Artificial intelligence is changing the international trade conversation. Many countries see AI as a boon for improving productivity, while others feel threatened by potential data breaches and cuts to the workforce. Jason Grant Allen is the director of Singapore Management University’s Center for AI and Data Governance. He explains the impact AI can have on trade [Transcript | Video].
4 takeaways:
➀ AI breaking barriers to trade.
“It’s potentially really open-ended, from contract automation and data analytics to logistics and supply chain optimization,” Allen said. AI-powered translation services can facilitate trade between countries with language barriers. One of the appeals of AI is that it combines with other technologies. “It doesn’t necessarily make sense to talk about any one type of AI system in isolation,” Allen said. “They interact with each other and they interact with other technologies.”
➁ Data protection can slow AI development.
In recent years, data control has been restricted to the countries of origin. This data localization—which is most stringent in China—can limit AI’s role in trade. “Data access is really crucial for AI development, whether based on symbolic logic or these neural network architectures, you need training data,” Allen said. Good data is necessary to train AI, and “to get better data…more data is good,” Allen said. Some countries are pushing for bilateral data sharing. Singapore has fostered data-sharing agreements with Australia, the UK, New Zealand and Chile. “We’ve got a big fault line going through this region of the world, and AI is definitely a frontline in that,” Allen said. “It implicates not just the software systems and the products in which they’re embedded, but the whole supply chain including for example, semiconductors and the hardware tech on which these systems rely.”
➂ Beware of ‘doomsday predictions’ for labor forces.
“This is where it’s good to be careful about the hype cycle,” Allen said. “There’s a lot of doomsday predictions.” Journalism, law and blue-collar manufacturing jobs are “on the chopping block” as AI develops, but the introduction of AI isn’t necessarily fatal for these industries. “We have to think carefully about the distribution of benefits and burdens,” Allen said. “So, we’ve got this potential productivity windfall from AI, how do we distribute that between capital and labor and governments?”
➃ AI may boost domestic productivity.
AI may mean domestic production is more realistic for countries. AI could have more impact on sectors like textile and automotive manufacturing. Countries may also be able to bring back formerly outsourced services, like call centers, through AI. “This ties in again, to this sort of geopolitical fragmentation that we’re seeing, and this desire in many countries towards…self-sufficiency, and less reliance on complex international supply chains,” Allen said.
*This fellowship is part of an ongoing program of journalism training and awards for trade coverage sponsored by the Hinrich Foundation. The National Press Foundation is solely responsible for the content. All programs are on the record and resources and transcripts from this and previous fellowships are published.
