The rapid expansion of datacenters across the United States, driven by significant investments in artificial intelligence (AI), has become a contentious issue. Nearly 3,000 new datacenters are currently under construction or planned, adding to over 4,000 existing facilities. Major technology companies such as Meta, Amazon, and Google are investing billions in these developments. Virginia leads the nation with 663 operational datacenters and an additional 595 planned or under construction. Texas follows with 405 existing centers and 442 more in development.

Epoch AI, a nonprofit research institute, is mapping this growth using open-source intelligence by analyzing satellite imagery, building permits, and local legal documents to create an interactive map detailing costs, power output, and ownership of these facilities. Their analysis highlights specific projects like Meta's "Prometheus" datacenter in New Albany, Ohio, which has reported construction costs of $18 billion and requires 691 megawatts of power.

The financial implications of this expansion are substantial; an estimated $560 billion in AI-related venture investments has been distributed across all states since 2019. Datacenters are projected to generate nearly $27 billion in tax revenue nationwide over the next decade. However, there is growing concern regarding their environmental impact and energy consumption. Critics argue that while these projects promise economic benefits such as job creation and increased tax revenues—accounting for approximately 92% of GDP growth in early 2025—they may also lead to significant local issues including energy use problems and quality of life concerns for nearby residents.

Some lawmakers have proposed moratoriums on new datacenter projects amid fears that AI could lead to job losses across various sectors. Senator Bernie Sanders has highlighted potential job displacement affecting up to 100 million workers over the next decade due to automation and AI adoption.

Epoch AI estimates that their current dataset represents only about 15% of global AI compute capacity delivered by chip manufacturers as of November 2025. They continue refining their methodology for identifying these facilities while acknowledging challenges related to incomplete information stemming from variable state laws regarding construction disclosures.

As debates continue regarding the future of datacenters—balancing economic growth against environmental considerations—the implications for employment and community impacts remain critical points of discussion among stakeholders navigating technological advancement's effects on society.

Original Sources: 1, 2, 3, 4, 5, 6, 7, 8 (meta) (ohio)

The article about Epoch AI's efforts to map datacenters across the United States provides limited actionable information for a typical reader. While it discusses the use of open-source intelligence and satellite imagery to track datacenter expansion, it does not offer clear steps or instructions that an individual could implement in their own life. There are no resources or tools mentioned that a reader can practically use, making it difficult for someone seeking immediate guidance.

In terms of educational depth, the article does provide some insight into how researchers estimate compute capacity and energy usage based on physical characteristics like cooling systems. However, it lacks detailed explanations of why these factors matter or how they relate to broader trends in technology and energy consumption. The statistics provided, such as construction costs and power requirements for specific datacenters, are mentioned but not sufficiently contextualized to enhance understanding.

The personal relevance of this information appears limited. While the growth of datacenters may have implications for local communities regarding energy consumption and infrastructure development, these effects are not directly tied to an individual's daily life or responsibilities. The article does not address how this topic might impact a person's safety, finances, or health in any meaningful way.

Regarding public service function, the article primarily recounts research findings without offering warnings or guidance that would help readers act responsibly concerning datacenter expansion and its environmental impact. It lacks context that could inform public discourse on technology infrastructure.

There is no practical advice provided within the article; thus, ordinary readers cannot realistically follow any steps suggested by its content. Without actionable insights or tips on navigating issues related to datacenter growth—such as energy consumption concerns—readers are left without useful guidance.

Long-term impact is also minimal since the information presented focuses more on current projects rather than providing strategies for individuals to plan ahead regarding technological developments in their communities.

Emotionally and psychologically, while the article presents factual data about a growing industry, it does not evoke fear or anxiety nor does it offer constructive thinking pathways. However, its lack of engagement with potential community impacts may leave readers feeling disconnected from the subject matter.

Finally, there is no clickbait language present; however, the overall tone remains somewhat dry and academic without compelling narratives that could engage a broader audience effectively.

To add value where this article falls short: individuals interested in understanding more about datacenter impacts can start by researching local energy policies related to large infrastructures in their areas. They can also engage with community forums discussing technology's role in local economies and environmental sustainability efforts. Keeping informed through reputable news sources about advancements in AI technologies will help them understand potential future implications better. Additionally, exploring ways to reduce personal energy consumption can contribute positively amidst growing concerns over large-scale energy usage by tech companies.

The text uses the phrase "energy-intensive buildings" to describe datacenters. This wording can evoke a negative feeling about these facilities, suggesting they are harmful or wasteful without providing context for their necessity in modern technology. By framing it this way, the text may lead readers to view datacenters as primarily detrimental rather than essential for advancements like artificial intelligence.

