en.Wedoany.com Reported - The rapid growth of the artificial intelligence industry is facing the dual challenges of rising energy consumption and pressure from electricity price increases. Analysts at consulting firm Wood Mackenzie point out that while some companies are questioning the actual returns of AI in improving productivity, the issue of data centers potentially driving up household electricity bills has already sparked political and social concerns.

At the corporate level, Uber Chief Operating Officer Andrew Macdonald recently stated that the company's engineers have not yet clearly found a correlation between AI tools and significant productivity improvements. Such remarks have intensified market discussions on the commercial viability of AI technology, which has attracted substantial investment in recent years.

Political resistance to data center construction in the United States is growing. At least 14 states are discussing measures to restrict or ban new data centers, including traditionally Democratic states like New York and Vermont, as well as Republican states such as Oklahoma and South Dakota.

Although energy is a fundamental element for AI operations, it accounts for a relatively small proportion of operating costs. SpaceX disclosed that Anthropic pays approximately $1.25 billion per month to use the capacity of the Colossus data center in Tennessee. The facility consumes about 300 megawatts of electricity, and even considering higher electricity prices, power costs account for only about 5% of the total bill.

The real issue lies at the political level. Concerns that consumer electricity prices may rise have become a major argument against data center expansion. Although the growth in electricity demand has not yet led to significant price increases nationwide, tensions have emerged in local areas such as the PJM grid, which supplies power to multiple eastern U.S. states. Governors of New Jersey and Pennsylvania have proposed plans requiring new data centers not to pass costs on to household users. Major tech companies and data center operators have pledged to fund the new power generation and infrastructure upgrades needed to meet their energy demands.

Implementing these commitments is challenging. The public utility regulatory system tends to spread all consumer costs, making it complex to attribute specific expenses to particular facilities. When demand growth drives up prices for key components such as transformers or fuel, the impact ultimately ripples through the entire power system.

Environmental concerns also exist. The growth in power generation needed to meet AI expansion could lead to higher pollution emissions and hinder climate goals. Although these issues are not currently prominent politically, experts believe they may regain a central position in future energy and technology debates.

Against this backdrop, major tech companies are beginning to push for accelerated energy innovation initiatives. Microsoft, Amazon, Google, and Meta announced a partnership with the organization Elemental Impact to jointly fund up to 10 startups focused on developing new energy technologies and materials for data centers, with estimated investments ranging from $500,000 to $5 million per project.

These investments are negligible compared to the hundreds of billions of dollars that major tech companies plan to pour into AI, but they reflect the industry's efforts to respond to criticism over energy impacts. The goal is to promote solutions that can provide cheaper, more reliable, and sustainable energy for communities hosting data centers. However, experts warn that new energy technologies will take years to reach a scale sufficient to transform the industry. Solutions such as small modular nuclear reactors or fusion energy are almost impossible to play a key role before 2040.

Therefore, the industry must rely on currently available technologies. Wood Mackenzie analysis shows that grid connection remains the most efficient and reliable solution for data centers. Companies that can operate with self-generated power and reduce reliance on the grid may gain a competitive advantage in the global race for AI development.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com