Law, Policy & Markets
Law, Policy & Markets
The Digital Infrastructure Revolution: AI Drives Massive Investment in Data Centers and Connectivity
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
We are experiencing a revolution in digital infrastructure. Artificial intelligence (AI) gets most of the attention, but the appetite for expanded capacity – and the magnet for new public and private investment – affects myriad digital technologies. Broadband fiber, data centers, wireless spectrum networks, and advanced chip manufacturing are attracting billions of dollars in new funding.
In this episode, “The Digital Infrastructure Revolution: AI Drives Massive Investment in Data Centers and Connectivity,” host Allan Marks speaks with Milbank partners Patrick Campbell and Jaime Ramirez. They examine how the CHIPS and Science Act of 2022 and other federal programs together with the explosion of new artificial intelligence technology and the impacts of the pandemic are driving a massive wave of investment in fiber and data center projects.
About the Speakers
Patrick S. Campbell is a partner in the Washington, DC office of Milbank LLP and a member of the firm’s Transportation and Space Group. Mr. Campbell’s practice focuses on transactional, corporate and regulatory matters in the digital infrastructure space. He has worked extensively on such matters on behalf of domestic and foreign fiber optic, wireless, data center, tower, satellite, computing service, media and Internet companies, as well as banks, private equity firms and other entities investing in, or providing securitization or other debt financing for, these companies.
Jaime E. Ramirez is a partner in the New York office of Milbank and a member of the firm’s Global Project, Energy and Infrastructure Finance Group. He focuses on representing parties in the acquisition, development and financing of large scale natural resource and infrastructure projects. His experience also includes the representation of investment banks and borrowers in cross-border financing transactions, particularly in Latin America.
Podcast host Allan Marks is one of the world's leading project finance lawyers. He advises developers, investors, lenders, and underwriters around the world in the development and financing of complex energy and infrastructure projects, as well as related acquisitions, restructurings and capital markets transactions. Many of his transactions relate to ESG and sustainability, innovative clean technologies, and sophisticated contractual risk allocation. Allan serves as an Adjunct Lecturer at the University of California, Berkeley at the Law School and p
For more information and insights, follow us on social media and podcast platforms, including Apple, Spotify, Amazon Music, iHeart, Google and Audible.
Disclaimer
Law Policy and Markets. I'm Alan Marks. Today, I'm joined by Patrick Campbell, a partner in Milbank's Washington DC office and a member of the firm's Transportation and Space Group, and Jaime Ramirez, a partner in Milbank's Global Project Energy and Infrastructure Group, based in New York. Let's get to it, patrick.
Speaker 2Jaime, thanks very much for taking the time to get together today, thank you. Thank you for having us.
Speaker 1Thanks, alan. So, patrick, let me start with you. Digital is suddenly all the rage, with data centers and AI impacts and new broadband cable networks being laid everywhere. But you've seen this for a long time. What's striking to you about the current market, compared to the legacy digital work you used to do?
Speaker 2So I've been a lawyer for nearly 30 years and I have worked exclusively in this digital sector.
Speaker 2However, for the first 25 plus years of my career, we called it telecom, or we called it TMT for telecom, media and technology.
Speaker 2Something shifted right around four years ago, right in the midst of COVID and right after, where the field became essentially rebranded as digital infrastructure, which is really something that I factors that led to that.
Speaker 2It was just the incredible growth in the need for bandwidth for streaming, in the need for the growth of cellular phone technology and also COVID, causing people to work from home and needing more infrastructure to communicate with their office networks, and so investors that in the past used to look at the infrastructure that drives our economy, such as roads and bridges and energy and the like, started realizing that the sort of digital world of connectivity whether it's fiber or data centers or wireless spectrum or towers and all of those things were actually becoming so critical to our economy.
