Technology has become a deflationary force, and AI has the potential to push this to the extreme. The cost of many services is likely to fall rapidly, but at the same time, this advancement will provide a unique driver for increased demand for goods. The broader societal impacts of AI are profound and deserve their own article, but today we will examine these counter-inflationary and deflationary trends.
AI has the unique ability to reduce costs in service-related industries. Services account for a significant portion of consumer spending. In developed countries such as the United States, services account for about 70% of consumption, including everything from health care and education to banking and insurance. These sectors have historically been costly due to their labor-intensive nature, including the skills and qualifications required for employees. But AI is changing this dynamic.
This trend is occurring across a wide range of services, from simple to complex. For example, the rise of AI-powered chatbots and voice assistants is already reducing the need for human labor in nearly all digital customer service forums. Call centers are important to many industries and can account for 2% to 3% of a company’s operating expenses. In complex businesses such as telecommunications, this number is often even higher. AI systems can now handle routine customer inquiries, complaints, and services at a fraction of the cost of human agents. This leads to significant savings and reduces costs for consumers in a competitive market.
For example, Klarna recently achieved significant cost savings through its AI initiatives, specifically the introduction of an AI assistant powered by OpenAI. The tool handled the workload of 700 full-time employees and handled two-thirds of Klarna’s customer service conversations. While every situation is different, even a fraction of the potential savings listed above can have a significant impact on industry-wide costs.
AI’s ability to reduce service costs can be applied to even the most complex services. Historically, complex professions were expensive because they required compensation for decades of research. A key example is healthcare, which is the largest cost leading to consumer bankruptcies in many countries and one of the most important items for multiple governments around the world. It’s simply too expensive, medical cost inflation isn’t slowing, and the training required will only increase. AI-enabled systems can now quickly analyze medical data, identify patterns, and help doctors make more accurate diagnoses. This reduces diagnostic errors and unnecessary treatments, increases efficiency, and significantly reduces healthcare costs. This trend is also occurring in several other industries, such as the legal industry, which is incorporating tools to reduce service costs.
Although AI technology is incredibly efficient in the above applications, it still requires large amounts of hardware, energy, and materials, which increases the demand for goods. First, AI models require a huge amount of computing power to operate, and training these models requires even more computing power. The Electric Power Research Institute recently increased its estimate of data center power consumption to account for further growth in AI, saying that data centers could consume more than 9% of US electricity. Third-party forecasts like this one have been steadily rising because many people are not used to a source of demand growing at this pace.
Additionally, AI is increasing the demand for certain metals that are essential to building the hardware infrastructure that runs AI systems. For example, copper is essential for creating wires and components for data centers and electrical systems. Countries such as China, which have historically taken a long-term view of commodity trends, have already built copper stockpiles to multi-year highs. Recently, robots have been shown to be able to make AI decisions and even provide services, as Figure and Tesla have shown. As if we were conceptualizing a new automotive industry, all of the same critical materials would be needed at scale. While AI reduces the labor costs of operating and maintaining these systems, it does not eliminate the need for goods to make the systems work.
This dynamic may result in divergent inflation trends. Meanwhile, consumers will see rapid reductions in the cost of services such as customer support, healthcare, education, and financial services thanks to AI-enabled efficiencies. Meanwhile, prices for commodities such as metals, natural gas, uranium, and industrial materials may rise due to AI infrastructure needs and associated electricity demand, which continues to outpace predictions made just a few months ago.
