I remember hearing the crowd around me gasp in disbelief as my professor, Human-Computer Interfaces researcher Jordan Aiko Deja, asked everyone to “give computers a chance to be creative.” There was no surprise there. After all, the room was filled with multimedia artists, psychologists, and user experience designers—people who rely on the humanities and on “human creativity” to further their craft. Hearing a computer science professor talk about computers being creative was just pure blasphemy. Almost expectedly, I heard my seatmate ask “Are we being replaced by computers too?” I could feel all eyes striking my professor with contempt while he spoke about creative computers on an international conference stage. Although I am an artificial intelligence (AI) practitioner myself, I couldn’t help but feel the concern that everyone else was feeling, and I do admit that this concern is real. That night, the question, “Are we being replaced by computers too?” kept lingering in my mind.
Job computerization is real. Everyday, thousands of people lose their jobs because of computers—smarter, faster, more eﬃcient, and less error prone than conventional human hands. As an AI practitioner, this is one of the ethical concerns that we face in our research. While it is true that we are advancing science and technology, pushing innovation forward as a catalyst for global change, it does come with a big price. We displace workers. Worse, most of the workers that we displace have blue-collar jobs, people who work seven to nine to earn minimum wage, working their fingers to the bone to put food on the table for their families. In the advent of the 21st century, it is not uncommon to hear Tesla advertising their latest self- driving car, or Amazon showcase their warehouse fully-operated by hundreds of robots. These days, it is not uncommon to discover that the telemarketer you hear over the phone, or the call center agent you were just talking to are both intelligent machines. Gone are the day that factories are full of men and women sewing their way through hundreds of pieces of clothing. While this scenario still holds true with some countries, it is undeniable that manual labor is being taken over by computers, and in the next decade or two, nearly all manual labor will be computer-operated. Unlike us, computers rarely make mistakes. They are more accurate, never tire out, and need minimum supervision to work. In a nutshell, they are just better at it than we are. This simple diﬀerence in eﬃciency is the reason why so many workers are being displaced, their jobs overtaken by smart machines. As the days go by, more and more jobs are threatened to be replaced by computers, and as a result, hundreds and thousands of laborers will be stripped oﬀ of their livelihood. Is this really the price we have to pay?
As my professor said, AI was not made to replace humans. Read that statement again, roll it on your tongue, and swallow it. AI was not made to replace humans. In fact, AI was made to create the best human-computer teams. In the long run, humans cannot be replaced from the picture. While it is true that computers are more eﬃcient and accurate than we are, these usually only apply to tedious and repetitive tasks. In a talk by Prof. Shengdong Zhao of the National University of Singapore titled “Innovation Principles in HCI”, he describes that the goal of AI is to create the perfect human-computer teams. In doing so, tasks must be subdivided into subtasks, and must be evaluated which are best done by computers and which are best done by computers. If the task is repetitive and tedious, computers usually perform better than humans. This is corroborated by various research and works on job computerization.
There are jobs that cannot be replaced by computers. In a research by Frey and Osborne in 2017, they found out that jobs exhibiting three characteristics are best suited for humans and not computers. These characteristics are social intelligence (jobs that involve negotiation and caring, like public relations for example), creativity (jobs that need originality, like fine arts and biology), and perception and manipulation (jobs that need finger dexterity and perception-assisted accuracy, like surgery). Jobs like recreational therapists, mental health and abuse therapy social workers, and teachers are hardly going to be replaced by computers. On the other hand, jobs that don’t need or hardly need these characteristics, like telemarketing, data typing and entry, watch repair, and general labor are to be replaced by computers. This research corroborates the ideal of the perfect human-computer teams. We have computers to aid us in jobs that they work best in, and we humans work in ones that suit us best. For example, medical diagnosis can be done by a computer assistant, while the surgery and check-up proper can be done by a human doctor. Oncologists at the Memorial Sloan-Kettering Cancer Center in the United States use an AI system called IBM Watson to run through 600,000 medical evidence reports, 1.5 million patient records and medical trials, and 2 million pages of text from medical journals to give the best diagnosis. Running through all of that material to give the best analysis and diagnosis is impossible and impractical for a doctor, even for teams of doctors, to do eﬀectively. This task is best done by computers. On the other hand, the actual surgery will be done by humans because it needs human perception and manipulation in order to be done. As seen in this example, AI is not meant to replace us. They are meant to augment us. They are meant to aid us build the best human-computer teams.
It is inevitable that some jobs will be replaced by computers, usually jobs that are tedious and require repetitiveness. It is inevitable that jobs will be lost to make way for innovation—it is the price of progress. Ever heard of a “vacuum tube engineer”? Back in the day, computers relied on vacuum tubes that cost millions and were the size of rooms to function. When the “transistor” was built in 1947, computers were scaled down to the size we know today. As a result, the vacuum tube engineering profession died. This is the same with hand calculator, a clerical job usually assigned to women back in the war days. Jobs will come and go. It is the price of progress. However, the notion that humans will be replaced is complete fiction.
Perhaps it is the product of years of science fiction movies filled with robots that rebel and exterminate humans that continues to perpetuate this notion of computers replacing us. While job computerization is happening, it will not continue to reach the point where humans will be gone from the picture. As I and hundreds of other AI researchers say, AI is meant to augment us, not replace us. And how about the answer to the question of “where will these displaced workers go”? Such ethical concern is the responsibility of the company that hires them. We shouldn’t stop AI progress just to keep workers in place. This bars innovation. What we should do is to ensure that companies hold on to ethical processes. If a company purchases a machine that does the job of a thousand people, it is their responsibility to ensure that the displaced workers get compensation, and if possible, be given new jobs that computers will not do. It is purely unethical and inhumane to just kick workers out all of a sudden to introduce computer systems. After all, the purpose of a company is to build opportunities. They don’t exist simply to make money. It is the responsibility of the companies that hire these workers to give them new opportunities when their old ones are taken over by intelligent machines.
We have to stop perpetuating the notion that AI will rebel, go against us, and replace us in the long run. In pushing innovation, humans can never be wiped off from the picture. There are simply things tech cannot do, and jobs that humans are too fragile to accomplish eﬀectively. True innovation can never be done with just pure human teams, nor be achieved solely by technology. It is through making the perfect human-computer teams that this can be done. We shouldn’t bar computers from helping us, or be scared of them replacing us. In the words of my professor, I think it’s time we give computers a chance.