Here's what nobody wants to admit: the artificial intelligence boom is quietly becoming the fossil fuel industry's best marketing campaign in years.
The stated problem is straightforward. Data centers powering AI systems consume staggering amounts of electricity. Renewable sources can't scale fast enough to meet demand. So what happens? Tech companies turn to natural gas plants and coal facilities that can flip on immediately. The industry gets a shiny new justification for keeping old infrastructure running. Everybody pretends this is temporary. Nobody believes it.
This matters beyond Silicon Valley because it reveals a fundamental misalignment in how we're incentivizing climate action. We celebrate corporate commitments to net-zero emissions while simultaneously creating economic conditions that reward companies for choosing the fastest, dirtiest power sources available. The math is simple: clean energy projects take years to build. Fossil fuel plants are already there. For a company desperate to deploy the next generation of language models, the choice writes itself.
The carbon removal industry illustrates this problem even more starkly. We now have emerging technologies that can capture CO2 directly from the air. These are expensive, requiring energy inputs that need to be clean themselves to make sense. But here's the perverse incentive: if you power a carbon removal facility with natural gas, you're creating demand for both the removal technology and the fossil fuel that made removal necessary in the first place. Everyone gets paid. The atmosphere gets nothing.
Let's trace who actually benefits from this structure. Energy companies lock in customers for decades. Tech firms get cheap, abundant power without waiting for infrastructure development. Carbon removal startups gain a business case for their expensive operations. The only party not benefiting is anyone hoping we might actually reduce emissions at scale.
The highway expansion through the Amazon that made headlines recently illustrates the same principle at work globally. Infrastructure gets built based on immediate economic convenience, not climate impact. The roads create new economic opportunities that depend on those roads existing, which makes reversing course politically impossible. We're not making decisions to reduce carbon; we're making decisions that guarantee we'll need carbon removal technology later.
What would real incentive alignment look like? It would mean making renewable power cheaper and faster to deploy than fossil fuel alternatives, even in urgent situations. It would mean carbon removal technology powered exclusively by truly renewable sources, not powered by the very emissions it's supposed to offset. It would mean infrastructure projects evaluated on their full carbon lifecycle, not just immediate convenience.
Instead, we've designed a system where every shortcut deepens our dependence on the longer-term solution. We pay AI companies subsidies to pursue innovation that drives fossil fuel demand. We fund carbon removal startups whose economics only work if we keep emitting. We approve highway projects that lock in transportation patterns for fifty years.
This isn't a conspiracy. It's worse. It's the logical outcome of well-intentioned climate policies layered on top of an economy that still fundamentally rewards the fastest, cheapest option regardless of carbon cost.
The reader should notice this pattern in real time. Every time you see a company announce it's pursuing carbon removal, ask what powers that removal facility. Every time you hear about new data center construction, check what grid it connects to. Every time an infrastructure project gets approved in the name of economic development, look at the emissions it's locking in.
We have the technology to do better. We're choosing not to because it's more profitable to keep doing things the old way while building a market for fixing the damage afterward. That's not climate strategy. That's climate capitalism, and we're funding it with our attention and our policy.