The AI Risk Nobody’s Talking About (But Should Be)

Al Gore has been sounding the alarm on climate change for decades, but when it comes to artificial intelligence, he’s surprisingly calm about one of the most pressing concerns: data centers’ carbon footprint. In an interview with TechCrunch, Gore pointed out that while AI emissions are indeed a problem, they’re dwarfed by other energy-consuming behemoths – namely, air conditioning.

This might seem like an odd thing to say coming from the climate icon, but it’s actually a crucial observation about the way we discuss and frame our environmental worries. We tend to get caught up in the shiny new things that worry us – AI’s potential for job displacement, its role in perpetuating societal biases, or the catastrophic consequences of losing control over these systems. Meanwhile, more prosaic concerns like energy consumption and resource management get relegated to the background.

Gore’s focus on air conditioning highlights a critical issue: as we rush headlong into an era of unprecedented technological progress, we’re overlooking some very basic facts about our daily lives. Global air conditioning already consumes more electricity than the entire European Union, according to the International Energy Agency. This trend is expected to continue, with ownership rates projected to triple by 2050 in the hottest and fastest-growing parts of the world.

The Unseen Costs of Progress

The environmental costs of AI are real, but they’re often buried beneath the surface. Take, for instance, the methane turbines powering some hyperscalers’ data centers. While these might seem like a convenient solution to our energy needs, Gore is right to worry about locking in decades more of fossil fuel generation. The fact that renewables and batteries have become increasingly cheaper should be a wake-up call: it’s no longer just about meeting short-term demands; we need to think about the long game.

A Shift in Perspective

Gore’s willingness to listen to industry leaders’ warnings about AI – particularly from OpenAI, Anthropic, and their respective CEOs – is telling. He believes them when they say that slowing down capability gains might be necessary, not because it’s a marketing ploy or some form of “apocalyptic” fear-mongering. In fact, he invokes Maya Angelou’s wisdom: “When someone tells you who they are, believe them the first time.” This is a rare moment of humility in an industry where self-interest often trumps genuine concern.

What’s at Stake

Gore seems to be getting at – and what Preston highlights with her focus on decoupling compute from energy intensity – is that AI has the potential to drive real change. Efficiency gains, waste elimination, and grid resilience are all areas where capital investment can create tangible benefits. It’s time for us to stop treating AI as just another source of anxiety and start thinking about its role in addressing our collective problems.

A New Era of Responsibility

We’re at a critical juncture: the world has been “brutally reminded” twice in four years that fossil fuels are a volatile energy source. Gore’s optimism is infectious, but it also highlights the stakes. If we can’t get this right – if we continue to overlook everyday concerns and instead focus solely on the latest AI breakthrough or existential threat – then we risk being trapped in our own making.

The Unspoken Truth

Perhaps the most disquieting aspect of Gore’s comments is that they’re a reminder of just how disconnected our conversations about tech have become from reality. We get caught up in debates over data centers, forgetting about the people and processes that actually make them tick. It’s time for us to acknowledge – and tackle – the AI risks nobody’s talking about but should be.