Artificial intelligence has the potential to improve productivity, or the amount of output relative to inputs such as labor, energy, and materials. Prices are expected to fall as AI maximizes productivity. It has a general effect on the price level, lowering the economy-wide inflation rate and potentially causing deflation. However, the actual route is not certain.
The key issue with prices overall is the increase in supply relative to the increase in demand. Milton Friedman's famous statement refers to the money supply, but it also applies to the theory that stimulus comes from fiscal policy. It's more about the amount of money than the amount of output. ”
Therefore, if AI increases production beyond the increase in money supply (or other stimulus), inflation could slow and potentially reach falling prices, or deflation.
When we think of AI, the first thing that comes to mind is the supply side. The creators of ChatGPT, his team at OpenAI, conclude: [large language models]Meanwhile, about 19% of employees expect at least 50% of their work to be affected. ” Less labor required per unit of production means lower costs. In a competitive field, all else being equal, lower costs lead to lower prices.
AI does not help reduce costs across the board. In practical fields such as construction, the impact of AI is small or negligible. Legal or regulatory restrictions may slow the adoption of AI in some sectors that could benefit. However, the supply of goods and services should increase significantly.
Aggregate demand is already being stimulated by AI. Part of the recent boom in computer chip manufacturing stems from the need for more processors for AI. Chips are needed to run the training, which requires a huge number of chips running 24 hours a day over several months. Then, getting answers from AI models requires large data centers full of chips. More generation and transmission lines will also be needed to power data centers.
Companies may find that implementing AI requires different capital than they currently have. A fast food ordering system based on voice recognition using AI may cost less than the old cashier-based system, but the equipment is different. Economic activity will be revitalized by discarding the old and introducing the new.
John Maynard Keynes argued that capital investment tends to generate waves of optimism and pessimism, causing business cycles. More recently, Finn Kydland and Edward Prescott won the Nobel Prize in Economics for their statistical analysis that concluded that business cycles are often driven by waves of changes in productivity.
On the demand side, new products can also cause increased consumer demand. At the moment, these examples seem trivial and unimportant, but that may change as smart minds put new ideas into practice.
At this point in our analysis, AI could significantly stimulate the economy's aggregate supply and modestly stimulate aggregate demand. The net effect appears to be dominated by factors that lower prices.
But remember Friedman's statement that stimulus is considered relative to output growth. If economic policymakers perfectly predict changes in productivity and demand due to AI, they will be able to properly keep inflation at the 2% target. If they are slow to recognize the productivity gains brought about by AI, inflation could fall and deflation could occur. However, if policymakers raise expectations for productivity growth too much, they will overstimulate the economy, leading to higher inflation.
For the past three years, fiscal policy (government spending and taxes) has not been used to stabilize the economy. Congress and the president passed spending bills that resulted in large deficits despite good economic growth. It's the Federal Reserve's job to keep the economy on track. Therefore, the question to consider is whether the Fed can accurately predict the impact of artificial intelligence over a one- to three-year time horizon.
The Federal Reserve Board in Washington, DC employs 400 Ph.D. economist. Each of the 12 regional Fed banks employs a small number of economists and also academics to support their research programs. Even with such a large pool of talent, the task of assessing the impact of AI in a short period of time will likely overwhelm their capabilities. Companies today don't know how to implement AI tools. For precedents, economists study the industrial revolution, the expansion of electricity, and the early days of computers. This means that when you need fine-tuning capabilities, you only have a rough idea of how changes will unfold.
Therefore, we should expect large swings in inflation. My guess is that the Fed will wait to add more stimulus until there is clear evidence of lower prices due to productivity gains, and then jump on stimulus. The Fed will then try to measure productivity gains based on very recent experience and miss the ups and downs that are inevitable with the use and benefits of AI. This is not to blame the Fed. The task is too big for anyone.
Businesses, consumers, and investors should expect a wild ride. Flexibility is key. Big bets can win in the long run, but only if bettors can maintain their positions during setbacks.