human form. In 100 years, the practice of building human robots will likely seem like an archaic curiosity, like those far-flung hobbyists still experimenting with flying bicycles.

It's not that building a human robot would be impossible, but that once we have a clearer understanding of how robotics already fits in our lives, and how that model is already evolving into newer and more unexpected forms, Terminators and Atlases will come to seem as curiously doomed as the Hindenburg. 

Michael Thomsen is's tech columnist. He has written for Slate, The Atlantic, The New Inquiry, n+1, Billboard, and is author of Levitate the Primate: Handjobs, Internet Dating, and Other Issues for Men. He tweets often at @mike_thomsen.