As we enter the age of the gigawatt-scale data centers, the way we sell them needs to change. People have wised up, and they know that the new mega-cluster down the road isn’t likely to benefit them directly.

The Internet just works. It’s great. But the trade-off is that the days of saying ‘if you want a better Internet, you need this data center’ are over. We’ve built out such a good digital world that people will require more convincing than ever that we will need to build out more and bigger data centers faster, is in their interest.

You want this, you just don’t know it

This week the BBC wrote a piece about data centers. Like many in the mainstream press, it was less than flattering, calling out the growing energy demands and potential water use the new wave of AI campuses will demand.

It’s not complimentary. Like other pieces that are becoming increasingly common, it hits home some uncomfortable truths; data centers have gone from essential infrastructure hiding in plain sight to major infrastructure having worldwide impact.

I saw a few posts about it on LinkedIn from the data center community. One particularly spicy take caught my eye.

Aside from the hashtags calling it lazy journalism and clickbait when it was neither, the post called out the BBC for criticizing data centers while using a data center to deliver its content and services. A media outlet, in 2025, using digital infrastructure as a medium to call out the dangers of digital infrastructure in a time of rampant hypergrowth and resource scarcity! The hypocrites! The BBC should have posted typewritten paper pamphlets (recycled, naturally) and hand-delivered them if it is so concerned about the impact of data centers!

Telling people they need a new AI super-campus because they use YouTube or a banking app or want to access their emails in 2025 is beyond reductive. As is calling out a company for relying on digital services in an age where not being digital-first is commercial suicide – even if they call you out.

There was a time when that argument held a little water. When the world was going cloud-first, saying a new data center was essential to enable the new realm of Internet services held a grain of truth. Most companies understood that, and most people could come around to that message, too. But in the age of AI and gigawatt-scale campuses, calling people or media outlets hypocritical for opposing data centers because they live in an age when using the Internet is a prerequisite to functioning is pure gaslighting.

Water scarcity is a real issue. Some data centers are very efficient on water use – but many are not – and transparency about data center water is still the exception rather than the norm, leaving an information void filled by concerned voices. Energy prices are going up in some places because of the infrastructure investment required to support such large new loads. Most of today’s growing new energy demand comes from data centers – and polluting natural gas is increasingly being turned to as a ‘stop-gap’ before renewables or new nuclear. And the age-old problem that most of these facilities are fugly still remains. And for what?

Is it the BBC asking companies to build it gigawatt-scale data center campuses? No. Even if they were heavy data center user, like every other company out there, they’re probably still trying to figure out if generative AI is truly useful, a novelty, or a solution looking for a problem. It is the BBC’s job, though, to educate its readers on the impact of a new digital revolution.

True, ChatGPT is popular with the general public. But the data center has been so secretive for so long, few people understand the linkage between today’s generative AI and its increased demand for resources. If people on the street truly understood the scale of the impact generative AI is, can, and will have on the world, perhaps they’d write their own emails and click through links on Google. Or maybe not, but at least then they’d be well-informed.

NIMBYs have changed, so too should data centers

The old days of NIMBYism are dead. To get the world onside, people need a compelling reason to want a new data center in their backyard. The data center industry needs a new message to convince people that the incoming wave of AI data centers is worth it. Strangely, ‘it will benefit the bottom line of Nvidia and the people that buy its GPUs’ or ‘think of the efficiencies’ won’t quite cut it.

Last year, I wrote an op-ed warning that as data centers become increasingly part of mainstream public attention, the amount of criticism will increase, and the way the sector reacts to such pushback will need to change. Unsurprisingly, it very quickly got attention in the data center community and ruffled some feathers. I won’t be shocked if this one does too.

But less than a year on, the issues I touched upon have compounded. The number of data center projects being amended or canceled is increasing. The number of people speaking against projects is going up. The number of politicians running on anti-data center platforms is higher than ever. Regulations around data centers are tightening. Even the number of films where opposing data centers is a central part of the plot has reached a new zenith – admittedly from a low base, but it is indicative of the times.

These people aren’t merely pearl-clutching and opposing change out of principle. Data center opponents are organized and passionate, and care about the impact of big projects where they live. And these people are having an effect. From Virginia to Arizona, North Carolina to Pennsylvania, projects are seeing more and more pushback.

The PW Digital Gateway in Virginia – probably the first data center project to make national mainstream news due to the scale of opposition – got the green light from local officials in late December 2023. But that wasn’t the end of the story. The developers might have thought the worst of it was over, but the opposition dug in tooth and nail to fight the project. This month, the project lost its zoning authorizations due to technicalities around public hearing advertisements.

If the locals were on side, such a small detail probably would have never been noticed, and the project would have gone ahead. But the scale of the project made it too big to go unnoticed, and the developers made too little effort to get locals on board. The companies behind the project will now likely have to appeal the ruling or be forced to cancel – much of the current Prince William County board makes opposing data centers a cornerstone of their election campaigns and will be unlikely to pass any new zoning request.

New era, new message

The heady days of the 1990s and early 2000s have long gone. But so too have the days of building cloud data centers that really were underpinning a new digital age. We’re in the age of AI – for better and worse. And the narratives need to change.

The data center industry has long been churlish in how it deals with NIMBYs – an unkind term in itself – for years telling people they want these facilities really if they want access to the modern world and its benefits. At DCD, we’ve heard ad infinitum at events, meetings, and interviews. It’s never really landed that well, and isn’t going to start now.

Trying to convince locals that a new 2GW cluster is key to quicker Internet speeds, making their Netflix work, or storing their cat photos, isn’t just disingenuous; it's white noise to people who have heard it a thousand times, and leaves a vacuum for the opponents to fill the void with reasons to oppose.

Netflix worked yesterday, and it’s working today; their cloud storage is doing just fine, and Internet speeds have never been better. Even their AI chatbots are functioning. Even if they don’t understand the minutiae of digital infrastructure, people can smell that something is off if a company is promising a new gigawatt campus is needed to keep the status quo. Building out more capacity on one site in a few years than has been built within 200 miles over the last 25+ years needs a new and compelling story and reason to happen in the eyes of the people.

The number of permanent jobs these facilities will bring will always be low, but the financial benefits of data centers are a real, tangible boon. A single campus can completely change the fiscal fortunes of a town or county. But people are fickle, and sometimes they need more than just the promise of cash going to government coffers.

In its place, companies should be talking about the potential these facilities can have. AI can and should be life-changing. It could revolutionize healthcare and unlock the secrets of the universe. AI might enable technological and societal change on a scale not seen since the Industrial Revolution. That’s a narrative people can get behind if sold on the potential benefits. We might not have found the killer AI use case yet, but companies need to be selling that potential.

If developers and operators don’t change their messaging, they will lose out. In the short term, the very real impacts around water and energy are a compelling enough reason for people to say no. In the long term, AI fear-mongers warning of job losses and worse will point to AI data centers as the harbingers of doom.

Or maybe I’m wrong. Perhaps just telling people that new mega-scale data centers are essential if you want to keep your AI clanker lovers will be enough to get the masses onside.