100% totally agree with that, and, uh, just to be absolutely clear about what I'm saying, I'm not at all pinning it all on the Trump administration. I don't think that's right. And as I... When I talk about reality checks, I think, as you say, some of those, uh, incidents you've been talking about are important. I think also we've been learning about the real economic and technical constraints that restrict the growth potential of many low carbon energy technologies. Think about hydrogen in particular, right? So hydrogen, we talked quite a bit about on this show. That's something where making that economically viable, really creating a significant large scale industry in low carbon hydrogen is turning out to be really difficult, and we've had a lot of people canceling, postponing, delaying, scaling back low carbon hydrogen projects. So that's another example of something which, let's say, putting the policy environment entirely to one side, that's kind of a realization we've been coming to over the past year, maybe more. So Shanu, what do you think then, given all this kind of negative discourse around the outlook for low carbon energy, given, say, what seems like pretty poor investment performance over the past year?

Yeah, Ed. Uh, happy to jump in on that. I mean, I think t- to your point, uh, there's like the near term kind of what's happened in the last 6 to 12 months and then what's happened over a multiyear basis, right? So I guess if we, if we rewind, right, over the last five years, you kind of ha- called out that, you know, the S&P has been up, you know, ex- exceptionally well the last two years, but let's call it, on a five year basis, right, so it'll turn kind of up 80%. Energy, kind of like some of the big popular ETFs there, ha- have kept up with that, whereas clean tech has, you know, been very meager, um, you know, on a five year basis, I guess like the Global Clean Energy Index up a s- low single digit percent, so dramatic underperformance. So this isn't like a n- you know, a recent shift in policy or sentiment. This has been a multiple year, I guess, headwind, uh, in terms of performance, and that's due to a variety of things that we've t- harped on before: inflation, um, supply chains getting out of whack, the interconnection queue growing. But to your point, right, I mean, at, at the end of the day, like, the stock market's a representation of investor views of, you know, the present value of future free cash flow. So I think absent execution in a lot of these sectors, you're starting to see, you know, the willingness to look through near term pain to, like, get really constructive on the sector, uh, q- quite honestly, right? So you've seen a lot of people that are not willing to step in just yet. I mean, part of that is the policy question mark in terms of, you know, what's gonna go if these n- some of the sectors, like as you mentioned, hydrogen, for example, or storage or, you know, SAF. You know, these things require some level of, um, subsidies to, to enhance re- returns or make them attractive for investors, and I think given the current administration's stance or disposition towards the sector, it's a little bit more challenging. So I guess kind of we just went through the last reporting cycle, right? Q1's coming through. 2024 finished. Uh, in terms of, like, sectors that did well within broader energy, right, I think you're still seeing a lot of growth in electrification equipment or gas turbine or kind of up and down the gas value chain, which I'm sure we'll get into today, right? So there are pockets of strength within there. Uh, the other area I would call it is utilities and IPPs. It's largely on the back of kind of low growth expectations. And so there are pockets that are doing well. I don't wanna say it's like a whole, you know, like throw the baby out with the ba- bathwater thing. But, you know, I think you have to be a little selective in terms of what you're exposed to. And then on the whole, just being, you know, acknowledging that the fact that performance has been lackluster in recent years.

Not exactly an ideal way to solve that problem. But yeah, no, I do agree, that's a very real possibility.

Can I offer just like two, I guess, tidbits in terms of like this discussion where I think there's like a bit of nuance required where... I mean if we just look at like the overall plant economics of like, you know, let's call it a hyperscale like gigawatt data center facility. Let's call it, you know, $20 million a megawatt, it's like a $20 billion facility. Uh, over 50% of that is the chips, and the chips have a useful life of like three to five years from, from what I understand. Energy over the cost of kind of like that entire data center is less than 10% of cost, maybe less than 5% of cost. So that's the reason that data centers want to run 24/7, right, is that like their economic imperative is to do that, just to maximize the value of those chips, and then the cost is disproportionately higher to that. So I think that's why you've seen this push-pull where they're not signing up right now to be flexible, just because, you know, it really quite frankly doesn't, you know, the economic incentives don't work that way for them to do so. I think where the, it becomes interesting is if a utility says, "Can I plug you in in two years if you're willing to be flexible for these number of hours versus needing to get plugged in in five years?" That changes the calculus, right? Because then the time to power value, or the opportunity cost of not being online faster may supersede the overall cost of chips relative to the, the energy cost. And so I think that's where like it can change. And I think this is where like if you look at some of the announcements that have happened, like for example, Google, Intersect Power, and TPG Rise, there's a $20 billion announcement there and if you li- listen to Sheldon Kimball who's the inter- the founder of Intersect Power, you know, he builds behind the meter type solutions, predominantly renewables and storage with some s- like perage of simple cycle gas turbines. I mean he argues that you can get to like a 70% type clean outcome run-... behind the meter for these large customers and these data centers with the re- liability they need, which again, like that's cleaner than the grid today, and is a near-term solution until you can get grid-tied. So, I think, like, you'll start to see some more creative outlooks there. I mean, he's talking about doing $10 billion worth of deals in the next year, I think, or so. So like these are, this is material too, and we'll start to see some o- some more of, uh, those solutions as well in the interim. But just kind of wanted to introduce that nuance there too, where it's not as easy as- or as, perhaps it's not as simple as like, why aren't they just choosing to be flexible? I think like, you know, there are a few things at play in terms of what's driving their motivations today. The hyperscaler companies represent 50% of the CapEx that's going into data centers. The other 50% is co-location, right? The hyperscalers, if you just kind of count up the aggregate announcements are gonna spend 330 billion of CapEx this year, predominantly on AI. So, I guess at least for the purpose of the infrastructure discussion, I actually don't know if the question is like, like, do consumers, uh, take, you know, whatever AI adoption in the form that it is today? I guess the CEOs of these tech companies are betting that it's gonna get better, that there will be adoption. And I think from an infrastructure challenge problem, that's like the key point to pay attention to, 'cause again, if- they can totally change their mind next year or the year after, right? And again, that, that goes back to the question, why aren't there deals? Is I think people are a little sc- scared, right? Like, oh, if you build all these plants in advance of, you know, this demand going into perpetuity, or at least really strong over the next three to five years, and all of a sudden 27, 28 CapEx numbers come down 20, 30, 40%, that crashes all the new demand. And I think that's why until you start to see these, these more bespoke intersect power behind the meter type deals, or like, you know, maybe some of these plants actually getting built for like via NRG or otherwise, that's the hesitancy you see right now. And the data point I'm paying attention to is like the overall CapEx number and where it's going.