Speaker 2They deemed it infrastructure and started putting money that used to be dedicated for the other infrastructure projects that we were used to and started putting it into digital, and so we saw this sort of increase in the need for infrastructure coincide with this new source of investing where all of a sudden, billions and billions of dollars that wasn't initially allocated to this sector suddenly became available, and so everything shifted. Lawyers who used to do real estate deals were suddenly looking at data centers. Lawyers who used to do project finance we're suddenly financing fiber projects and data center projects, and everything shifted very quickly, and so now pretty much when you think about telecom and connectivity, it's all now branded digital infrastructure, which sort of pulls all of it together in one nice big package.
Speaker 1That's interesting because we have then stair-step technology innovation driving it.
Speaker 1We have changes in demand patterns that we really saw.
Speaker 1I think you're right with remote work coming out of COVID, but this idea that the digital infrastructure is a different sector or a newer sector or a kind of a hybrid sector is pretty interesting to me, and I mean I kind of want to pick up and get your thoughts on this, because you've done work, as I have, in other sectors.
Speaker 1The essentiality of connectivity is one hallmark of infrastructure, but there are things that make it very different and maybe even more like real estate or more like corporate finance in some ways, and less like infrastructure traditionally, like toll, roads or power plants. What have you in the contracting structures depend on, or maybe have more risk in them, because you have corporates that are, at the end of the day, betting on the demand for all this new capacity and you're betting on their balance sheets. Stick around for the length of time that it takes to repay these massive amounts of debt Some of them are taking on and, as a result, we actually see significantly more equity relative to debt in the digital infrastructure asset class than we do in traditional project finance. So how do you treat this? Is it infrastructure?
Speaker 3It's a good question, alan, and I fall into the bucket that Patrick was talking about, in that I've been doing this for almost 20 years, but really focusing on digital infrastructure in the last couple of years, as Patrick noted, it migrated into other asset classes. Noted, it migrated into other asset classes. The literal way to look at this, on the one hand, is we're in a moment, we're almost we're rebuilding the Eisenhower highways for the future. So in a very literal sense, it really is infrastructure, because, regardless of whether we know what the final application is for digital infrastructure, we seem to have a solid sense that it is and will be required for the future. The common question that everyone asks is okay, we know that the hyperscalers Microsoft, Google, amazon they need all this data center capacity. The next question is what for and post-COVID, it was for data storage or to facilitate remote work. Then chat GPT happened and almost overnight it became an AI story in large part.
Speaker 3So, to address your question about from a contractual perspective, I do think it's infrastructure. It has hallmarks of infrastructure, of the infrastructure asset class. There are defined off-tick agreements with investment grade entities. They are long-term contracts, they are capital intensive, but they are very much a hybrid of other asset classes, in particular in the data center space, and that's why the financings look very much like a hybrid between real estate and infrastructure. I think the more important question, I think, is what will this infrastructure be used for? So that we can answer the question that often comes up are we in a digital bubble for the asset class?
Speaker 1And I think A digital bubble, in the sense of there being overinvestment and excess capacity.
Speaker 3Exactly, I think that the obvious question a lot of people ask over and over again is there's so much money going into this at such a rapid pace? Are we in a digital bubble? And I think there are a couple of ways to approach that. On the one hand, supply is not outpacing demand by any stretch, in large part because of some of the roadblocks in development for digital infrastructure, power being at the top of the list. So even if we keep building at this pace, because of the lack of power going forward relative to the need for data centers, the demand is going to far outpace the supply for the next five to 10 years.
Speaker 1I want to pause on that for a second because if I look at, for example, I used to do work years ago for companies like Level 3 and Enron on their telecommunications side and Global Crossing, and there we really did see a build out of broadband fiber networks both within the US and between continents and around the world. That where the capacity far outstripped the demand growth for that capacity. But this is different here. If you talk about, for example, shortages of power supplies in places where people want to build out new data centers, just for example, that's a very real issue. It's an issue for the next person coming in, which in some ways creates more pressure for a first mover to do this and accelerate construction, if you possibly can. With what's available now. How does? Are the players gonna change? Is that or is the power grid gonna adapt? How is that gonna play out?
