AI could supercharge offshoring

The double hit of AI superpowers plus low-cost labor could mean that the types of roles at risk of being offshored shift from repetitive tasks like data entry, which have traditionally been gestured to as actions AI can replace, to meatier work like prompt engineering, high-end customer service, and marketing, industry experts told Business Insider.

Andrew Yeung, a former product lead at Google and Meta, predicted in May that overseas workers who get their AI glow-up will someday take over numerous jobs.

“In a few years, every scrappy founder and high-powered executive is going to have an army of offshore talent equipped with AI tools that completely replace the need for traditional engineers, designers, marketers, and assistants,” he wrote on X.

His thesis seems to be backed up by Sagar Khatri, cofounder and CEO of Multiplier. The company’s platform lets employers hire workers from anywhere by automating functions like payroll and labor law compliance. He told BI that companies will face more pressure to go global — not to boost sales, but to find the skilled workers they need.

Thanks to productivity gains from AI, a company in New York might more easily hire accountants in the Philippines, customer-success workers in Mexico, and customer support teams in India, Khatri said.

“With AI, a customer support agent can do their job much better,” he said.

Everyone’s getting schooled

Online learning is making it easier for workers abroad, especially young ones in developing countries, to build their skills, according to Jeff Maggioncalda, CEO of Coursera.

He told BI that workers will increasingly face competition from others overseas who are using technology to automate and be more efficient.

And it’s not just learning about AI, but learning from AI. Maggioncalda said the technology can help workers in emerging areas get up to speed faster and compete with those who command higher salaries in developed economies. For many workers, a big factor will be what he calls “talent agility.” Essentially, it’s how fast workers can learn new things to add value to business models.

“It sounds kind of crass, but that’s what it’s going to come down to,” Maggioncalda said.

He added that a worker in a low-cost region who might have previously taken five years to become as effective as someone in a more expensive job market can now do so far sooner. The person overseas doesn’t necessarily even have to be better, Maggioncalda said.

The threat simply comes from “someone who matches you but costs less,” he said.

Maggioncalda added that it’s easier to hire, fire, and move workers in many developing markets. Many workers in these markets, he said, are young and hungry to gain skills that will help them build their careers. Then, you give them AI.

“Now they have a tool that has a differentially positive impact on their productivity compared to someone who’s at the higher end,” he said. “The only other question is, how fast is this going to happen?”

Based on Coursera’s numbers, it could be soon. In 2023, the company enrolled a person every minute in a Gen AI class. In 2024, it’s drawing four people every minute, he said.

Fifty-two percent of enrollments in Coursera’s Gen AI classes are from emerging markets like India, Pakistan, Brazil, Vietnam, and Egypt.

AI is also making it cheaper to bring lessons on AI and other topics to people in languages other than English. Two years ago, it cost Coursera $10,000 to translate a course into another language. Now, with GenAI, the company can do it for $20, Maggioncalda said.

The massive drop in cost has enabled Coursera to translate 4,500 courses into 22 languages, he said.

“Everybody can now learn this stuff because language is not a barrier anymore,” Maggioncalda said.

AI can deliver the real answer

Multiplier’s Khatri said AI is making it easier for a bigger pool of workers in countries enjoying demographic tailwinds to get trained to become software developers and other lucrative roles. That’s helpful, he said, for companies operating in countries with aging populations like the US, the UK, Japan, or Germany.

Khatri also said AI could make companies less hesitant to look overseas to fill needs like customer service reps because the technology can quickly serve up a “real answer” that an agent can give a caller. That can make the conversations shorter, which, in turn, reduces the risk that customers will hang up feeling dissatisfied by how hard it was to get the information they needed or because of hurdles like language.

That’s because even if AI provides the answer, “customers still want to talk to a human being,” he said.

Daron Acemoglu, an institute professor in the economics department at the Massachusetts Institute of Technology, told BI that AI could open up parts of the world where language barriers might have kept regions from becoming offshoring hotspots like India or the Philippines.

“Workers in Indonesia couldn’t do some services because they’re not fluent in English. Perhaps with better AI translators, they can do that,” he said.

Acemoglu also warned that there is a danger of AI becoming so good that it undercuts the need for offshore workers.

“It has to be a sort of rather narrow path — that AI is good enough to do the translation but not good enough to do these pretty low-skilled tasks,” Acemoglu said.

‘This isn’t the cotton gin’

Scott Vincent, CEO of Digital Futures, a UK company that helps people of varied backgrounds land roles in tech, told BI that the global trial of remote work brought by the pandemic proved to companies that they don’t need all workers to be together to remain productive. As a result, he said, more organizations are looking at lower-cost locations to tap into workers.

Vincent said he’s seen “a real acceleration” of offshoring offerings across industries since the pandemic.

“It poses a significant threat to the labor market,” he said.

The threat doesn’t stop with workers in expensive markets like the UK or the US getting undercut by someone in a developing market, Vincent said. Offshoring can make it harder for companies to produce home-grown talent that can rise through the organization, he said.

“Generative AI impacts, or can replace, a lot of foundational skills that are learned at the start of one’s career. So you’ve got this double whammy,” Vincent said.

Many big companies, he said, see Gen AI as having two horizons. One is the immediate impact on the workforce — the sugar rush of productivity gains. But the other, more consequential effect is what he sees as an “outflow of human capital” due to AI’s eventual ability to do much more than it can now.

“It is an exponential trend in terms of the pace at which it’s moving,” he said.

The time it took for earlier technologies — including automation and robotics — to rejigger the labor market was longer than what we’re seeing with Gen AI, Vincent said.

He expects companies’ spending on overseas labor will grow. Digital Futures examined spending from the top 100 publicly traded companies in the UK and found that, on average, each spent about £750 million (about $951 million) a year on offshore offices. Vincent said that works out to about 1.2 million jobs in the UK and about £16 billion in lost tax revenue.

But short-term gains overseas may not lead to long-term jobs. Drew Cesario, who consults on revenue operations and marketing systems and is the founder of Botanical Grp, told BI that the fate of some offshore workers who might work alongside AI is likely similar to that of the minders who sat in self-driving taxis during testing phases. “They are training themselves out of a job,” he said.

Because of how many tasks AI can take on, Cesario expects there will be widespread job displacement both domestically and internationally.

“I would love to say that there isn’t, but this isn’t the cotton gin,” he said. “It is a general technology versus a specific technology.”

How you can use tech

Vincent said governments in developed economies and businesses need to work together to deal with the coming changes to the job market. Businesses that say they’re offshoring in part because they can’t find the skilled labor they require need to see improvements in education systems to produce better-prepared workers, he said.

Governments, Vincent said, could consider regulating the percentage of a company’s workforce that can be offshore.

Coursera’s Maggioncalda noted that if you buy a car from another country, you often have to pay a tariff.

“Maybe if you pay wages to another country, you have to pay a wage tariff,” he said.

Maggioncalda said it could be a way to help even out the cost of labor.

In the meantime, he said, workers need to consider how they can “add real value in a world where more and more pieces of your job get automated.”

“What you need to be thinking about is, ‘How can I use technology to automate certain parts of my job?’ Maggioncalda said. “Because if you don’t do it, someone else is going to do it someplace else in the world.”