From OUR EDITORIAL STAFF we restart the course with a very interesting topic that is on everyone’s lips, in the minds of business leaders, as well as politicians who have to define political action programs, in addition to consultants, professors and especially very dedicated lately any student who is taking a postgraduate course. It is obvious that we are referring to AI (Artificial Intelligence). As is our style, we have chosen for today some very outstanding contributions from authors and also companies and/or consultants that always mark the direction of where the actions of organizations and institutions should go, at a time when technological disruption gives no respite.
AI revolution: rise of productivity and beyond
This contribution corresponds to the Barclays economic research team, which we detail below:
– Christian Keller, who is Director of Economic Research at Barclays and leads a global team that covers developed and emerging markets. He is based in London and joined Barclays in 2007 from the International Monetary Fund (IMF), where he had worked on programmes with emerging market economies in Europe, Latin America and Asia. He graduated with a PhD in Economics from the University of Cologne, Germany, and holds a joint Masters in Economics and Finance from the University of Cologne and HEC, Paris.
– Mark Cus Babic, is an economist in Barclays’ European Economic Research team, based in London. His areas of expertise are the euro area real business cycle and macroeconomic coverage of Germany. He joined the bank in 2020. Mark holds an MSc with honours in Economics from the University of Edinburgh.
– Akash Utsav, who is also an economist in Barclays’ Global Economics Research team. He joined the bank in 2017 and contributes to thematic notes on global macroeconomics as well as quarterly publications. He holds a Masters in Economics from University College London (UCL), London, UK, and completed his Bachelors at Loyola College, Chennai, India.
The world operates in a digitalised economy, but information technology (IT) has not provided the long-term boost that many had hoped for.
In fact, productivity growth has been in decline since the 2000s. In partnership with the IBM Institute for Business Value, Barclays Research is exploring the potential of artificial intelligence to drive a productivity boom that will power economies in the future.
There has been growing enthusiasm for “generative” artificial intelligence
which uses powerful computer models to produce high-quality text, images and other content, based on the data they were trained on.
Many are wondering whether this technology could mark a turning point in labour productivity, similar to the inventions of the steam engine, electricity and the personal computer.
Historically, positive effects on productivity have lagged behind the invention of new technologies, but our research analysts, working in collaboration with analysts at the IBM Institute for Business Value, see good reasons for optimism when it comes to GenAI’s potential to drive growth.
On the one hand, the basic technology is accessible to a very broad audience on an infrastructure that already exists
A user can give instructions to a tool like ChatGPT, the bot developed by OpenAI and launched in November 2022, without having to learn any special programming language.
On the other hand, these tools are not limited to any particular task, function, problem, or sector. This makes them usable across different disciplines.
Once a large language model is trained on a body of text, for example, it can summarize a legal document just as well as a medical document or an insurance document. Few occupations are likely to remain unaffected. As Stanford scholar Jerry Kaplan said several years ago: “automation doesn’t see the color of your collar.”
Providing a genuine boost to the production of goods and services.
Those two basic attributes – accessibility and versatility – suggest that a broad implementation of GenAI could encounter fewer obstacles than previous technological advances and thus provide a genuine boost to the production of goods and services.
However, to ensure that we unleash the full potential of AI technology and limit any of its negative effects, the right mix of policies will need to be put in place, both from a regulatory and business point of view.
Two of the biggest challenges for the global economy in the coming decades are, on the one hand, ageing populations in advanced economies and, on the other, low productivity in developing economies. AI could help on both counts.
In countries like Japan, Germany and Italy, for example, labour forces are shrinking fast enough
Of course, this is going to require huge leaps in labour productivity, just to maintain the levels of GDP growth that prevailed before the pandemic.
But such advances are possible. According to our analysts’ estimates, most countries would need to achieve similar levels of labor productivity growth to those they achieved between 1990 and 1994 to regain pre-COVID average GDP growth rates by 2033.
In the emerging world, the picture is different: working-age populations are still expanding, overall, and in some cases very rapidly. But skill and education levels tend to be limited, on average, compared with advanced economies, manifesting in low GDP per hour worked.
In addition, economists are observing what some have described as “premature deindustrialization”
in which developing nations no longer experience the industrialization that typically resulted in large productivity and real income gains, as workers shifted from agriculture to manufacturing.
With AI, however, that productivity-boosting effect may now be possible, if those workers move into AI-assisted service industries. “Service-ization” could take over the role that industrialization played in the past.
Balancing AI-driven productivity potential with societal needs and security
Policies adopted by businesses, industries, and regulators will have a major influence on whether the promised benefits of AI are delivered and how those benefits are distributed. Cost will also be a critical factor, as acquiring data and computing power (and the energy it requires) is not cheap.
Our analysts argue that two guiding principles are especially crucial in the development of AI. The first is that technology be used as a complement to labor, rather than a substitute for it.
A recent survey by the IBM Institute for Business Value indicated that the primary goal of companies is to enable a greater focus on uniquely human talents such as creativity, social and interpersonal skills, and empathy. In that context, critical skills include time management, the ability to prioritize, and an affinity for working in teams.
Second, it is vital that policies encourage the spread of AI across the economy
Regulation around access to data to train or deploy specialist technology is likely to play a critical role, and issues around security, privacy and ethics need to be addressed.
Our analysts are encouraged by initiatives such as the AI Alliance, a global network of technology companies, universities, non-profits and government groups that have come together to, in their words, “responsibly maximise the benefits for people and society everywhere.”
New business models and workflows
The opportunities offered by AI appear broad. However, fully realising its benefits is likely to require a very human response – a collaborative effort by industries and regulators, and a complete reimagining of business models and workflows.