Speaker 2I think about it this way we're living in two completely different worlds, right? If you think about the way we grew up, if you wanted to see a movie, you'd go to the theater. If you wanted to read a book, you'd go to the library, and if you wanted to plan a vacation, you would go to a travel agent. Now we do it all digitally, right? We book our tickets, we read our books, we watch our movies.
Speaker 1And we do it while we're on the move on our phones. But that whole telephony piece of this, which we don't talk about as much, is also pretty important with 5G.
Speaker 2Yeah, but when you think about the level threes of the world, they were mostly focused on enterprise connectivity. Now it's all about connecting everybody. If you look at the world of video gaming and the importance of low latency for people playing games, which is just an incredibly growing industry, and then you'll see the demand is, in my view, will continue to grow because people individuals are now demanding the kind of power that companies used to say hey, we need X amount of data, we need X level of speed. Now, every home needs the same level or even more so data and speed in order to function in today's world, and so I think the way we operate is changing so much that we are now seeing incredible growth in demand for the technology.
Speaker 3Patrick raises a good point, because we tend to think about digital infrastructure as if it's responding to one particular need. But in this case there are actually various needs for data center, fiber, etc. One is the enterprise data storage capacity. The other is AI training for future applications, right, so it's very different models and you're going to need both in the future. And data certainly is not decreasing. Incredible statistics out there, that sort of data in the world multiplies within a relatively very short period of time. That's not going to stop. And then you see on the AI side, just an intense need for training at this point. Then you're going to talk about whether it's refining the training models et cetera going forward, but at the moment it's a heavy need to train.
Speaker 1I want to circle back to this idea that you mentioned about latency, because I think it's something that's often missed in the conversation. People talk about, for example, putting big data centers in the middle of North Dakota or in the middle of other places that are pretty remote from population centers, because latency is just not an issue. If you're using AI and big data and large machine learning, language models and things to try to solve problems, latency is not really a big issue. Right, because you can send it up there and you get your answer back and the world is fine. But if you're Netflix and you're trying to get movies onto people's screens without a lag, then it actually matters very much that you've got your servers with redundancy close to where that is.
Speaker 1And you mentioned gamers. If any of us have kids or ourselves over the age of 10, we've probably done it and we see how that can make a big difference in the experience. So that does definitely put pressure on the last mile of fiber, but also on the location of some of these server farms and things that are meant to service it. How does that play into capital allocation? Because so businesses that are location sensitive and where real estate might be expensive or scarce or connectivity is congested, are going to look very different than ones that are, say, being located close to a remote power plant that's otherwise shutting down, in a place where there's otherwise low competition for power and the land is cheap.
Expanding Fiber Infrastructure for Equality
Speaker 2What it's really led to is investment in parts of the industry that previously did not see as much investment. So, for example, edge data centers, bringing smaller data centers closer to the cities to service particular applications and customers. That's grown incredibly over the last few years. Fiber to the home it used to be sufficient that you would build fiber into a neighborhood and then put a coax or copper cable into someone's home. That's no longer sufficient. We used to have satellites that were far away. We call them geostationary satellites. Now we have lower earth orbit satellites. We've brought the satellites themselves closer in order to reduce the latency.
Speaker 2I'll tell you a quick story. I was at a conference recently and was talking to a young man who said he had a wonderful fast internet service at his home using coax cable and he thought it was amazing for work. And his kids came to him and started complaining and said dad, we have to upgrade to Google Fiber, which was what was available in the neighborhood. And he said but why? We have a very fast connection. They said look, we played Call of Duty and Call of Duty publishes our latency and our latency is just enough slower than others that we're being picked on. And so gamers with fiber into their homes are picking on gamers with coax cables and shooting them way faster than they could ever shoot back.
Speaker 1There's some kind of a deep social metaphor in that, I think, shooting them way faster than they could ever shoot back.
