The macroeconomics of Artificial Intelligence …..and the trends I am seeing

Exploring ‘What-If’ scenarios on how we shape AI and its impact at a macro level specifically on -->

?? Productivity growth

?? Income inequality

?? Industrial concentration

I am a proponent of technology that can bring more efficiency in my day-to-day work, help me grow as a professional and add to my learning/productivity. What I to see is if the benefit of this can be done/seen at a macro-level. The collective decisions we make today will determine how AI affects productivity growth, income inequality, and industrial concentration.?

Economists' forecasts are often inaccurate, and Silicon Valley's enthusiasm for emerging tech is prone to fluctuation. While skepticism about AI's economic impact is warranted, its potential—evidenced by recent significant advances—requires serious consideration. Here I focus on AI's impact on productivity, the job market, and industry concentration. The future of AI is not fixed and will be shaped by current technological choices and policies.

Productivity Growth

The U.S. and many advanced economies have struggled with low productivity growth for the last half-century, with a slight uptick around the turn of the millennium (late 1990s-early 2000s). Productivity is crucial as it influences national wealth and citizen living standards. Enhancing productivity can significantly ease issues like deficits, poverty, health care, and environmental concerns, and is possibly the world’s most pressing economic issue.?

Low Productivity Future

AI's impact on productivity may be minimal due to slow business adoption, potentially leading to narrow labor-saving uses rather than transformative ones. This could relegate displaced workers to less productive roles, dampening overall economic productivity. Historically, technological benefits have often been delayed, exemplified by Solow's 1987 paradox, where technological prevalence didn't match productivity stats—a trend that may intensify with AI. Organizations might also fail to adapt structurally to maximize AI's benefits.

Further, legal and regulatory challenges analogous to those faced by self-driving cars could hinder AI's development. Intellectual property laws might restrict AI training on extensive datasets, leading to potential "patent thickets," and deterring innovation. Simultaneously, stringent regulations or outright bans by national bodies or other entities could slow AI progress. Additionally, there may be protectionist efforts by early AI developers to maintain their competitive advantage.

High Productivity Future

In a more optimistic scenario, AI could lead to a surge in productivity growth by taking over a large portion of tasks across many jobs. AI might fulfill its potential as a pivotal innovation, augmenting human work and freeing up workers to focus on creative and novel tasks by utilizing and learning from vast amounts of digitized information. This could transform the workforce, making it more akin to a community of researchers and innovators, and thus sustain a higher economic growth rate.

Additionally, integrating AI with robotics could make a larger segment of the economy ripe for AI-driven improvements. This could not only enhance existing processes but also unlock entirely new possibilities in areas such as medicine, through AI-driven advances in understanding human biology and pharmaceuticals. AI could even catalyze its own evolution and that of scientific progress in general, entering a cycle of self-improvement that was previously the realm of science fiction.

?

Income Inequality

The?increase in income inequality between individual workers over the past 40 years is a major concern. A large body of empirical research in labor economics suggests that computers and other forms of information technology may have contributed to income inequality by automating away routine middle-income jobs, which has polarized the labor force into high-income and low-income workers.

Higher-inequality future

AI may substitute for human labor in many areas, reducing wages and expanding into traditionally creative domains such as writing, art, and customer interactions. This threatens more jobs and may transform entire industries, as evidenced by recent entertainment industry strikes. In this scenario, not mass unemployment, but greater inequality arises as many are pushed into low-paid service roles that machines can't economically replace, leading to a labor market split between a highly skilled elite and a large, low-paid service workforce, exacerbating income disparity.

Lower-inequality future

In an alternative scenario, AI reduces income inequality by enhancing the abilities of less skilled workers. AI tools, like programming assistants, level the playing field, bringing less experienced coders closer to the proficiency of experts. A study at a call center showed that AI assistance most improved the productivity of the least skilled workers, suggesting that if these gains are shared, income distribution could equalize. Additionally, by taking over routine tasks, AI allows workers to focus on more creative and satisfying work, which could improve job satisfaction, reduce turnover, and increase the quality of service, as evidenced by the same study.

Industrial Concentration

Since the early 1980s, industrial concentration—which measures the collective market share of the largest firms in a sector—has risen dramatically?in the United States and many other advanced economies. These large superstar firms are often much more capital-intensive and technologically sophisticated than their smaller counterparts.

Higher-concentration future

AI enables these firms to become more productive, profitable, and larger than their competitors. AI models become ever more expensive to develop, in terms of raw computational power—a massive up-front cost that only the largest firms can afford—in addition to requiring training on massive datasets, which very large firms already have from their many customers and small firms do not. Moreover, after an AI model is trained and created, it can be expensive to operate. For example, the GPT-4 model cost more than $100 million to train during its initial development and requires about $700,000 a day to run. The typical cost of developing a large AI model may soon be in the billions of dollars. Executives at the leading AI firms predict that the scaling laws that show a strong relationship between increases in training costs and improved performance will hold for the foreseeable future, giving an advantage to the companies with access to the biggest budgets and the biggest datasets.

It may be, then, that only the largest firms and their business partners develop proprietary AI—as firms such as Alphabet, Microsoft, and OpenAI have already done, and smaller firms have not. The large firms then get larger.

More subtly, but perhaps more important, even in a world in which proprietary AI does not require a large fixed cost that only the largest firms can afford, AI might still disproportionately benefit the largest firms, by helping them better internally coordinate their complex business operations—of a kind that smaller and simpler firms do not have. The “visible hand” of top management managing resources inside the largest firms, now backed by AI, allows the firm to become even more efficient, challenging the advantages of small firms’ local knowledge in a decentralized market.

?

Lower-concentration future

In this scenario open-source AI models become widely available. A combination of for-profit companies, nonprofits, academics, and individual coders create a vibrant open-source AI ecosystem that enables broad access to developed AI models. This gives small businesses access to industry-leading production technologies they could never have had before.

It may also be that AI encourages the kind of broad, decentralized innovation that better flourishes across many small firms than within one large firm. The boundaries of the firm are the outcome of a series of trade-offs; a world in which more AI-backed innovators need the residual control rights to their work might be one in which more innovators decide they would rather be owners of small firms than be employees of large ones.??

?

Outcome

?? The path that leads to a worse future is the one of least resistance and results in low productivity growth, higher income inequality, and higher industrial concentration.??

?? Society needs innovations in economic and policy understanding that match the scale and scope of the breakthroughs in AI itself.??

?

Policy makers need to understand that AI can develop in different directions, and this can help frame their discourse accordingly. Things they need to consider:

?? How can policies encourage the types of AI that complement human labor instead of imitating and replacing it?

?? What choices will encourage the development of AI that firms of all sizes can access, instead of just the largest ones?

?? What kind of open-source ecosystem might that require, and how do policymakers support it?

?? How should AI labs approach model development, and how should firms approach AI implementation?

?? How does society get an AI that unleashes radical innovation, instead of marginal tweaks to existing goods, services, and systems?

There's a significant gap between the amount of research going into advancing AI technology and the research dedicated to understanding its impact on economy and society. This gap wasn't as critical when AI's macroeconomic effects were smaller.

However, with AI's impact now potentially reaching into the trillions of dollars, it's crucial to invest more in studying AI's economic aspects. We need economic and policy insights that are as groundbreaking as AI's technological advancements. By shifting research focus and crafting smart policies, we can aim for continuous and inclusive economic growth.

Some things I am observing in the market that worry me about the negative ‘what-if’ scenarios above:

2023-2024 Tech lay-offs

The widening income gap in the US

Reference: IMF report by Finance and Development

Opinions expressed here are not policies.