It's going to take nuclear energy to power AI at scale, says TeraWulf CEO Paul Prager

So Paul, great to have you on Power Lunch. First, explain, I love this, because you kind of are at the intersection of Bitcoin, mining, and energy. Explain in sort of plain English for the audience, what does TerraWolf do? We're an energy infrastructure shop. Thanks for having me, by the way. It's exciting to be here, especially when you're talking about the New York teams. Now you left out the Yanks. We're an energy infrastructure shop, and we have always been looking for the highest value for our megawatts and energy infrastructure. There isn't a more flexible load out there than a Bitcoin mining facility, and so that is how we started. But now there's tremendous value and opportunity for our shareholders as we pivot into HPCAI. And so we have reserved our next, if you will, 250 to 500 megawatts to build out HPCAI facilities for customers. Okay, I would ask about HPCAI, but I don't know what HPCAI stands for. I'm guessing the AI is artificial intelligence. Right, high power compute, artificial intelligence. You know, it's what all the hyperscalers are looking, right now they're trying to capture energy infrastructure and power. Our focus is in the co-location side of that business, where we'll look for people that have enterprise customers, and we can provide energy and facilities for them. So here's why I love having you on, Paul, because, and I think you can put it in plain English, because energy is your world even before Terawolf, which is this, you said 500 megawatts. A lot of people don't understand what that means. That is probably going to mean, what, about, that's a couple hundred thousand homes worth of electricity. We're talking about Three Mile Island coming back online, Amazon making nuclear deals. Can we do any of this stuff in a macro way without nuclear? The answer is, I don't think so, but nuclear's going to take a real long time. I think the near-term solution are large gas-fired facilities, and the acceleration, to the extent possible, is some of the renewables. But nuclear's going to take a real long time. I think we're, you're looking at 10 years away. In terms of new generation for new, in terms of restarting the facilities, I mean, we have to see. It's a technology that we've lost for a little long, for a little bit of time, and that we have to bring back to the United States. Real quickly, Paul, why are you selling this one sort of project at a time when, is it in order to fund the bigger projects that you now see as more attractive? Yes, but it, you know, we were just a 25% interest in a joint venture, and in fact, we got a great return on the deal. It was 3.4 times. More importantly, we were selling at the time when our interests were most valuable before they go to nothing. I mean, we had three years left on a contract of two-cent power, and then there was no residual value left. So we got a chance to get a great return on our investment, take something, sell at a real premium, and then return it to our shareholders in the form of investment in our HPC AI business, renewing our Bitcoin fleets. It's going to be 18 point joules per terabyte, the most efficient fleet out there, low cost of operations, and we just have increased liquidity for growth. It's a great deal for our shop.