Speaker 2There's some kind of a deep social metaphor in that, I think, and so that's why that tells the story, of why now we're having fiber built directly into homes, which is something that just wasn't a big concern before, because who needed that much speed and that much two-way speed into homes? Well, now we do, and so there is incredible investment now going into fiber projects. Even companies like T-Mobile, which historically has been a wireless provider solely, is now putting billions of dollars into investing in fiber and creating that package that people need, because, from their perspective, the Verizons and AT&Ts of the world will take all their business away, because customers are now looking for these fiber options as well, and if they can combine it with their wireless packet services, then they can sell to a lot more subscribers, and so the importance of fiber has it cannot be overemphasized.
Speaker 1So let's look at that because obviously getting fiber in Manhattan or Los Angeles, especially if you're close to a nicer neighborhood or a commercial center or a bunch of high-rises, that's going to be pretty easy. There's a lot of part of the countries which is rural and there's a lot of part of the country which is underserved, urban communities that are not getting that level of service. So enter the government right and we've seen really a lot of different models for how you build that out. So Kentucky as a state went in the public-private partnership model to wire the state including parts of it that would not be otherwise economic for private companies to hook up.
Speaker 1We're seeing the federal government now being extremely active under the current administration and the recent laws passed by Congress in supporting federal funding, usually reimbursing states or municipalities for the cost of those kinds of connectivity improvements, and that's different. We'll get to subsidies on the data side. But I want to just look maybe on the connectivity fiber and the intersection with telephony, including cellular telephony, and how that the government's role is changing, because it's not just a regulated monopoly with a rate of return that they're allowed on their costs that then they pass through to rate payers. This is a different kind of a model for how the government is supporting that expansion, much more like we did 100 years ago for rural electrification.
Speaker 2If you think about it, if you go with a purely economic model the cost of investing in fiber into certain neighborhoods, as you mentioned, certain urban areas where people are probably less willing to pay as much, or areas that are harder to reach it wouldn't really make sense for a company to build fiber into some of those communities because they just would never recoup their investment.
Speaker 2And so, if you allow it to be purely economical, what's going to happen is you're going to have two classes of society where some are connected and some just simply aren't, because it's just not worth it to the companies. And so, as you said, the government has stepped in and has created a number of programs. The World Digital Opportunities Fund is one. There's another one called BEAD, where billions and billions of dollars are being poured in by the government to essentially subsidize the build out of fiber, with the notion that it's almost become a bit of a right that we have as a society, right, and we think that if there are kids in neighborhoods who can't connect as quickly because companies don't find it worth it to connect them, then it becomes the government's responsibility to ensure some level of equality.
Speaker 1Which, by the way, is why we want to have the government make sure that we have running water and electricity and a road that's paved outside the door.
Speaker 2It has. It has it's necessary.
Speaker 1Yeah, but not regulated like one in the same way. That is different.
Speaker 2That is true. It's not regulated in the same way and I think there is some sense that there is competition. That's enough to pull back on some of the regulation. As far as pricing and rates and the like, the rate pressures are really strong on connectivity Companies if they wanted to raise rates. There are people coming in and overbuilding in areas where rates could be raised and so when we do these deals, there's a big focus on companies making sure that, hey, if rates go down, there are financings where they say, look, we're setting the rate at this rate, but it could go down, and you have to understand that's going to be a risk to a lender, because if we're we used to collect 50 bucks a month, we ended up having to go down to 30 bucks a month and that's a risk that people have to take because, at the end of the day, there's pressure from competitors.
Speaker 1Patrick, that's really good. I want to ask you, jaime, then too, I want to pivot, because we've talked about fiber and connectivity. We didn't want to talk about cell towers we can come back to that if you'd like to, but that's also a piece of that and we've talked about data centers. What we haven't talked as much about is something which also serves a different level of government involvement, and that's the advanced logic chip fabrication facilities. So these foundries that are receiving significant incentive in the United States under the CHIPS Act. For people outside the United States, obviously, other countries are doing similar things. Korea's got a significant subsidy for chip manufacturing. You will continue to see. Europe has adopted something similar in order to remain competitive. A year ago, the United States made exactly zero advanced logic chips and I think the hope that the Commerce Secretary has expressed is that we'll control. 20% of the world market will be made here within short order.
