Lane, Water, Power, Permitting Issues Drive AI Data Center Infrastructure Decisions

It used to be the case that choosing the site for a new data center was based on three things: the availability of sufficient and affordable power, land and access to good optical fiber connections. 

The need for land has not changed. But in the present era of high-performance computing, data center requirements arguably center on access to power, water (needed to cool the dense arrays of processors) and the relative ease or difficulty of gaining needed permission to proceed (permits, regulatory issues. ). 

Goldman Sachs Research estimates that in 2027 a rack of artificial intelligence servers will require 500 kW of power, about 50 times more than a 2022 data center rack, which required between 5 kW and 15 kW of power. 

source: Goldman Sachs

 

source: Goldman Sachs

Goldman Sachs Research further estimates that 40 percent of the increase in power demand from data centers will be met by renewables, including renewable natural gas. The bulk of the remaining 60 percent is expected to be driven by natural gas. 

In 2024, natural gas supplied over 40 percent of the electricity for U.S. data centers. Towards 2030, the natural gas role is expected to increase, with major projects in development, such as Amazon's 754 MW natural gas plant in Mississippi and a 4.5-gigawatt campus in Pennsylvania. 

To be sure, alternatives including small nuclear reactors are under development. But natural gas advantages are many, at least for U.S. data centers. The United States is the world’s largest producer of natural gas and has an extensive pipeline network. It’s an energy source that can be switched on and off easily to match demand and does not fluctuate in supply as do wind and solar energy sources. 

Data center operators are looking at both on-grid and off-the-grid alternatives, including renewable natural gas and local generation. 

.  

source: Orennia

And use of renewable natural gas, in combination with other renewable sources (wind, solar, batteries) improves the economics of using renewable energy. 

source: Orennia

In July 2023 the Environmental Protection Agency noted there were 532 U.S. landfill-based renewable natural gas operations.  

While natural gas has the lead in the United States today, the gains by renewable generation, as well as the expansion of nuclear generation using small modular reactors, are expected to cut into gas consumption.

Renewables are now projected to add 110 TWh to the domestic data center electricity supply between 2024 and 2030, according to the International Energy Agency . 

Land, power and water remain the key physical requirements for new AI-focused data centers. Of these inputs, power and water see heightened importance.

Blue Pill, Red Pill or No Pill?

The character Morpheus in the movie The Matrix offers the character Neo a choice between a blue pill and a red pill. Take the blue pill and you stay in the artificial world you believe exists. Take the red pill and you see reality.

Likewise, “one pill makes you larger, and one pill makes you small,” sang Jefferson Airplane lead singer Grace Slick. 

Both are examples of the "there's a pill for that" therapeutic culture in the United States, a societal preference for quick, convenient, and often pharmaceutical solutions to a wide range of physical and emotional discontents.

At the same time, there is a growing counter-movement that embraces "self-improvement" through lifestyle changes, alternative therapies, and holistic wellness practices. 

To be sure, some people have health conditions unrelated to personal choices. High cholesterol levels are largely genetic for at least one out of 250 persons, one study suggests, although 10 percent of adults have high cholesterol levels. 

The percentage of U.S. adults on GLP-1 drugs such as  Ozempic. Wegovy. Mounjaro and Zepbound, for example, could reach nine percent to 16 percent by about 2030, according to some estimates. 

Some studies indicate that more than half of U.S. adults could be eligible, assuming they face risks from diabetes, obesity, or cardiovascular disease. Without question, consumer demand is fairly high. 

One study suggests 12 percent of U.S. adults have taken GLP-1 drugs. 

It isn't as easy or as quick, but for many, the choice is not blue pill or red pill or any pill, but the lifestyle changes (not especially easy or necessarily always convenient) that obviate the need for taking a pill.

Agentic AI Will Cause Additional Disintermediation in Value Chains

Agentic artificial intelligence, which as software agents acting on behalf of human users, will threaten some participants in existing value chains, in the same way that internet platforms and apps disrupted commerce and content value chains. 

“Disintermediation” is the removal or reduction of intermediaries in any value chain. Disintermediation allows buyers, producers, and consumers to bypass traditional middlemen such as brokers, consultants, customer service agents or logistics coordinators. 

Consider any procurement operation. 

