By the time Brad Smith, vice-chair and president of Microsoft, took the stage in Singapore yesterday, the narrative around artificial intelligence had already shifted from just months ago.
The “if” of AI will transform the global economy has been replaced by “how”. Specifically how evenly and how quickly that transformation will spread.
“AI will be perhaps the most defining technology of the next 25 years,” Smith said, framing it not as another digital tool, but as a general-purpose technology on par with electricity or the steam engine.
Beneath the AI buzz lies a more uncomfortable truth, though. History suggests that transformative technologies do not lift all boats equally.
Economists have long observed that technologies like electricity and computing drove not just productivity, but prosperity, creating divides between those who had access, and those who did not.
The same pattern is emerging again, said Smith, in a discussion in front of about 600 public and private sector leaders on the development and impact of AI in Singapore.
The “global north” – advanced economies across North America, Europe and Japan – has historically been decades ahead of the “global south” in adopting transformative technologies.
In the case of electricity, the gap stretched to nearly 70 years, reflecting not just differences in wealth, but in access to infrastructure, he noted.
“What we think of as an economic divide was, in fact, a technology divide,” he said, pointing to electricity as a foundational layer of infrastructure that underpinned modern growth.
Today, that same pattern is re-emerging in AI where disparities in infrastructure investment are once again shaping who benefits, and who risks being left behind.
While Singapore ranks among the world’s leaders, with about 67 per cent of its workforce actively using generative AI, much of Southeast Asia and the broader global south is lagging, with adoption rates remaining nearly half those of developed markets.
“If we don’t address it,” Smith warned “we risk widening the gap even further.”
The Microsoft leader’s visit to Singapore was timed with a reminder from the company that it was on track to spend US$5.5 billion commitment to expand its AI infrastructure in Singapore.
As part of this commitment, it is also enabling every tertiary education student in Singapore to have free access to Microsoft 365 Premium with Copilot and the company’s suite of productivity tools.
All educators will be provided free AI training through Microsoft Elevate for Educators, while Microsoft Elevate for Changemakers will upskill non-profit leaders.
There has been tremendous focus on infrastructure development but it is not the only critical factor, stressed Smith. “Skilling is equally important,” he stressed. “You need people across the economy who know how to use the technology.”
The challenge is not just access to AI, but the ability to work with it. In Europe, there are roughly 14 data scientists per capita for every data scientist found in Africa.
“In some ways, we need to help accelerate the education of people in the fields that play an important role, and that we need to spread skilling much more broadly,” said Smith.
For business leaders, the implications of AI fluency among the workforce is critical. Every job can be broken into three categories of tasks – tasks AI can do on its own, tasks humans do better with AI and tasks only humans can do, explained Smith.
The strategic question is not whether AI will replace jobs, but how it will reshape them, he added. “The opportunity is to let AI do what it can and shift human effort to what only humans can do.”
That shift is already redefining what matters at work. Drawing on research highlighted by leaders in Linkedin, a Microsoft-owned firm, Smith outlined what he calls the “five Cs” of the AI era – curiosity, creativity, communication, compassion and courage.
“These things will be even more important in an age of AI than they were without it,” he argued. From these 5Cs, he highlighted curiosity, which he said ought to be combined with humility to create a specific mental attitude for workers.
In a world where AI can generate answers in seconds, the value of a human shifts from “the person with the answer” to “the person with the best question”, he argued.
This requires an insatiable curiosity to keep poking at the boundaries of what is possible. However, curiosity without humility is just ego.
Smith noted that people have to be humble enough to admit that the “old way” of doing things, even if it served us well for several decades, might now be the slow way.
It is about having the humility to realise that a 22-year-old developer might have a more intuitive grasp of AI orchestration than a veteran executive, and the latter being curious enough to pull up a chair and learn from them.
Another insight from Smith’s remarks is that the current gap between AI capability and enterprise adoption is not unusual. “The innovation curve accelerates much earlier than the adoption curve.”
Even in the United States, home to the world’s most advanced AI models, adoption ranks only 23rd globally. The reason is a familiar one – uneven diffusion between urban and rural populations, said Smith.
Technology may be built at the frontier. But it is adopted in the real world, by people, organisations and systems that take time to change.
For enterprises, this creates a paradox. AI capabilities are advancing exponentially, while organisational change moves linearly. The result: a widening gap between what AI can do and what businesses actually use it for.
Adoption, argues Smith, ultimately comes down to whether organisations make AI approachable and even enjoyable. While infrastructure and skills form the AI foundation, it is use culture that is the accelerant, he stressed.
At Microsoft, internal adoption is gamified through initiatives like the “Copilot Cup,” rewarding teams that improve usage the most each quarter. Leaders share personal use cases. Weekly tips are circulated across the company.
The goal is simple: remove intimidation. Said Smith: “You have to create a culture that makes it fun to work with AI.”
This is especially critical for senior workers, who may need to “unlearn” established ways of working. Smith offers a practical solution – learn from younger colleagues who never built those habits in the first place.
“The best way to unlearn is to learn from people who never learned it the old way,” he said.
Beyond economics and productivity, a more fundamental question is emerging: is AI meant to outperform humans or to empower them?
“There are some people who talk about building machines that outperform people in almost everything,” Smith said. “Or is it to build machines that help people do more?”
At Microsoft, the answer is clear, he said. It is echoed by the company’s CEO Satya Nadella, that technology should augment human capability, not replace it.
“It is to bet on people,” said Smith, who ended his 15-minute talk using an anecdote not from Silicon Valley, but from a classroom in India.
A history teacher, told it was “impossible” to stage a complex Indian theatrical production, turned to AI. With his students, he used AI tools to write a script, design sets, compose music and produce a full play.
“They did what they had been told was impossible,” he pointed out. “They did it with the help of AI but they did it themselves.”
The real race is to make AI usable by everyone. AI will define the next phase of economic growth but the winners will not be those who build the most powerful models alone, said Smith.
They will be those who close the gap between innovation and adoption through infrastructure, skills, culture and trust, he added.
