Soft, pliant fingers extend to delicately lift the apple off the shelf and gently place it in the basket. Then the task is repeated with lemons, then with pepper. The fingers never complain or get tired. More and more companies are turning to smart machines to save on the slow and costly human workers. Automation. What it is? What does this mean for your job?
The manipulator prototype described above is the development arm of Ocado, a UK-based online supermarket. The irregular shape and thin skin of these mediocre products suggests that they are usually packed by humans in Ocado warehouses. But the company is using robotic technology to help these people not only handle products safely, but also do it faster and cheaper for the company.
Ocado is far from the only company with a preference for automated workers. The same thing happens in hospitals, law firms, and the stock market. The list is going to be long.
The question is, how does this affect working people. For example, how can this affect you?
We often hear about grim robots of the future stealing our jobs, but is that so? Who is at risk? What will your workplace be like in five years?
The answers may surprise you.
Middle class at risk
Promotional video:
Research suggests that 47% of people working in the US can be replaced by machines, while in the UK about 35% of jobs could be at risk - and in developing countries, the threat is even higher, as two-thirds of jobs can be automated.
But jobs stealing machines are nothing new. “Automation has happened before,” says Bhagwan Chowdhry, professor of finance at the University of California, Los Angeles. Chowdhry points to the shifts that occurred in factories during the Industrial Revolution, when automatic looms and other machines took over weaving from humans.
What has changed this time? “This will affect not only blue collars, but a lot of whites,” says Chowdhry. Blue collar refers to the working class, and white collars to employees, bureaucrats, administrators, managers.
We often believe that the lowest paid, low-skilled jobs are the most at risk. For example, warehouse workers or cashiers. However, automation can also affect the work of middle-income people such as clerks, cooks, office workers, security guards, junior lawyers, inspectors.
It is clear that the people in the line of fire are worried. “The fears are not only about the transition period,” says Carl Benedict Frey, co-founder of the Oxford Martin Program on Technology and Employment. “Most of the jobs that will be automated require different skills than newly created ones. It will be important to ensure that people out of work can find employment.”
So, should companies that seek automation have a moral responsibility and help their staff learn new skills?
Verification by the future
The answer may not only affect companies - the search for the answer may start at school.
Our modern, structured education may be meaningless in a world where technology is changing so rapidly.
“The concern is that we are not updating our educational, training and policy institutions to keep up,” warns Erik Brunholfsson, director of the Digital Economy Initiative at the Massachusetts Institute of Technology. "We may end up leaving a lot of people."
Brunholffson and Paul Clarke, CTO of Ocado, agree that school and college education should better prepare students for a world in which artificial intelligence and robotics are widespread.
In the workplace, employees will also be required to continually update their skills instead of using the same skills throughout their careers.
For most of us, this can be a critical factor in thinking.
A study by management consultants McKinsey and Company found that less than 5% of professions can be fully automated by existing technology. Simply because our workplaces are too diverse and volatile for robots to do all the tasks.
Instead, according to their forecasts, about 60% of professions will be automated by a third. This means that most of us will be able to cling to our jobs, but the very process of our work will change.
Robots will complement, not replace
Learning how to work side by side with robots will be critical.
“There are times when machines take over repetitive work to free people to do other, more rewarding aspects of their job,” explains James Mannica, a senior partner at McKinsey who has done much of his research on the impact of automation. “This can make a big difference to the wage determination process because the machine will do all the hard work. It also means that more people will be able to do this work with the help of technology, so the competition will increase."
There are also broader issues. With lower middle-class incomes, governments may face such fundamental problems as lost taxes and a dissatisfied electorate.
Fortunately, there are many things that machines cannot do yet.
One good example is the work of researchers in Singapore trying to train two autonomous robotic arms to assemble flat chairs from Ikea. Despite the use of modern equipment, the machines cannot cope with the simplest tasks.
Even isolating different objects from a chaotic mixture of parts is a daunting task for robots. In recent tests, it took two robots over one and a half minutes to successfully insert a piece of tongue into one of the chair legs.
And that's just one piece of furniture. “The real problems start when you want the robot to assemble multiple pieces of furniture. A robot could assemble an Ikea chest of drawers, but it will not be able to assemble a wardrobe from the same series, since the details will be different, even if some of the assembly steps remain the same. People don't have that problem."