Speaker 1So the CHIPS Act was passed just over two years ago and, hami, you've done a lot of work on some of these significant investments where we're marrying public money with private money. So far, just in those two years, we've seen $30 billion or so deployed directly as federal grants under the CHIPS Act for mainly manufacturing facilities. That's created over 30,000 or expected to over 30,000 jobs for those factories, not counting the construction jobs that are already now blowing and going where they're building them. It's gone to 13 companies in 14 states and 23 projects. It is significant investment and that $30 billion has been catalyzed, including that, a total of over $200 billion of new public or private investment. That's a lot of change in a very short period of time. Would you consider the CHIPS Act so far to have been a significant success, or are there things you would wish were different?
Speaker 3I think it's definitely helpful and I think it's certainly attracted business to the United States and activity levels for the CHIPS Act has picked up. I think the commentary, if you were to ask people, is perhaps that the money hasn't been deployed fast enough, Because the capital again, the capital needs. When we talk about data center capital needs, we're not even talking about how much it costs to put the chips and the servers in the data centers, which is in some cases multiples of what it costs to build a data center, and so there's still a lot of capital that needs to be deployed to even get anywhere remotely near what those objectives are.
Speaker 1Can I ask you to clarify one thing though? Does that mean that people there's either a shortage of liquidity, which I doubt, or does it mean that they're not processing these things quickly enough, which, based on the progress so far, may not be the issue, or there are other constraints? For example, we mentioned availability of power, permitting and labor and things that are preventing the users from being able to suck up even more money that the government or investors are willing to provide.
Speaker 3I think it's a combination of all the above and also note that there are not that many are willing to provide.
Speaker 3I think it's a combination of all the above and also note that there are not that many as you mentioned, it's a very specialized industry and there are not that many companies out there and this is part of the problem, right that can actually do this.
Speaker 3So, in terms of deploying the money, there's not a whole universe of companies, like in the data center world, that you can actually deploy this money to accomplish what the stated objective is. So part of that is attracting those companies in attractive states where there's labor where there's, and trying to marry that. You may have seen in the New York I think it was the New York Times would have an article where there was cultural tensions between the Taiwanese manufacturing company coming into Arizona and the local labor population and just a cultural clash. So I think there's going to be some period of adjustment and I think it's just going to be a product of the fact that, one, these are very capital intensive projects. Two, there are only a handful of companies in the world that can do them and three, there are obviously sort of regulatory labor barriers to entry that we're going to have to address.
Speaker 1So if you look at the trends going forward, would you expect that they will. How long will this kind of runway go of a massive investment in new manufacturing capacity in the area?
Speaker 2Well, think about the fact that Apple just released the iPhone 16 that has built-in AI right.
Speaker 1Messing up my photo feed.
Speaker 2Well, right Up until this point. Ai, while it's been growing, still isn't necessarily something of the masses. Right, it's still with the people that kind of are in the know. Now it's going to be something that almost everybody uses to some extent, be required, as that starts catching on, to support that kind of AI. We're already seeing incredible investment in GPU installments which are designed to service some of these AI models, which are only growing. And as that becomes more mainstream and there's more and more students and people at home start using AI for recipes and vacation plans, you're going to see more and more of a need for data centers, for chips, for power, for all of the things that drive these services, and so I think the investments will continue. We're talking to clients on a regular basis who are looking to invest in what's going on and what's to come, and so I think there is no end in sight as far as I can tell.
Speaker 1Let me ask about power, though, because power is an interesting constraint here, and we're recording this right now, sitting all together in a conference room in New York and I happen to be here for climate week and one of the interesting panels that I heard was looking exactly at this and one of the speakers made an interesting point.
