Robots helping humans, humans helping robots
People have been terrified by the rise of machines for a good part of the last 200 years since Luddites destroyed sewing machines in textile factories. Technology has changed remarkably but newspapers still keep predicting humans losing jobs with the same degree of drama:
- 1921: “Will Machines Devour Man?”
- 1928: “March of the machine makes idle hands”
- 1961: “Automation Might End Most Unskilled Jobs in 10 years”
- 2011: “Will a robot steal your job?”
- 2020: “The Robots Are Coming. Prepare for Trouble.”
These frightening newspaper headlines reflect how strongly people have feared losing their jobs to robots. This is not a surprise — losing a job is a fundamental threat to our being. Your job often defines who you are. This deep-seated anxiety has only been attenuated in current pop culture by movies and comics.
While technology innovation in the past has improved the quality of life for most people and created more jobs, this fear again looms over us in 2020, fresh with media images of a new generation of artificial intelligence and smart robots. Politicians and companies promise that technology will again create more new jobs than it displaces. People are often left wondering “But what exactly are these new jobs?”
Addressing the question is a political hot potato. Without absolute certainty, few dare to hazard an answer. But when looking for new jobs, few bothered to look where the new robots in reality are. In fearing what most people have never seen, we run the risk of missing the big picture — it could just be that robots need humans now more than ever.
Where are the robots?
Robotics has made an incredible leap in the last decade — robots can now walk as humans can, navigate through city traffic, and sometimes see better than humans. This innovation should have allowed robots to get out of the factories to the world of humans to help with specific tedious and repetitive tasks.
Yet, most people (outside of Silicon Valley, Boston, and Zurich) have never seen any of these new robots in person. Besides iRobot’s Roomba, the latest massively commercialized robot was based on a robotic arm introduced in 1962. Despite billions invested in Silicon Valley startups over the past decade, robots still have not escaped robotic labs and factories. The main real-world result of intense investments in robot development has been a bunch of impressive Youtube videos, further stoking the flames of distress that robots will soon rule the world. Ironically, startups fail to push their vision to the market.
So, why is this the case? The truth is that robots fail in the real world!
Outside of a warehouse or a factory, we lose control over the robot’s environment. All possible outcomes of the robot’s actions increase infinitely and current AI cannot safely cope with that. However, a human can.
Robotics is a fairly new industry still mainly driven by academic research. It is a unique field as it combines advanced concepts from mechanical, electrical, and software engineering, and adds artificial intelligence on top. Universities were the first to bring all this together first on a bigger scale. This is unlike other industries like automotive or aerospace which were forged by engineers in factories, not academics in labs. As a result, most of the autonomous robots used today were originally built in university labs and only later spun off into startups.
Universities tend to train researchers, not engineers. Full autonomy is the holy grail in robotics research and this reflects the mindset in problem-solving when it comes to autonomous robots. An autonomous robot is not just a “smarter tool” enhancing human capabilities but “almost self-conscious being” with a (typically female) name. Problems of our complex world are currently resolved by increasing the complexity of algorithms, instead of step-by-step learning from failures. And deploying “human intervention” is a disgraceful option in these settings.
Robot failures actually require the physical intervention of skilled robotics engineers. This is precisely what makes operating robots an expensive affair. Businesses cannot fully rely on robots that are unable to recover on their own from their failures.
Yet, until general AI is achieved, robots will keep failing in the real world. Before robots can help us eradicate dull and dangerous jobs they first need help from humans. Most importantly, robots need our trust.
Robot pilots
At Cognicept, we tackle this problem head-on by building systems to help humans operate robots remotely. Human robot pilots attend to robot failures and help robots navigate new situations not previously programmed for. This is a 24/7 service where our operators are unencumbered by geographical, cultural, and time differences. Our approach enables companies to deploy robots outside of labs and empowers people through generating new robot pilot jobs.
Our core mission is to make robots more accessible to the everyman. Most pertinently, becoming a robot pilot should be a skill attainable within weeks of retraining. We could reskill and upskill the current workforce in a matter of months.
While the forces of creative destruction will remain with us in the foreseeable future, they also lend us an extraordinary opportunity to reinvent how we work. At Cognicept, our answer to the perennial question posed by robots is our ambition to make the robot pilot a mainstream job in the 21st century.
