This article originally appeared on Inc Magazine.

Generative AI will have multifaceted effects on work and the economy, including the potential to reshape traditional roles, create new job opportunities, and transform productivity.

The economic and ethical implications of generative artificial intelligence include its effect on the labor market. It’s not just changing how people work, but also reshaping the nature of work itself. Traditional roles are evolving, some jobs face disruption, and new roles are emerging in their place.

This impact is already in progress, but a lot is still unknown. What applications of AI will have the most commercial success? What kinds of jobs will be created out of new opportunities? What kinds of work can generative AI fully replace versus augment? Despite these uncertainties, here is a framework to assess AI’s impact.

First, a look at what generative AI does for businesses. There’s a lot of conversation about the automation of tasks that’s possible with AI, as well as the text-generation abilities of tools like ChatGPT. But even more fundamentally, generative AI can analyze enormous sets of data and learn how to recognize patterns much like the human brain. 

That data digestion and pattern recognition then inform the AI decision-making model. Unlike humans, AI models can look at huge amounts of data in a short period, massively accelerating the process of deciding what actions to take in a given situation.If AI is thought of as “prediction machines” that enable organizations to make cheaper, more abundant, better-automated decisions, then the best generative models can reshape the economy because of their widespread relevance. If AI automates decisions in the economy, this increases productivity, which has significant effects on labor and investment. Since information and knowledge work dominate the U.S. economy, the potential for AI systems to boost overall productivity is vast. 

The measurement of productivity

Productivity can be divided into two components: total factor productivity, or TFP, a measure of the impact of technology, and the contribution of the labor composition and capital intensity. Recently, the U.S. has experienced slow TFP, or tech-related growth, which has made it harder to fight inflation, eroded wages, and worsened budget deficits. 

However, the 1990s saw a surge in productivity growth driven primarily by an investment in computers that drove business transformations. Despite the stock market bubble and a reallocation of labor, workers were better off. The federal budget was also balanced from 1998 to 2001. Digital technology can drive economic growth. 

As the capabilities of generative AI systems grow, allowing them to perform many tasks that used to be reserved for cognitive workers, the broad applications of the system will impact a large portion of the U.S. workforce in some form. According to recent research, LLMs could affect 80 percent of the U.S. workforce. A recent report by Goldman Sachs suggests that generative AI could raise global GDP by 7 percent, a truly significant effect for any single technology.The good news is that these gains are not hypothetical–they are significant in the real world. The ability of AI systems to create value by capturing knowledge and conveying information on a wide range of tasks was previously only learned on the job. This is a key indicator of AI’s effect on the labor market. 

Channels of AI’s impact

AI enhances output efficiency. More efficient workers mean increased output. AI also accelerates innovation as cognitive workers not only produce output but also invent new things and engage in discoveries that boost future productivity. As these workers become more efficient, they contribute more significantly to technological progress, compounding productivity growth over time.

However, for these productivity gains to materialize, AI advances must disseminate across the economy, a process that takes time and adaptation. Upskilling and training staff to effectively use these new technologies is crucial for realizing productivity gains.

The productivity effects of generative AI go hand in hand with disruption in the job market. However, much of the current discussion is around displacement, because it’s much easier to conceive of straightforward machine-human replacement in completing tasks, reducing the demand for labor. AI’s potential for augmentation, capital deepening, and creating new tasks and industries indicates an overall positive impact on labor demand. For instance, an improved AI recommendation system like Netflix’s could increase company returns and, consequently, the demand for workers. Investment in AI systems enhances worker productivity, increasing labor demand. Furthermore, as AI creates new tasks and industries, this further boosts labor demand.

Contrary to fears of a job apocalypse, this model suggests that AI systems enhancing labor productivity could also bolster labor demand. However, it’s essential to recognize that the net effect of new technology on labor demand could be negative if the AI systems are “mediocre” in that they are only productive enough to displace workers, but not productive enough to increase labor demand through the other channels.

While AI’s impact on the labor market is complex and multifaceted, it holds the potential for significant economic growth and job creation, provided the challenges of displacement and the need for skill adaptation are adequately addressed.