Speaker 1He said today Data centers and related or ancillary digital infrastructure consumes around 2% 2% to 3% of the world's power, and within seven years, that's going to be somewhere between 5% and 8%, maybe even people say as much as 10% electricity in other areas. His point, which is provocative, is that, yeah, but if you're using that power because of the super hungry power demands of AI in particular, and you can apply AI to smart applications like making battery storage more efficient, optimizing discharge and recharge cycles on utility scale batteries, like making the grid smarter, like significantly improving the energy efficiency of the building sector, which is 15% or more of our emissions, the net gain from AI is actually significant, and one reason you see those large percentage changes is not just because you're using more power for it, but because you're using less power other places than you otherwise would in the absence of AI.
Future of Data Center Technology
Speaker 3Well, what I would add to that is it is persuasive, and also I've heard those numbers as well. While I do expect power consumption to increase dramatically in this space, I think that there's also going to be a lot of technological efficiency coming in, in addition to the efficiencies that AI could bring itself, but the actual use of energy by these chips etc. Is expected to become more efficient going forward. So I think there's going to be a combination of yes, there's going to be a strong growth in power consumption, but I think it might be driven more by the volume of use than the actual use by the GPUs in the future.
Speaker 2Yeah, I think that's a great point, and in fact I actually was just at a conference and met with a company that has invented and is developing a sort of diamond mining concept to put diamond particles into semiconductors in a way that would exponentially reduce the need for power to keep them cool. And so to Jaime's point. I think as this world, as the sector, grows, so will the technology and so will the innovation, and so the power needs, because a lot of the power really is needed to cool GPUs that get very hot.
Speaker 1Yeah, if you look at your typical data center, half the power is electrons flowing through the machines and the other half is keeping them cool because they're wasting so much energy and giving off a lot of heat.
Speaker 2Right, Exactly right. So if we figure out new ways to keep them cooler without using as much power, then, boom, we can reduce the power.
Speaker 1So let's make those diamond particles synthetically out of captured carbon. So let's make those diamond particles synthetically out of captured carbon. And now we've.
Speaker 2There you go, and smart people will always come up with smart ways to do it and the other thing that's going to happen is, I think, the grid itself is going to increase, in the sense that right now, if you're building a data center, you have to figure out where your power is coming from, and in fact, what we're seeing are things we call like behind the fence power generation, where a data center is almost becoming a mini power plant, and so, instead of relying on the existing grid, data centers are creating new grids and new power sources in order to power themselves, and so I think that is going to change the world of power in a way that kind of coincides with the growth of data centers, so that it's not like they're just going to be taking more and more of the existing power. They're actually going to be creating new technology, renewable power, et cetera. That's going to, I think, improve for everyone over time.
Speaker 1Yeah, it's interesting this week or in a few days, england is scheduled to turn off their very last coal-fired power plant. So after what? Since 1860, now here we are in 2024, and that'll be done. And what replaces it is renewables, by and large, but also different generating sources, and you can imagine, inside the fence or behind the meter, generation for data centers that are in the future, perhaps done with small modular reactors that are relatively safe and inexpensive compared to traditional nuclear power and are not intermittent other things that could be useful. So I think it brings the full circle back to that asset class question.
Speaker 1Right, because in some ways there's a separation, where we now have digital, which is something distinct from power or traditional infrastructure, transportation, mobility and distinct from real estate. Right, it's a hybrid of those, but it's its own thing because it's got scale. But we also see a convergent, because at some point digital infrastructure is actually just a proxy for energy. And when you ask how big is your data center, the answer is in power measurements, it's megawatts, right, it's not in bits or memory, it's how much power is it using? Or that's for the processing speed, really what matters. I find that fascinating and I don't know to wrap it up, if you think we're going to see more convergence among the investors and the people allocating capital, maybe even among regulators, in looking at this, or if it will continue to grow in ways that are distinct from other traditional infrastructure sectors.
Speaker 3I actually think it's a very elegant way to conclude, because I think it will be its own separate asset class, in part because I'm amazed every day by how much, once you peel the layer of the onion back, there's always something else that's growing. So you look at a data center, that's really just the box. Then you look inside the box it's the chips and the servers. Then you look at the cooling systems, then you look at the chip manufacturing and that's without even getting to what the ultimate application is. And in this day and age, we're starting to talk about robots, right, and that may be one of the big end users for AI. And so what AI and what all of this is driving towards, I think is going to be revolutionary in a lot of ways.
