OpenAI and the Economic Impact of AI: An Unseen Potential

The Enthusiasm Surrounding AI and Its Economic Reality

Artificial intelligence (AI) is generating increasing interest, captivating both investors and financial institutions. However, the actual contribution of AI to the American economy last year may surprise those expecting spectacular results.

Despite the ubiquitous rhetoric praising the merits of AI, the American economy does not appear to have benefited from growth as explosive as some tech companies suggest. Dario Perkins, a macroeconomic expert at TS Lombard, expressed his reservations in an interview with the Financial Times, stating that there is no tangible evidence showing that AI has significantly boosted productivity or influenced employment in the United States.

Massive Investments in Automation

Companies have heavily invested in automation, with spending expected to reach nearly $410 billion by 2025. The goal is clear: to accelerate work processes, increase production, and reduce reliance on human labor.

This promise of efficiency and cost reduction is enticing. In the long term, these technologies are expected to improve companies' profit margins. Yet, the banking sector remains cautious. Goldman Sachs, after a period of observation, has taken a firmer stance, asserting that AI has not had a notable effect on American economic growth until 2025.

Brian Peters, a former regulator at the Federal Reserve Bank of New York, shares this skepticism. While he acknowledges the impressive capabilities of AI, he believes that the short-term economic spillovers are difficult to quantify, despite substantial investments.

The Productivity Paradox

Even in academic circles, doubts persist. Researchers from the National Bureau of Economic Research refer to a "productivity paradox." The perceived benefits of AI seem to exceed those that are actually measured, likely due to a lag between the investments made and the expected financial returns.

Reasons for the Gap Between Investments and Results

Several factors explain the gap between massive AI spending and tangible economic results. First, geography plays a crucial role. For instance, when American companies purchase semiconductors manufactured in Taiwan, the direct economic impact primarily benefits that country.

Second, the increased individual productivity due to AI does not necessarily translate into a transformation of global supply chains. Thus, gains often remain confined within companies, without widely diffusing throughout the economy.

By 2025, even the most cautious analysts believed that AI was at least partially supporting GDP growth. Today, that certainty is being questioned. However, the market has not yet fully embraced this pessimistic view, with investment forecasts reaching $660 billion by 2026, according to the Financial Times.

AI: A Revolution in the Making?

It is important to nuance these observations. Aaron "Ronnie" Chatterji, chief economist at OpenAI, compares AI to major technological innovations of the past, such as electricity or the Internet. According to him, the impacts of AI are slowly diffusing and are not immediately visible in traditional economic indicators like productivity or GDP growth.

He explains that to measure real gains, companies must first reorganize their work methods, rethink their processes, and deploy AI at scale. Only after these steps will measurable effects appear in official data, similar to what happened with computing or the web, whose macroeconomic impact was only visible years later.

A study by McKinsey supports this perspective. It estimates that the potential economic impact of AI is immense on a global scale. While some analyses focus on the lack of immediate effect on GDP, this research shows that AI could add trillions of dollars to the global economy by 2030.

Simulations predict a cumulative increase in global GDP of about 16% by 2030, an effect comparable to that of historical technological revolutions.