A new report from the International Energy Agency (IEA) reveals a significant shift in global economic priorities, with an estimated $580 billion projected to be spent on data centers this year. This figure surpasses the $540 billion allocated for finding new oil supplies by $40 billion, highlighting the growing investment in digital infrastructure. This comparison is particularly relevant amid concerns that the rapid expansion of generative artificial intelligence (AI) could exacerbate climate change and put additional strain on energy resources. These findings were discussed on TechCrunch’s Equity podcast, featuring Kirsten Korosec, Rebecca Bellan, and Anthony Ha.

The proliferation of new data centers is expected to create an immense demand for power, potentially putting further stress on already strained electrical grids. However, a significant potential upside exists: the increasing adoption of solar power and other renewable energy sources to fuel these projects. This trend could open new opportunities for startups focused on innovative renewable energy solutions and sustainable data center design. Major tech companies are committing substantial capital to this expansion, with OpenAI reportedly dedicating $1.4 trillion, Meta $600 billion, and Anthropic announcing a $50 billion plan for data center development.

Kirsten Korosec elaborated on the positive outlook for renewables, citing climate tech reporter Tim De Chant’s research. Many data centers are increasingly turning to renewable energy due to fewer regulatory hurdles and cost efficiencies, making it easier and more economical to install solar panels adjacent to facilities. This pivot is a boon for companies developing innovative renewable solutions or advanced data center technologies aimed at reducing global emissions. As a former energy reporter, Korosec emphasized the sheer scale of investment in data centers compared to traditional energy exploration, underscoring the magnitude of this economic reorientation.

Rebecca Bellan highlighted the geographical distribution of escalating electricity demand, with approximately half expected from the U.S. and the remainder from a mix of China and Europe. A key challenge arises from the practice of locating most new data centers in or near urban areas with populations around one million, creating significant obstacles for grid connection and infrastructure development. Bellan concluded that focusing on renewables is not just an environmentally friendly policy choice but a sound business strategy for these operations.

Addressing the critical issue of grid stress, Kirsten highlighted Redwood Materials’ new business unit, Redwood Energy, as a promising solution. The company repurposes old electric vehicle (EV) batteries not yet ready for recycling, transforming them into microgrids designed specifically to power AI data centers. Such initiatives are crucial for alleviating pressure on electrical grids, especially during peak demand seasons in regions prone to rolling brownouts and blackouts, like Texas. The success of Redwood Energy could catalyze new investment in companies pursuing similar strategies to enhance grid resilience.

Anthony Ha raised questions about the broader implications of large-scale construction, particularly its impact on landscapes even outside major cities. He also questioned the feasibility of these ambitious plans, given the immense capital commitments involved. Ha noted the ongoing scrutiny of OpenAI’s financial model, particularly how projected earnings compare to its stated trillions of dollars in commitments over the next decade. He referenced the controversy surrounding OpenAI’s CFO, who initially mentioned government “backstopping” loans for data center construction, later clarifying that the company sought an expansion of tax credits under the CHIPS Act. This suggests that building out data center infrastructure will require substantial collaboration and financial planning from both private companies and governmental bodies in the coming years.