A new study suggests workers are becoming more comfortable with the idea that they will work alongside artificial intelligence and robots in future. (Image: Shutterstock)

Citizens in developed countries are not terribly frightened that artificial intelligence and robots will take their jobs. But they do realize AI is coming, and want more help updating their skills. Meanwhile, senior executives in critical industries believe their companies are doing well preparing for an AI future, including helping their employees adjust.
Those insights emerged from a new research study conducted for global professional services firm Genpact by Wakefield Research. It surveyed 4,000 adults as well as 500 senior executives of large companies in Australia, Japan, the UK, and the U.S. Tiger Tyagarajan, Genpact’s CEO, says the company undertook the research because it sees AI’s transformations as gigantic, positive, and inevitable, but also believes it’s urgent for business and society to do a better job preparing.
Generally, people seem to be getting more comfortable with the concept that AI will be everywhere. In the new survey, 62% of adults say they expect to work comfortably alongside robots in three years. That’s up from 40% who answered the same question that way in the same survey one year earlier. There’s also been a change in how willing consumers are to have AI used in customer service. In the recent study, 54% of people said they would be comfortable if companies used AI to analyze personal data to improve customer service. Responding to the same question in the 2017 study, only 30% said they were comfortable with that idea.
When it comes to jobs, just 28% of workers in the new survey said they worry AI will threaten theirs. The same relatively low number say AI may undermine the relevance of their profession. But that doesn’t mean people are sanguine. Almost half of the respondents said they worry about employment prospects for their children and grandchildren as AI grows even more dominant.
Among executives, 53% said their companies were already providing the reskilling and training that workers need, a confidence that has not been borne out, generally, in separate research about what companies are actually doing. And people clearly want and need more training. The Genpact research finds only 35% of them say reskilling is available at the companies where they work. And just 21% have actually taken advantage of the training that does exist.
To Tyagarajan, while the results are encouraging, there remains a huge amount of work to be done to manage this vast transition. “Our very clear view,” he says, “is that people working with machines, or what we sometimes call human in the loop, is the future. But if employees are saying they aren’t seeing enough education opportunities, then companies, as well as governments and educational institutions, have to work together to teach them.”
Tyagarajan says that while AI can be scary, it can also assist people in the transition to an AI-centric world.  “AI itself will help with the teaching,” he says. “The machine can figure out exactly what you need to learn. We’re already using AI that way at Genpact for our 80,000 people. And the more customized the training becomes, the easier it is for the person to benefit from it.”
The research also uncovered a growing awareness among consumers about the risks of algorithmic bias in artificial intelligence – unfair decisions made by poorly programmed computers. While 78% of consumers in the countries surveyed say it’s important to fight such bias, they’re worried that not enough work is being done by AI developers to prevent it. Two-thirds of people say they think machines may discriminate against them.
The executives surveyed are more confident that problem is in hand. An amazing 95% say their companies are taking steps to combat algorithmic bias. It seems unlikely that so many companies truly have this risk in hand. But more than a third of leaders say their companies are taking one or more of the following steps:
-Establishing internal controls and governance mechanisms to manage the risk of bias
-Employing diverse teams to select data samples and train algorithms
-Modifying algorithms to eliminate bias
-Educating employees about the potential for bias to creep into the software
While on balance the report is encouraging, Genpact’s Tyagarajan is by no means confident the path forward will be smooth. He remains certain that, on balance,  “AI is a good news story.” But he adds, “like everything in life, what is good can also be used for bad. AI can be misused to create monopolies or by governments to wage wars. All kinds of bad things can happen. But the same tech that creates the problem is the one we’ll use to defend against its risks.”
“The speed of change will be more rapid than any other tech change we’ve seen,” he continues. “People with many years of work experience will now have to learn new skills to keep working.”
This revealing new data underscores that companies and societies need to work harder to build systems that take into account both the risks as well as the benefits of AI. They also need to offer workers more training to help them adjust to a world radically transformed by AI.