Business

The carbon cost of our clicks: why data centers matter

Misryoum examines how energy-hungry data centers, AI usage, and “always-on” digital design add hidden carbon—and what circular, lower-impact tech can change.

Last year at SXSW, a simple thought experiment landed hard: what if every time we checked our phone, a puff of smoke appeared. Misryoum asked the same question in business terms—because even when the impact is invisible, the costs still show up in electricity, water demand, and emissions.

Over the past two decades, the digital ecosystem has become society’s invisible infrastructure.. With more than half the world online. our everyday browsing. messaging. streaming. and AI assistance have turned into a constant drain on physical resources.. Misryoum frames the issue as a major economic transition: the internet isn’t intangible. it’s built on real assets—servers. power supply chains. cooling systems. and the land where data centers sit.

The common intuition is that environmental harm comes from obvious products and industries. not from “a few clicks.” But the internet runs on physical infrastructure.. Every email, text, post, video, website visit, and AI response is processed somewhere—then stored, replicated, and transmitted again.. What users call “the cloud” is, in reality, a network of data centers.. That shift matters commercially because it turns digital growth into an energy-and-water planning problem. not just a software or bandwidth story.

Experience-heavy design has an energy price

Misryoum analysis suggests the problem isn’t only where data is stored—it’s also how digital products are built.. Today, users expect immersive experiences: bold visuals, autoplay video, rich motion, and fast, frictionless interactions.. From a design perspective, those features can be legitimate and even beneficial for engagement.. But nearly every client brief that chases the “flashiest” effect tends to increase data transfer and compute load.

When a site is highly animated or media-heavy, it typically requires more data movement per visit.. Multiply that by daily traffic, and the energy demand scales quickly.. Misryoum sees a disconnect: many organizations treat optimization as a performance exercise. measuring speed and conversion. while the environmental dimension is treated as a secondary concern.. Yet file size. image compression choices. video autoplay defaults. and unnecessary animation all influence how much electricity the system uses behind the scenes.

AI usage and data-center expansion raise the baseline

AI has pushed digital consumption into a higher category of demand.. Misryoum points out that a generative AI query isn’t just a slightly longer search; it often involves substantially more computation and energy use to generate responses.. As AI becomes embedded into everyday workflows, the “baseline” electricity demand associated with online activity rises.

Data centers already consume significant energy, and cooling adds another layer of pressure.. Servers generate heat, and keeping them within safe operating temperatures can require large quantities of water.. Misryoum also flags a location challenge: many data centers are built where large tracts of developable land exist. which can include water-stressed regions.. That creates a direct tension between corporate growth plans and local resource realities.

Meanwhile, the investment cycle is moving fast.. High-profile plans to expand AI-focused data-center capacity signal that this is not a short-term adjustment—it’s infrastructure buildout intended to support the next wave of technology.. Misryoum’s core takeaway is that these decisions determine resource use for years, not months.. The grid capacity, cooling approach, and hardware choices made during construction lock in consumption patterns well into the future.

Efficiency alone may not stop the climb

Even if technology becomes more efficient, emissions and resource use don’t automatically fall.. Misryoum highlights the rebound effect: when digital tools are faster. cheaper. or more widely available. usage can expand to fill the space created by efficiency gains.. That can erase environmental savings if demand keeps growing without guardrails.

This is where business strategy and climate risk intersect.. Data-center expansion is not only an engineering project—it’s a financial and operational bet on energy availability. grid resilience. water sourcing. and regulatory constraints.. If firms treat sustainability as a branding layer rather than a design and capacity constraint. they risk mispricing both costs and externalities.. For the market. that can translate into higher operating exposure over time—through energy costs. cooling limitations. permitting delays. and stricter reporting requirements.

Misryoum also sees a human implication that extends beyond corporate balance sheets.. As digital services become more immersive, more always-on, and more AI-assisted, everyday habits can unintentionally multiply demand.. The “hours online” problem becomes a “compute per task” problem—one that customers. employees. and consumers can influence through usage choices. organizational procurement. and product defaults.

A digital redesign: circularity, deletion, and responsibility

Because technology isn’t going away, Misryoum argues the solution has to shift from only operational efficiency to broader responsibility.. Circularity—often discussed for physical products—can apply to digital systems too.. For designers and engineering teams. that can mean building modular components that are reusable. modernizing legacy architecture instead of endlessly adding on. and planning for content archiving and deletion.

Deletion is easy to overlook. but it’s a practical lever: unnecessary retention of data and content keeps storage and compute requirements alive longer than needed.. Misryoum also notes that sustainability must go beyond carbon accounting to include water use, critical materials, and e-waste.. Extending hardware lifespans. enabling repair where possible. improving recycling. and increasing water reuse can reduce the long-term footprint of the infrastructure that powers “online life.”

Responsibility doesn’t belong to designers alone.. Misryoum sees organizations needing dedicated budgets for sustainable design practices. training teams to measure and reduce environmental impact. and reporting emissions with transparency.. Consumers, too, influence outcomes by extending device lifespans and being more selective about higher-impact digital services and AI usage.

The carbon cost of our clicks can be measured in electricity drawn from strained grids. in the potable or constrained water used to cool servers. and in emissions that intensify climate change.. Misryoum’s bottom line is that the physical footprint of the digital economy may be hidden from daily life—but it is measurable. cumulative. and shaped by decisions being made right now.

Tariff refund portal stumbles: glitches delay claims for US importers

US Data Centers Surge Past Offices—What It Means for Jobs

6 Payment Systems For Small Businesses in 2025: What to Choose