BANGKOK: Behind the hoopla over the promise of artificial intelligence lay difficult realities, including how such technology might affect people already disadvantaged in a data-driven world.
A new report by the United Nations Development Program notes most of the gains from AI are likely to be reaped by wealthy nations unless steps are taken to use its power to help close gaps in access to basic needs, as well as such advanced know-how.
The report released Tuesday likens the situation to the “Great Divergence” of the industrial revolution, when many Western countries saw rapid modernization while others fell behind.
Questions over how companies and other institutions will use AI are a near universal concern given its potential to change or replace some jobs done by people with computers and robots.
But while much of the attention devoted to AI focuses on productivity, competitiveness and growth, the more important question is what it will mean for human lives, the authors note.
“We tend to overemphasize the role of technology,” said Michael Muthukrishna of the London School of Economics, the report's main author, told reporters. “We need to ensure it’s not technology first, but it’s people first,” he said, speaking by video at the report’s launch in Bangkok.
The risk of exclusion is an issue for communities where most people are still struggling to access skills, electric power and internet connectivity, for older people, for people displaced by war, civil conflict and climate disasters. At the same time, such people may be “invisible” in data that will not take them into account, the report said.
“As a general-purpose technology, AI can lift productivity, spark new industries, and help latecomers catch up,” the report says.
Better advice on farming, analysis of X-rays within seconds and faster medical diagnoses, more effective weather forecasts and damage assessments hold promise for rural communities and areas prone to natural disasters.
“AI systems that analyze poverty, health, and disaster risks enable faster, fairer, and more transparent decisions, turning data into continuous learning and public value,” it says.
Still, even in wealthy nations like the United States, the potential for data centers to devour too large a share of electricity and water has raised concerns. Ramping up power generation to meet higher demand may hinder progress in limiting the emissions of carbon from burning fossil fuels that contribute to global warming, while also causing health hazards.
The technology raises ethical, privacy and cybersecurity concerns: researchers have found hackers using AI to automate portions of cyberattacks. There also is the problem of deepfakes that can misinform or facilitate criminal activity.
Asian nations including China, Japan, South Korea and Singapore are well placed to take advantage of AI tools, the report notes, while places like Afghanistan, the Maldives and Myanmar lack skills, reliable power and other resources needed to tap into the computing potential of AI. Inequalities between regions within countries mean some places even in advanced economies are prone to be left behind.
About a quarter of the Asia-Pacific region lacks online access, the report says.
If such gaps are not closed, many millions may be excluded from the kinds of devices, digital payment systems, digital IDs and education and skills that are required to participate fully in the global economy, falling further behind, said Philip Schellekens, the UNDP's chief economist for the Asia Pacific.
Other risks include misinformation and disinformation, surveillance that violates rights to privacy and systems that can act as “black boxes,” reinforcing biases against minorities or other groups. So transparency and effective regulations are crucial guardrails for ensuring AI is used in fair and accountable ways, he said.
“We believe we need more balance, less hysteria and hype,” Schellekens said.
AI is becoming essential for modern life, like electricity, roads, and now the internet, so governments need to invest more in digital infrastructure, education and training, fair competition and social protections, the report says.
“The goal," it says, "is to democratize access to AI so that every country and community can benefit while protecting those most at risk from disruption.”