AI agents can gather and compare data across sources in real time, reducing reliance on human experts or intermediaries for information gathering or product curation. That might threaten some parts of Amazon and other e-commerce platforms, for example. 

As was the case with internet retailing, this is going to create new pressures for “price-based” comparisons and some potential diminution of “brand value.” 

AI agents then will negotiate price, delivery, and quality parameters autonomously, replicating the human “buy this” operation, and also circumventing many of the marketing practices that assume a human is persuadable during the buying process. 

As was the case for internet retailing, agentic AI should create more direct producer–consumer ability to transact directly, without distributors.

Process orchestration also should happen, where the AI unifies and handles procurement, contracting and payment operations that previously might have required multiple apps or systems. In a growing number of cases, this will involve the buyer’s agent negotiating with the seller’s agent, without distributors, advisors, consultants or specialists in between them. 

And where internet commerce featured lots of “personalization,” so agentic AI will replace “trusted advisor” or “expert advice supplier” functions and suppliers of those values. “Personalization” and “AI customization” will be analogous outcomes. 

Industry / Value Chain Stage	Traditional Intermediary Role	How Agentic AI Enables Disintermediation	Agentic AI Scenario
Retail and E-commerce	Online marketplaces (Amazon) aggregate sellers and handle logistics	AI shopping agents directly compare sellers, place orders, and track delivery	Consumers’ personal AI negotiates bulk discounts from multiple retailers and arranges delivery without using a central platform
Financial Services	Brokers, financial advisors, loan officers	AI evaluates options, performs due diligence, and executes trades or loans	A consumer’s AI portfolio manager automatically reallocates investments across platforms using live market data
Real Estate	Real estate agents and mortgage brokers	AI agents handle property search, valuation, negotiation, and contract execution	Buyers use AI that identifies undervalued homes, negotiates price, and manages closing paperwork
Supply Chain & Procurement	Procurement agents, sourcing platforms	AI autonomously sources suppliers, evaluates risk, and executes contracts	A manufacturer’s AI identifies suppliers worldwide and directly contracts best-value inputs without human brokers
Healthcare	Primary care gatekeepers, medical schedulers, or insurers as coordination intermediaries	AI triages symptoms, recommends providers, and books care directly	AI health assistant evaluates symptoms, finds available doctors, and schedules an appointment — skipping insurer’s pre-authorization layers
Entertainment / Media Distribution	Streaming platforms, music labels	AI agents match creators directly with audiences and handle rights/licensing smart contracts	Artists’ AIs distribute content directly to audience AIs, who pay micro-royalties automatically
Travel & Hospitality	Travel agents, comparison websites	AI directly plans and books multi-leg trips, comparing prices and reliability	A traveler’s AI negotiates with airlines and hotels’ AIs to assemble the best route and price
Legal & Professional Services	Lawyers, notaries, consultants	AI creates, reviews, and files contracts autonomously	SMEs use AI to draft and file incorporation paperwork directly with government APIs
Education / Training	Universities, training marketplaces	AI tutors create personalized curricula and credentialing directly	Learners use AI tutors that build custom programs, verify mastery, and issue credentials via blockchain
Advertising & Marketing	Agencies, ad brokers	AI agents buy media and tailor campaigns autonomously	A small business’s AI negotiates ad buys with media outlet AIs in real time, eliminating agency fees

So among the “dangers” or challenges for e-tailers are new value compression issues, with the correlating danger of profit margin compression. 

The danger for e-commerce platforms is a loss of gatekeeper power as more peer-to-peer or agent-to-agent interactions develop. 

Brands might also find they face some diminution of “brand value,” just as price comparison sites will shift buyer evaluations in the direction of “lower price.”

What Do We Do When AGI Automates Much Economically-Essential Work?

It is reasonable to suggest that, at the moment, agentic artificial intelligence is not yet ready to displace many full human jobs. Hopes are higher (or more worrisome, depending on one’s point of view) for artificial general intelligence. 

The equally far-reaching implications, though, might happen if artificial general intelligence does acquire such capabilities. For as hard as it might be to imagine a world where nearly all essential work can be done by the “compute,” the economic ramifications would be stunning and unprecedented.

“Before AGI, human skill was the main driver of output, and wages reflected the scarcity of skills needed for bottleneck tasks,” says Pascual Restrepo, author of the paper “We Won’t be Missed: Work and Growth in the AGI World,” published by the National Bureau of Economic Research. 

