Doug Neal of eLab Ventures argues that the AI race’s true battleground is power infrastructure, not just software development.
The fragmented growth of clean energy sources lacks coordination, slowing AI’s potential scalability.
Data centers’ power demands are rising rapidly, threatening to outpace current grid capabilities.
While chip efficiency improves, renewable energy support from public policy remains insufficient.
The certainty of AI’s power needs offers a rare chance to expand energy infrastructure with minimal risk.
There’s a lot of talk about the ‘AI arms race’ and how to regulate control of the algorithms, and that’s where a lot of the discussion gets bogged down. I think we have to be talking about how to get access to the energy as well. Everyone’s racing to get to artificial general intelligence. The implications of being there first are pretty profound. But without solving the energy equation, no one crosses the finish line.
The AI arms race is charging full steam ahead, with nations chasing AGI as the key to economic and strategic dominance. But while the world obsesses over software, the real battle may be over who has the power—literally—to run it.
Doug Neal, Co-founder and Managing Director of eLab Ventures, has spent over a decade investing in emerging tech. While the fight for AI dominance plays out in the cloud, he argues the more critical battleground is on the ground where power grids, data centers, and infrastructure will decide who can actually keep up.
The power play: “There’s a lot of talk about the ‘AI arms race’ and how to regulate control of the algorithms, and that’s where a lot of the discussion gets bogged down,” Neal says. “I think we have to be talking about how to get access to the energy as well.” To him, the story isn’t just about reaching AGI. It’s about who has the power to sustain it. “Everyone’s racing to get to artificial general intelligence. The implications of being there first are pretty profound,” he says. But without solving the energy equation, no one crosses the finish line.
A new bottleneck: As AI moves from a niche technology to a general utility, the nature of the challenge has fundamentally shifted. “It’s no longer an innovation bottleneck; an infrastructure bottleneck now that we’re moving into broad execution of AI,” Neal says. The problem is a lack of coordination. While big tech and government focus on large, centralized projects, a fragmented mix of clean energy sources—from regional solar farms to home microgrids—continues to grow without a unifying plan. “It’s not organized to be as efficient and collaborative as it could be,” he says. “Done right, it could scale much faster.”
The statement ‘build it and they will come’ is rarely true, but in this case, we actually have that opportunity. If we can make this work, there’s plenty of need for the energy out there. It could unlock a huge opportunity for powering what’s going to continue to be a much more energy-hungry world.
Race against time: The real conflict isn’t between strategies but against the clock. “The real question is, are we going to get there slowly in a piecemeal way, or are we going to get there fast enough?” Neal said. The urgency is already visible in the pace of global development, with projects like the $100 billion Stargate supercomputer planned in the UAE signaling how quickly others are moving. Neal points to the pressure: data center construction is accelerating at an alarming rate, and upcoming facilities could require 50% more power than the entire current U.S. grid—a trajectory he calls unsustainable.
Renewed worries: According to Neal, market forces are already driving innovation in areas like chip efficiency, led by companies such as Nvidia. But that progress only underscores where momentum is missing. “The biggest challenge is renewables,” he says. “Especially today, with the lack of focus on subsidies for renewable solutions, that’s where I get worried.” In other words, while the private sector pushes hardware forward, public policy is failing to meet the energy demands that hardware will soon require.
They will come: What looks like a looming energy crisis also creates a rare moment of opportunity: demand is certain, and growing fast. “The statement ‘build it and they will come’ is rarely true, but in this case, we actually have that opportunity,” Neal says. With AI’s appetite for power all but guaranteed, the usual risks of overbuilding disappear. “If we can make this work, there’s plenty of need for the energy out there,” he adds. “It could unlock a huge opportunity for powering what’s going to continue to be a much more energy-hungry world.”