How AI can boost productivity and fuel growth
This contribution is from JPMorgan Bank’s Economics and Markets research section. The authors are Joe Seydl, Senior Markets Economist and Jonathan Linden Executive Director, Senior U.S. Equity Strategist
Optimists say AI is a revolution. Skeptics respond that it is a bubble.
We believe that the impact of AI could be truly transformative, as we discuss in our Mid-Year Outlook for 2024. The path ahead is uncertain, but there are powerful forces that could drive it forward.
In this article, we apply an economist’s perspective to a question at the heart of the debate between optimists and skeptics: how might AI affect the broader economy?
Among the questions we address:
– How much productivity gains could AI bring? And how quickly could it come?
– Which jobs could be displaced, and how might policymakers respond?
– Will AI be inflationary or disinflationary?
– What will be the pace of corporate adoption?
Along with a macro perspective on the potential of AI, and generative AI in particular, we also consider the prospects for investing in this powerful trend.
Although AI stock valuations have performed well, we see no signs of a bubble.
And while AI-related companies now account for a relatively high proportion of the overall US market, we don’t think market concentration is necessarily a cause for concern.
In short, you can find a wide range of AI-related investment opportunities to consider now and in the years ahead.
Productivity and growth: how soon, how fast?
When assessing the economic prospects for AI, productivity is a key metric. Faster productivity growth allows the economy to grow faster and living standards to rise more quickly, without generating excessive inflationary pressures. The U.S. economy has not seen sustained productivity gains since the 1990s. A repeat of 1990s-style productivity gains could usher in a new era of economic growth.
To understand how AI might (or might not) transform the economy, we consider historical precedents.
In his latest letter to shareholders, our Chairman and CEO Jamie Dimon compared AI to the steam engine, electricity, and the personal computer. If we look at those episodes of technological innovation, productivity gains don’t appear overnight.
It took more than 60 years for the steam engine to deliver any observable productivity benefit to the entire economy
With each subsequent technological innovation, productivity gains occurred more quickly.
If the trend holds (and we think it will), by the late 2020s, U.S. economic data could show evidence of productivity gains thanks to AI. Think of it this way: It took 15 years for the personal computer to boost the productivity of the economy. AI could do it in seven.
The time between innovation and productivity growth has been shortening
Chart depicts years from innovation to productivity growth for innovations such as the steam engine, electricity, PCs/Internet, and AI
Is AI the 21st-century equivalent of the steam engine or electricity, each of which radically changed the economy? Probably not. But we believe AI has the potential to be as transformative as the web and the personal computer, with the potential to deliver even more economic value over the next 20 years.
The Internet provided a powerful boost to productivity, but AI will likely outpace it
As for the percentage increase in labor productivity relative to the baseline without technological advances, the bar chart depicts the percentage increase in macro productivity relative to the baseline without technological advances
Quantifying job displacement
To quantify the potential impact of AI, we modify an International Monetary Fund (IMF) framework. We conclude that the impact of AI could be much larger than the productivity assumptions built into projections by government agencies such as the Congressional Budget Office.
The IMF identifies which jobs could potentially be displaced by AI. We assume that half of vulnerable jobs in the United States will be automated over the next 20 years. The cumulative productivity gain would be about 17.5% or $7 trillion more than the Congressional Budget Office’s current projection for GDP.
What if half of the vulnerable jobs in the United States are automated in 20 years?
The Congressional Budget Office (CBO) calculated this using the projection of real GDP divided by the number of hours worked. The AI-induced upside potential is estimated by assuming that 15% of jobs will be replaced by AI in the next 20 years, which is half of what the IMF suggested in the research.
It’s important to remember that technological innovation tends to boost economic productivity
But it generally does so only when it changes the amount of labor and capital needed to create a given service or product in the economy. So, for example, Uber didn’t change the labor and capital inputs (it’s still one driver, one car). But driverless cars, if they eventually appear on our roads, could.
What jobs might be most at risk of being replaced by AI? Not surprisingly, white-collar professional service jobs, such as budget analysis and technical writing, appear more vulnerable than child care work or pipe-laying.7
We also expect to see an education gap
A Pew Research Center study finds that workers with bachelor’s degrees or higher are more than twice as likely to be in jobs exposed to AI than those with only high school diplomas. The chart below is illustrative:
Across the global economy, AI will likely disrupt some economies more than others
The IMF has found that workers in advanced economies are more vulnerable to AI displacement than those in emerging market economies. For example, the IMF estimates that 30% of US jobs could be displaced by AI, versus less than 13% in India.
All estimates and projections, including our own, should be taken with a grain of salt. No one can accurately predict the economic trajectory of AI, and estimates of its economic impact vary widely. One specific uncertainty relates to the cost of implementing AI technologies in the workplace.
We are already seeing the infrastructure costs related to building AI computing platforms skyrocket.9 Just because a particular job can be automated using AI technologies doesn’t mean it will be if it is not profitable.
Among optimistic forecasters, Goldman Sachs projects a 15% increase in GDP thanks to AI over the next 10 years. Our estimates are a bit more moderate: we see an 8% to 9% increase in GDP over the next decade. MIT economics professor Daron Acemoğlu takes a much more circumspect view of the potential macroeconomic impact of AI, projecting only a 1% to 1.5% increase in GDP over the same period.10
Economists make varying estimates of how AI will impact growth over the next decade
The chart outlines the potential impact of AI on growth over the next decade and the cumulative percentage increase over baseline real GDP by 2034.
We should also remember that economies evolve in ways that can be best understood in hindsight. According to a study by economists at MIT, more than 60% of current job occupations in the United States did not even exist in 1940.
New technologies explain much of that change. Through each successive technological transition, aggregate demand increased and the economy created jobs that did not exist before.
History tells us that technology continually crea