Human advantage
From increased flexibility to better personality, there will always be things we are better at than robots.
“As we automate repetitive work, we see a growing demand for creative skills,” says Brunholfsson. “We also see a growing demand for people with social skills, interpersonal skills who nurture, care, teach, impose their beliefs, have negotiation skills and sell well.”
Frey believes there are several areas where humans will have an edge.
“The first is social interactions,” says Frey. “When you think about the variety of complex social interactions we encounter every day when we negotiate or try to convince people, help others, or take care of customers … We run teams and all that. It's incredible that computers can replace human workers who do all this.”
Another is creativity. Computers are great at dealing with problems and not getting bored with repetitive actions. However, people find such monotonous work tiring.
An initiative by the Massachusetts Institute of Technology has set a $ 1 million target to encourage businesses to make the most of these "human traits" along with technology.
“The amount we currently pay nannies and caregivers for the elderly is very low,” says McKinsey's Maniika. “Likewise, there are many artistic and creative works that have never been paid for. The challenge is to pay and appreciate creative work as it deserves, because a machine will never be fully capable of it."
Alex Harvey, chief scientist at Ocado Technology, which develops software and technologies for the company's retail division, notes that the world was designed and built for humans, and making robots function in this complex natural environment is a major technical challenge.
One of Ocado's projects is a maintenance assistant robot called SecondHands. It shows how humans and robots could work together.
“For example, he has the ability to lift things to a greater height than a human,” explains Harvey. "This is a fairly simple robot in terms of its behavioral repertoire, but it can form a great team in which a human technician will be the leader and they can use the muscle strength of the robot."
But the closer humans and machines work together, the darker the ethical side will be.
Ethics issues
About 1.7 million robots are already in use around the world, but most often in industrial environments where people are virtually denied entry. The numbers are growing, and so are the roles that robots play. It turns out that people will have to work side by side with them, and the risk will rise accordingly.
“There needs to be more transparency so that we can understand how these things do what they do and how they behave,” said Madi Delvaux, deputy chair of the European Parliament's legal committee.
Recently called on parliament to create rules for robotics and artificial intelligence.
A report prepared for the European Parliament highlighted the urgent need for new legislation on liability in the event of accidents. Similar liability issues arise if a robot takes action that violates the law. An artificial intelligence algorithm, for example, can skip a series of financial transactions bypassing the intricate web of rules governing a sector.
Delvaux and her colleagues are also calling for a code of ethics that will regulate our relationship with robots.
“There should be points that require respect, like human autonomy and privacy,” Delvaux said.
All of this highlights another problem that worries many artificial intelligence developers: bias. Machine learning systems are only as good as the data they are given to learn. Recent research has shown that artificial intelligence can develop sexist and racist tendencies.
Bill Gates, meanwhile, recently proposed taxing robots to compensate for forgone employee income taxes. Others have suggested that as robots take on more and more tasks, there may be a need for a universal basic income in order for everyone to receive government benefits.
Cars are moving forward
On the example of furniture from "Ikea" it becomes obvious that AI still needs to develop for a long time.
Perhaps one of the biggest challenges facing machine learning and artificial intelligence is understanding how their algorithms work. "Things like artificial intelligence and machine learning are, for the most part, black boxes," says Manieka. "We cannot open them to find out how they got the answer they got."
This creates a number of problems. Machine learning systems and modern AI are usually trained using large sets of images or data that are fed to the system to learn to recognize patterns and trends. They are then used to identify similar patterns as new data is entered.
This can be good if we need to find CT scans showing signs of disease. But if we use a similar system to identify a suspect from a snapshot, we need an understanding of how the algorithm works to provide irrefutable evidence.
Even in the field of autonomous vehicles, this generalizability remains a major problem.
Takeo Kanade, professor of robotics at Carnegie Mellon University, is an expert in self-driving cars and computer vision. Giving robots a "true understanding" of the world around them, he says, is still a technical challenge that needs to be overcome.
“It's not just about identifying the location of objects,” he explains. “Technology needs to understand what the world around it is doing. For example, does a person want to cross the road or not?"
ILYA KHEL