Go Where the Energy Is. Then Find Out Who Owns It.

There's a startup up in Portland called Panthalassa, and the plan is genuinely cool. They build big wave-powered platforms, float them way out in the open ocean, and let the swell bob them up and down forever. That motion drives a turbine, the turbine makes power, and the power runs AI compute right there on the platform. The answers come back to shore over satellite. No grid hookup, no land, no substation, none of the permitting fights that bog down a build on land.

Their tagline is "go where the energy is," which I love, because it sounds obvious right up until you notice almost nobody actually does it. The pilot, Ocean-3, is headed for the northern Pacific later this year, with real commercial deployments planned for 2027. The team came out of SpaceX, Blue Origin, NASA, naval architecture programs, and at least one polar research station, so these are not folks who are bad at hard things. And they set it up as a public benefit corporation from day one. Good on them.

I want to be clear about something before I get to the next part. I think this is great. The honest version of the AI energy problem is that demand is climbing way faster than the grid we built on land can carry, and most of the proposed fixes are some flavor of "build more of the thing that's already buckling." Panthalassa looked at that and said let's just move the whole operation out to where the energy already lives and never quits. By their numbers, a platform like this makes power close to ninety percent of the time, way more than offshore wind or solar pull off. Clean, cheap, always on. Count me in.

So here's the question this whole gold rush keeps tiptoeing around.

Cheap clean power is the easy half. Who owns the thing it runs is the half nobody wants to talk about.

You can run a datacenter on the cleanest, cheapest electrons ever pulled out of the sea and still hand-build the exact world we already have, the one where a few companies own the compute, set the price, and meter your access to your own future. The electricity gets greener. The ownership sits right where it always was. Everybody else pays by the call.

That gap is the whole reason Current exists. Panthalassa is building the power and the boxes that run on it. We care about who owns the boxes. Two different problems, and solving the first one beautifully does nothing at all for the second.

We have done this exact thing before

Current is built on the rural electric cooperatives, the ones that wired up the American countryside back in the 1930s, and that history is worth two minutes because it rhymes hard with right now.

When electricity showed up, the big investor-owned utilities ran wire out to the cities and the nice suburbs and then stopped. Rural America was too spread out, too few customers per mile, not worth the money. So you had farms sitting a couple miles from a power plant, lighting the house with kerosene lamps. The technology existed. The business model just pointed somewhere else.

So the farmers did the obvious thing and built it themselves. Neighbors threw money in a pot, formed a cooperative, strung their own lines, and owned the whole works together. The people using the power were the people who owned the power. The profit, what co-ops call patronage, came back around to the members instead of riding off to some shareholder three states away. Inside a generation, the lights were on just about everywhere.

AI is having its own version of that moment right now, and the same fork in the road is sitting in front of us. The compute is getting built. The energy is getting solved, by Panthalassa and plenty of others. The only real question left is whether regular people end up owning a piece of this thing or renting it from a few landlords for the rest of their lives.

The meter runs one way

That's the line we keep coming back to. The meter runs one way. You feed it your data, your prompts, your work, and your money, and the value flows out of your life and into somebody else's quarterly numbers. A wave platform in the Pacific does not change that direction. It just makes the meter cheaper to run.

Now point that same abundance at a co-op and watch what happens. The members own the nodes. The members vote on how the network runs. The surplus circles back to the members as patronage instead of leaving as profit. Same electrons, completely different answer to who the whole thing is actually for.

That's why I keep an eye on companies like Panthalassa. The abundance they're chasing is real, and it's a gift to anybody trying to build something outside the hyperscaler model. Every serious push toward cheap, clean, always-on power lays down more of the ground a member-owned network gets to stand on.

Go where the energy is. Bring the ownership with you.

I'm going to steal their tagline, because it's too good to leave lying there, and then I'm going to finish the thought.

Go where the energy is. The ocean, the desert, the empty places the old grid never bothered with. Build the clean, almost ridiculous abundance the next ten years are going to demand. Please, go do it.

Then make sure the people plugging their lives into it own a slice of the thing they're plugged into. Build the energy like a public benefit. Build the compute like a cooperative. Do both and you get an AI economy where the value finally runs back toward the people doing the living instead of away from them.

The energy is out there in the deep blue, and I'm glad serious people are going after it. The ownership part is the one we're here to build.

Current is a member-owned cooperative AI compute network, built on the same idea as the rural electric cooperatives that lit up America. The people who lean on intelligence every day should own a piece of the machine that makes it.