As AI technology continues to progress, its effects may forever change the world of economics and economic policymaking. I recently had the pleasure of joining David Beckworth on the Macro Musings podcast to discuss these transformative changes. Here’s a summary of our conversation and my thoughts on the matter. Listen to the podcast here.

My Move from Macro to AI

I’ve been an economist for nearly 30 years, with a background in government, the banking sector, and running my own consultancy. Data has always played a significant role in my work. However, with the advent of the internet, increased computing power, and advancements in machine learning, we’re seeing a shift where all markets are beginning to behave like financial markets, with prices becoming less sticky.

In 2015, I resigned from my banking job to start Markets & Money Advisory, with the goal of doing more academic work and exploring sports economics and analytics. The release of ChatGPT sparked my interest in AI.

Initially, I used it for Python coding and econometric analysis, leading to significant productivity gains. This led me to establish an AI advisory company called PAICE with my partner Christian Heiner Schmidt, where we advise financial and retail companies on using AI and data to make decisions.

The Basics and Implications of Dynamic Pricing

Dynamic pricing, or flexible pricing, reacts to supply and demand changes. While there are costs associated with changing prices (menu costs), technology is reducing these costs. In Scandinavian supermarkets, for instance, electronic price tags allow for frequent price adjustments from headquarters.

This shift resembles the transition in finance from open outcry trading to electronic trading. As the cost of implementing dynamic pricing decreases, we’re moving towards a world where retail prices could change as frequently as gas prices do today. Imagine a future where Jerome Powell’s statements influence the price of milk in Copenhagen instantaneously. This would result in a world where prices are much more flexible, significantly impacting monetary policy.

Using AI for Econometric Analysis

I’ve been exploring AI’s potential in econometric analysis. For instance, I used ChatGPT to model Gary Becker’s Rational Addiction Model for calories and anti-obesity medicine. More recently, I asked ChatGPT to create a New Keynesian DSGE model (see my post on this here). The results were impressive, as it provided Python code and helped troubleshoot errors, demonstrating AI’s potential as a powerful tool for economic analysis.

This technology allows us to perform complex econometric analysis much faster, improving productivity and enabling more sophisticated research. However, the quality of analysis still depends on the economist’s expertise and understanding of economic principles.

The Implications of AI for the Economics Field

AI is essentially advanced statistics, evolving from traditional econometrics. While it enhances our productivity, it doesn’t replace the need for critical thinking and economic insight. Economists must still understand the underlying principles and implications of their analysis.

In the future, the role of economists may shift from government and finance to the broader economy. Just as the quants moved into finance in the ’90s, we might see economists applying their skills in diverse industries like retail, where dynamic pricing and data-driven decision-making become more prevalent.

How Will AI Impact the Federal Reserve and Its Policymaking?

AI’s ability to process and analyze vast amounts of data could transform the Federal Reserve’s operations. As markets and prices become more flexible, the Fed’s traditional tools and models may become less relevant. AI could enable more accurate macroeconomic indicators and better policy decisions.

However, the need for a large staff of economists at the Fed might decrease. As technology reduces the costs of monetary policy mistakes, the Fed’s role could shift towards monitoring and adjusting to real-time data rather than extensive analysis and forecasting.

Deflation as a Response to an AI-Driven Productivity Shock

George Selgin’s idea of allowing mild, gentle deflation in response to rapid productivity gains is intriguing. In a world where AI drives significant productivity improvements, output prices could fall, benefiting consumers and sharing real gains widely. Keeping demand stable while allowing prices to decrease gradually could ensure that everyone benefits from AI’s advancements.

In conclusion, AI has the potential to revolutionize economics and monetary policy, making processes more efficient and data-driven. While there are challenges and adjustments ahead, embracing this technology will enable us to navigate the future with greater insight and agility.

Contact:

Lars Christensen

LC@paice.io

+45 52 50 25 06