The researchers at Epoch AI are described as aiming to "provide transparency regarding the impact of these energy-intensive buildings on local communities." This statement suggests that there is a lack of transparency in the industry, which could imply wrongdoing or negligence by companies operating datacenters. The choice of words here may create an impression that companies are hiding information from the public, even though it does not provide evidence for such claims.

When mentioning Meta's "Prometheus" datacenter and its $18 billion construction cost, the text highlights this figure without discussing how such investments might benefit local economies or technological progress. By focusing solely on costs and power requirements, it could mislead readers into thinking that high expenses equate to negative outcomes without acknowledging potential positive impacts.

The phrase "notable projects" implies a sense of importance and prestige associated with large datacenters like Meta's facility. However, this could also suggest that smaller facilities are less significant or worthy of attention. This language choice may downplay contributions from smaller operations in the overall landscape of AI computing capacity.

Epoch AI states their dataset represents only about 15% of global AI compute delivered by chip manufacturers as of November 2025. This claim lacks context regarding what this percentage means in terms of overall industry growth or significance. Without additional information, readers might misunderstand the implications of this statistic and its relevance to understanding AI infrastructure development.

The text mentions that researchers use satellite images to identify cooling systems since they indicate power consumption levels. While this methodology is presented as innovative, it does not address potential inaccuracies or limitations inherent in using satellite imagery for such analysis. By omitting these caveats, it may lead readers to overestimate the reliability and comprehensiveness of their findings.

When discussing confidentiality agreements between companies involved in renting computing resources, the text implies that important operational details remain undisclosed due to corporate secrecy. This framing can create suspicion towards these companies without presenting any evidence or examples illustrating why such confidentiality would be problematic. It subtly encourages distrust among readers toward Big Tech firms based on speculation rather than facts.

The phrase "actively expanding their search for larger datacenters worldwide" suggests urgency and ambition but does not clarify what criteria define "larger" facilities or why they are prioritized over smaller ones. This vagueness can mislead readers into believing there is a clear rationale behind their research focus when there may be complexities involved that are not fully explained in the text.

In describing Epoch AI's mission as one focused on shedding light on Big Tech’s infrastructure growth, there is an implication that such growth is inherently negative or requires scrutiny. The use of “Big Tech” carries connotations associated with monopolistic behavior and exploitation while failing to acknowledge any positive contributions these companies might make through innovation and job creation within communities affected by datacenter expansion.

The text expresses a range of emotions that contribute to its overall message about the expansion of datacenters and their implications. One prominent emotion is concern, which arises from the mention of the significant construction costs and power requirements associated with large datacenters, such as Meta's "Prometheus" facility in Ohio. The figure of $18 billion in construction costs and the need for 691 megawatts of power evoke a sense of urgency regarding the environmental impact and energy consumption tied to these facilities. This concern serves to alert readers to the potential consequences for local communities, emphasizing that while technological advancements are beneficial, they come at a considerable cost.

Another emotion present is pride, particularly in relation to Epoch AI's efforts. The researchers' work in utilizing open-source intelligence to create an interactive map showcases their dedication and innovation in addressing a complex issue. Phrases like “actively expanding their search” reflect enthusiasm and commitment, suggesting that they are pioneers in this field. This pride builds trust with readers by demonstrating that Epoch AI is taking proactive steps toward transparency in an industry often shrouded in secrecy.

Additionally, there is an underlying tension or frustration expressed through references to incomplete information due to variable state laws and confidentiality agreements between companies. This emotion highlights challenges faced by researchers who seek clarity amid obscurity, fostering empathy from readers who may recognize the difficulties inherent in navigating bureaucratic systems.

These emotions guide readers’ reactions by creating sympathy for local communities affected by datacenter operations while also inspiring admiration for Epoch AI’s mission. The combination of concern over energy consumption and pride in research efforts encourages readers to consider both sides: the technological benefits versus environmental costs.

The writer employs emotional language strategically throughout the text. Terms like “significant,” “contentious,” and “energy-intensive” amplify feelings surrounding datacenter expansion, making it sound more urgent than neutral descriptions would convey. By using phrases such as “shed light on” or “provide transparency,” there is an implication that previous practices lacked openness; this contrast heightens emotional stakes by suggesting that knowledge can lead to better decision-making.

Furthermore, repetition of ideas—such as emphasizing both large-scale projects like Meta’s facility alongside smaller ones—reinforces urgency around understanding AI compute capacity comprehensively. By comparing different scales of datacenters while acknowledging gaps in data collection methods due to state laws or company confidentiality, the writer effectively illustrates how complex this issue is.

In conclusion, through careful word choice and emotional framing, the text not only informs but also persuades readers about the importance of transparency regarding datacenter growth and its implications on society at large. It encourages reflection on how technological advancements can coexist with community welfare concerns while fostering trust through acknowledgment of challenges faced by researchers striving for clarity amidst uncertainty.