Consider the potential impact on jobs, wages and sources of value. “In an AGI world, compute takes that central role, and wages are anchored to the computing cost of replicating human skill,” he argues. “While human wages remain positive, and on average exceed those in the pre-AGI world, their value becomes decoupled from GDP, the labor share converges to zero, and most income eventually accrues to compute.” 

There are some caveats. 

AGI assumes we can replicate what people do if we throw in enough compute at the tasks. That does not mean it is practical or efficient to automate everything. 

Depending on the computing costs 𝛼(𝜔), it may be better to leave some tasks to humans and

allocate our finite computational resources elsewhere.

Also, some work requires interacting physically with the world. AGI optimists assume that, when needed, and if economically rational, computer systems can control machines and hardware to accomplish this work. 

Some work requires empathy and social interaction and, it is argued, must be carried out by humans. The “human touch” and “empathy” of a therapist or healthcare provider may be impossible to replicate, creating a premium for work completed by people. 

The issue is whether we can substitute so much compute that the alternative is really between a human and an AI system that “perfectly emulates the best therapists in the world (from a functional point

of view).” 

Assuming we can afford to do so, one might rationally argue there are some, or many, instances where the AI is an acceptable substitute. 

One must also assume that compute capabilities and costs continue to scale over time on something like the Moore’s Law rate. 

All that noted, we might still argue that even if some work can be automated, it might not be. There will of course be a cost for using AGI. And if the costs are significant enough, and the tasks being considered for AI substitution can be handled by humans at equivalent or lower cost, then using AGI will not make sense. 

Hospitality, live performance or entertainment might provide examples. 

Also, AGI compute might be a scarce resource. If so, then normal cost-benefit logic should hold:AGI replaces human labor when it makes economic sense to do so.

A new theory of value might include the idea that human labor is worth what it saves in compute costs, Restrepo suggests. But algorithmic progress, which arguably is less linear than “compute infrastructure,” should also be an issue, as uncertainty introduces volatility. 

The social implications are huge. In an AGI economy, most income accrues to owners of compute. How society manages such a transition, in terms of impact on social inequality, is unclear. 

As Restrepo says, “today, if half of us stopped working, the economy would collapse.” That might not be true in a future where AGI can be economically deployed to displace humans in economy-central roles. 

All of which raises new issues around “abundance” that humans have not generally had to deal with in the past: what do people do when they do not actually ne

AI Equity Volatility Shows 30-Point Swing Between Fear and Greed

“Fear and greed” notoriously are drivers of equity market sentiment and that is clear in the yo-yo behavior surrounding artificial intelligence equities recently.

A positive development such as a new chip announcement, a major partnership like the AWS/OpenAI compute services deal, or strong earnings from an AI leader pushes the market into "extreme greed" territory, driving up prices quickly.

But then reports of high AI capital expenditure, delayed profitability for end-users, or a general sentiment survey warning of a "bubble" causes profit-taking and selling, plunging the market into "fear" sentiment, leading to sharp, temporary pullbacks.

Month	Major Event	Sentiment	Notable Impact
2025-01	DeepSeek Launch	Fear	Sharp drop, infrastructure risk flagged
2025-04	Trump Tariffs Threat	Fear	Market volatility spiked, quick rebound after walkback
2025-09	NVIDIA-OpenAI Chip Deal, Fed Rate Cut	Greed	Strong surge, positive sentiment returned
2025-10	Bubble Talk Surge	Fear	Renewed caution, market exhaustion warnings

The cycle resets because the fundamental belief in AI's future remains generally strong. Investors who sold out of fear often rush back in for fear of missing the next leg up (greed), making the dips short-lived and creating the current high-volatility, upward-trending cycle. 

But skepticism and hope continue to coexist and oscillate. 

Beyond the volatility, we might argue that “high-performance computing capability” has become a strategic commodity.

High-performance compute capacity arguably has become the single most critical, scarce, and expensive strategic resource in the AI industry. 

If so, long-term, multi-billion-dollar compute contracts are now a competitive necessity, resembling procurement models for essential commodities like energy or raw materials. But volatility will persist until some future time when there is much more predictability about AI investments and revenue gains.