On November 30, 2022, OpenAI launched the AI chatbot ChatGTP, making the latest generation of AI technologies widely available. In the few months since then, we have seen Italy ban ChatGTP about privacy issues, which are requested by leading tech stars a pause in the development of AI systemsand even leading researchers who say we should be prepared to do so launch air strikes on data centers associated with rogue AI.
The rapid deployment of AI and its potential impact on human society and economies is now clearly in the spotlight.
What does AI mean for productivity and economic growth? Will it usher in an era of automated luxury for all, or simply reinforce existing inequalities? And what does it mean for the role of man?
Economists have been studying these questions for many years. My colleague Yixiao Zhou and I examined their results in 2021, and found that we are still a long way from definitive answers.
The big economic picture
Over the past half century or so, workers have gotten around the world a smaller fraction of their country’s total income.
At the same time, productivity growth – how much output can be produced with a given amount of inputs such as labor and materials – slowed down. During this period there have also been tremendous developments in the creation and implementation of information technologies and automation.
Read more: The internet has done a lot, but so far little for economic growth
Better technology should increase productivity. The apparent failure of the computer revolution to deliver these benefits is a puzzle economists call the Solow Paradox.
Will AI save global productivity from the long slump? And if so, who will reap the benefits? Many people are curious about these questions.
While consulting firms have often portrayed AI as an economic panacea, policymakers are more concerned about possible job losses. Economists, perhaps unsurprisingly, are more cautious.
Radical change at a rapid pace
Perhaps the biggest source of caution is the sheer uncertainty surrounding the future trajectory of AI technology.
Compared to previous technological leaps – such as railways, motorized transport and, more recently, the gradual integration of computers into all aspects of our lives – AI can spread much faster. And it can do this with much lower capital investment.
The application of AI is largely a revolution in software. Much of the necessary infrastructure, such as computing equipment, networks and cloud services, is already in place. There’s no need for the slow process of building out a physical rail or broadband network – you can now use ChatGPT and the burgeoning horde of similar software right from your phone.
It is also relatively cheap to use AI, which greatly lowers barriers to entry. This is related to another major uncertainty surrounding AI: the scope and domain of the effects.
AI seems likely to revolutionize the way we do things in many areas, from education and privacy to the structure of global trade. AI may not just change individual elements of the economy, but rather its broader structure.
Adequate modeling of such a complex and radical change would be extremely challenging, and no one has done it yet. But without such models, economists cannot make clear statements about the likely impact on the economy at large.
More inequality, weaker institutions
While economists have differing opinions on the impact of AI, there is general agreement among economic studies that AI will increase inequality.
A possible example of this could be a further shift of advantage from labor to capital, which would gradually weaken labor institutions. At the same time, it can also reduce tax bases, weakening the government’s redistributive capacity.
Most empirical studies find that AI technology will not reduce overall employment. However, it is likely that the relative income going to low-skilled labor will fall, which will increase inequality in society.
Read more: The benefits of task automation are unlikely to be shared equally
In addition, AI-induced productivity growth would lead to job redistribution and trade restructuring, further increasing inequality both within and between countries.
As a result, controlling the rate at which AI technology is adopted is likely to slow the pace of societal and economic restructuring. This allows for a longer adjustment period between relative losers and beneficiaries.
In light of the rise of robotics and AI, there is an opportunity for governments to reduce income inequality and its negative impacts with policies aimed at reducing inequality of opportunity.
What is left for man?
The famous economist Jeffrey Sachs once said
What humans can do in the AI era is just be human, because this is what robots or AI can’t do.
But what exactly does that mean? At least in economic terms?
In traditional economic models, people are often synonymous with “labor”, and at the same time also an optimizing factor. If machines can not only do work, but also make decisions and even come up with ideas, what is left for humans?
The rise of AI challenges economists to develop more complex representations of humans and the “economic actors” that inhabit their models.
As the American economists David Parkes and Michael Wellman have done noted, a world of AI agents may actually behave more like an economic theory than the human world. Compared to humans, AIs respect idealized assumptions of rationality better than humans, interacting through new rules and incentive systems very different from those tailored for humans.
Importantly, a better understanding of what is “human” in the economy should also help us think about what new characteristics AI will bring to an economy.
Will AI bring us some kind of fundamentally new production technology, or will it tinker with existing production technologies? Is AI just a substitute for labor or human capital, or is it an independent economic agent in the economic system?
Answering these questions is essential for economists – and for understanding how the world will change in the coming years.