Sam Altman was fired from openAI, then hired by Microsoft, and then returned to openAI to continue being CEO. Commercial agreements aside, the latest artificial intelligence tools, including language models like ChatGPT-4, have become the focus of attention around the world. It’s clear that artificial intelligence may be about to shake up our understanding of knowledge work, as well as the broader economy. Just as the car kicked away the coachman, how to apply AI will become a top priority. How should we hold the steering wheel of the AI economy?
Artificial intelligence (AI) is an unprecedented and great technological revolution. Various existing forecasts about AI on economic growth seem to underestimate its impact, and the impact of AI on the economy is by no means measurable by GDP. The impact of AI on the economic cycle is the Kondratiev Wave: Unlike agriculture, the economic development of information-based wealth such as the digital economy, metaverse, and AI will be the main force driving the global economy in the next 10 to 50 years, I think. In fact, agriculture is already highly automated in today’s advanced economies. Various machines are used for irrigation, sowing, fertilizing, and harvesting. We are in an era of high productivity, but there are no longer as many farmers as there were a few decades ago. To some extent, machines have incredibly killed jobs, but the high productivity of modern agriculture has made it easier for everyone to get food. More labor can be invested in other fields or new positions.
Then, how can AI transform the economy? This is about the extent to which it can increase labor productivity in the economy. It is worth noting that the impact of AI on the economy is not necessarily reflected in the growth of traditional economic indicators, such as GDP. More may be reflected in the substitution and change of original production methods – most of these effects may not bring about economic growth but unemployment. Judging from the changes in the production function brought about by AI, we must not ignore the impact of AI on employment. AI has not yet begun to replace employment on a large scale. If AI is good enough to replace humans, at least in the short term, this is a very bad situation for the human workforce. Back to basic demand and supply: If AI replaces traditional production methods, it will improve production efficiency on a large scale. But total demand does not increase sharply, so it may bring about a sharp drop in the prices of related products due to improved efficiency and a substantial increase in supply. Therefore, even if AI comprehensively rewrites all walks of life and causes changes, it will be difficult for the total economic volume to grow, because economic growth is still constrained by the growth of aggregate demand. Similarly, AI technology can improve industrial efficiency, but industrial output will also be limited by natural resources and people’s total demand for industrial products.
There is a more optimistic scenario here. If AI does prove to bring about radical technological innovation. For example, if robots can teach all students the knowledge they need, do we still need traditional teachers? Then we can expect explosive growth in the economy: educational resources are greatly developed, labor expenditures are reduced, and some professors need to find another way out. But this doesn’t seem realistic because of the perceptual nature. In other words, we need at least one, or more, real humans to manage the subject, the direction of teaching, and the mental health. This is to say, even if machines take over the vast majority of jobs, as long as at least some tasks need to be completed by humans, there will always be a way out for humans.
The AI economy, like the Internet economy, digital economy, metaverse economy, etc. over the past 20 years, is information wealth. Traditional material wealth is a production-centered era. However, in today’s era of information wealth, R&D no longer serves production, but production serves R&D. For example, Apple in Silicon Valley has hundreds of parts manufacturers and equipment manufacturers around the world. Is Apple serving these equipment manufacturers and parts and equipment manufacturers? There is no doubt that production serves R&D—Apple’s R&D creates value, while production and assemblers simply transform that value. R&D is the core of economic value, and production only realizes these economic values. The same is true for AI productivity. The core is human ideas. How to realize ideas is the focus of economic transformation.