Speaker 3I'm trying to temper a little bit to not sound so exaggerated, but I do think that there's going to be revolutionary applications of AI coming, and I think there are things that we haven't thought of today that are going to come out in the next couple of years.
Speaker 3I think things that we expect to be here 10 years from now are going to be here a lot sooner than we think expect to be here 10 years from now are going to be here a lot sooner than we think, and I think financing all of this, I think investing in all of this, is going to be very distinct, because it requires a different level of understanding, but also an understanding of how it all fits together, and so I think the reason why it's so in many ways apropos, that it's come into project finance and infrastructure, is because everything is connected right. The data center is connected to the powers, connected to manufacturing, and so I do think this sector is going to evolve somewhat independently, just because it is such a large ecosystem, different than perhaps just traditional highways or power plants or pipelines.
Speaker 1Yeah, I appreciate that. So more hope, not hype.
Speaker 2Yeah, I like that. What I would add is look, at the end of the day, it comes down to humans and applications. It comes down to what we want to do as people. Do we want to get around faster? Do we want to order food quicker? Do we want to get content more conveniently?
Speaker 1And waste less time on hold and less time waiting for things and actually get it done.
Speaker 2That's exactly right, and if you're an investor in any sector, your goal is to make money and figure out what do people want and how do I respond to that. And so all we're seeing with this digital infrastructure is a recognition that we have become a digital world and that we are relying more heavily on the digital infrastructure than we ever have. And so that's why we're seeing all this investment, we're seeing all of this convergence, and I think we're going to continue to see it as long as people continue to demand AI, robots, automation, low latency, increased power for computing. As long as that grows, innovation will continue and we'll find newer ways to serve those needs.
Speaker 1And it's actually a global phenomenon too. It strikes me that it used to be that you'd have economies, countries, that were based on natural resources traction, and then you'd move up the food chain and they were based on manufacturing and goods output, and then the highest, wealthiest economies gradually shifted, or maybe rapidly shifted, to being basically service sector based, which means you need information, and that's where digital can add huge amounts of productivity, both as far as speed scale again, but also the creativity that people can bring. And now what you'll see is the ability of a lot of countries that otherwise might have been stuck trying to to inherit legacy technology, in fact leapfrogging and improving rapidly some of their their systems communications, energy and others so that they can actually jump into services and creativity that are better for their populations without that step of basically lower value added, either commodities or manufactured goods. So it could be an interesting revolution globally, I think.
Speaker 2And look, there's also that national security angle, where you talked about the CHIPS Act and the need for us to start making and investing more in our own chips. We can't allow others to start dominating these essential items that we need for our own security and our own sort of independence, if you will. And so I think, even within the world of satellites, we're starting to see systems that are developing almost regionally, where you have a company called Lightspeed that's sort of Canadian based, then you have another one that's European, the sense that governments have to invest in systems that are somewhat tied to their regions because, at the end of the day, you don't want to be in a conflict and have to rely on the system, potentially your enemy, frankly.
Speaker 1Well, I think, if you look at even just the United States context, you have CHIPS Act in the context of the bipartisan infrastructure law and of the Inflation Reduction Act, which really focused a lot of people, which is on climate, but there's also things that are healthcare and a lot of other stuff. But it's really a marriage of supply chain security from national security purposes, as you say, and for economic purposes and industrial policy and job creation. It's not just about infrastructure or climate or energy or what have you. Yeah, well, I want to thank you both very much. Jaime Patrick, this has been a great conversation and I appreciate you taking the time away from the massive digital deal flow to share your expertise.
Speaker 3Thank, you, thank you.
Speaker 1Thank you. Thank you for joining us on another episode of Law Policy and Markets Milbank Conversations. Follow us on your favorite podcast platform and learn more at milbankcom.
