As the AI revolution proceeds, which nations will benefit? Of course, artificial intelligence will be good for nations that make and sell it, with the US taking a clear lead. But what about the effects everywhere else? How will AI alter the trajectories of economic growth?
Broadly speaking, most economic endeavours fall into one of two categories: those with known routines, and new projects. AI will favour nations that excel at the latter and hurt those which rely on the former.
Most activity falls into the routinised category — it describes a lot of bureaucracies, attempts at medical diagnosis, back-office work, and so on. To be clear, this is not a criticism: Routinising activities lowers their costs. New projects — startups, attempts to build new towns or cities, trying to establish a colony on Mars, founding a new university — are different.
To the extent activities are replicable, AIs stand a good chance of mastering those tasks over time, perhaps in conjunction with the right physical machines. AIs can capture the data associated with such activities and figure out how to copy the routine, just as automated bots are widely used now for customer-service inquiries. The current instantiation of GPT-4 may not be good enough to do all these jobs, but improvements in AI have been coming rapidly and will likely continue.
So, to the extent a country specialises in providing routine activities, such as the call centres and back-office support provided by firms in India and the Philippines, AI presents a risk. It could take away many of those jobs and shift the associated profits to foreign firms.
Since so many economic activities involve replicable activities, the eventual scope of AI in this regard could be quite high, although the timing of these transitions remains uncertain. My biggest worries are about smaller, poorer countries that lack the resources to pull off major projects. It will be hard for them to climb the value chain by selling routinised activities at low wages.
In contrast, consider new projects. Current AI models are not anywhere close to being able to conceptualise a new idea, communicate a new vision, assemble and inspire the necessary talent, raise money and deal with the corporate politics — to name just a few important components of new projects. So, AI cannot substitute for the essential creative forces of entrepreneurs.
That said, AI makes many new projects easier to pull off by aiding with the routine work along the way. Say you have a brilliant new idea for a fintech firm, but need help with the slide deck and marketing copy and all the email inquiries. AI will be of use to you.
So, countries and regions that are good at executing new projects are the most likely to benefit from the AI revolution. Which countries might those be?
One possible candidate is China, which has successfully carried out a large number of infrastructure projects. But there is a tension between free-flowing commercial AI and the Chinese government's policy of censorship. What if someone asks the AI some political questions that the regime is not so keen to see discussed? Exactly how much will the Chinese government allow or encourage decentralised access to quality AI models?
India is another possible winner, even though it is vulnerable in the area of back-office support. Indian infrastructure has improved by leaps and bounds in the last 10 years, a sign the nation now has greater ability to pull off new projects. The Indian Aadhaar programme, which has done bio-scans of well over 1 billion Indians and helped them make and receive payments, was a major new project that largely succeeded. India has some censorship issues as well, although they are not as serious as China's.
Saudi Arabia is planning some major projects, such as the ambitious desert city of Neom. Perhaps the Saudis will need yet further technological advances to pull off those plans, but at least they are trying to make some significant changes. They are possibly a big winner from AI advances.
How about Canada, or for that matter Greenland? Climate change may bring about new real-estate developments in those regions, with possible assistance from the US. The US itself is likely a major beneficiary, above and beyond its role in producing and selling the AI itself. Whether the area is biomedicine or green energy, evidence abounds of the US's newfound ambition.