00:00Let's keep it on tech. NVIDIA CEO Jensen Wang saying $700 billion in AI-driven spending from hyperscalers this year
00:05might not be enough.
00:07Lumen Technologies staking its claim in the race to develop AI infrastructure.
00:11The company announcing a new partnership with Anthropic to build out a fiber network across North America.
00:16The Lumen Technologies president and CEO, Kate Johnson, joins us now for more. Kate, good to see you.
00:20I should say hello properly this time. Do you want to start in Lumen Field in Seattle first?
00:23Should we start there? I think we covered it.
00:25When did that relationship begin? The Seahawks and Lumina, maybe 20 years ago.
00:30The sponsorship, yeah. And it's been quite an evolution. We're excited.
00:34This company's changed a lot. A couple of years ago, summer of 24, $1. $1 a stock.
00:40What's changed? What have you changed to get from that position to where we are today?
00:44We have changed everything. We've turned around the capital structure through a couple of important transactions.
00:51Sold $13 billion of private connectivity fabric. Sold our fiber to the home business so we could double down on
00:57enterprise and double down on digital.
00:59Reposition the company to become the trusted network for AI. Total culture transformation so our people are empowered to actually
01:07focus on growth.
01:09It's a new day for Lumen.
01:10How painful has the transition been given that this is new technology that changes all the time and the skills
01:14are quite different?
01:16Yeah. So, you know, it's adaptation, right? I mean, it's about acquiring new skills. It's about injecting new skills into
01:24the company.
01:24It's about reimagining how assets that have been around for a long time provide new value. It's a total refresh.
01:32We find it very energizing.
01:34AI is very disruptive but poses a huge opportunity for our growth because it proliferates data and we move data
01:41for a living.
01:43How much easier is it for you to raise money given the fact that there is this buy-in right
01:47now from Wall Street in the build-out of fiber connectivity,
01:51the build-out of what is necessary to power the AI hopes and dreams?
01:56Well, honestly, we don't need to raise money. We fully funded the next several years of our transformation with the
02:01cash that we've gotten from deal-making
02:03and from redoing our capital structure. So we're in a great place.
02:07I think you'll continue to see us tweak our capital structure, continuing to opportunistically reduce leverage and go after great
02:15rates.
02:16So, I mean, we're in a great place right now.
02:17When it comes to one of your major customers, partners, Anthropic, are you concerned a little bit about the pushback
02:23Anthropic is getting in Washington, D.C.?
02:25So, look, Anthropic is a great partner and they tapped us to be the trusted network for their AI services.
02:31So we're super proud of it. You know, we're connecting their data centers and we're moving at lightning speed with
02:36them to help them grow.
02:38But, I mean, the question remains. New tech often raises complex questions.
02:44We've seen this movie before with Public Cloud.
02:46And, you know, as that discussion evolves, I think it really points to the fact that we probably need stronger
02:53policymaking muscle,
02:54stronger governance muscle, and adjudication. So we're watching it quite closely.
02:59Well, what if the DOD actually says they are supply chain risk?
03:03If they're supply chain risk deemed, I mean, we'll obviously, as we always do, comply with the full extent of
03:09the law, federal regulations and contract requirements.
03:12Does it give you pause about establishing a relationship, given that's on the line at the moment?
03:16It doesn't because no matter what the landscape is at the model level, everybody needs their data to be moved
03:23from anywhere to anywhere quickly, securely, effortlessly.
03:27And we're sticking to that, which is several abstraction layers away from the discussion about the model.
03:32Things are happening really, really quickly. And the numbers, the sums involved are absolutely ginormous.
03:37We had an individual sat in your chair, I think maybe a week ago to the day, who turned around
03:42and said that this capex boom is the biggest bubble he's ever seen.
03:46And the constraints are going to come from things like energy.
03:48Can you give us the counterpoint to that? Because I'm sure that's something you hear a lot as well.
03:52For sure. So there's a couple of different layers to this problem, right?
03:57Whether or not there's a bubble, it's disruption.
04:00There are going to be winners and losers.
04:02There are going to be value propositions and business models that are changed and incumbents that are threatened.
04:10And who wins at the model layer? Who wins at the services layer?
04:14That's up for debate.
04:16What stays constant is the need to move data from anywhere to anywhere.
04:20And that's what we do for a living.
04:21We also got paid up front for building out all the infrastructure.
04:26So we're helping build the supply side of the AI economy.
04:30And we were able to take in all of that cash up front.
04:34And it'll take several decades.
04:35We have more in pipe.
04:36So we know there's more coming along.
04:39I think on the demand side, it gets really interesting.
04:41It's as enterprises start to consume those services, what does that look like?
04:47And I think that's where a lot of the adaptation that we talked about earlier needs to happen.
04:53Aside from some of the bombastic headlines about AI being a bubble and data centers being a bubble,
04:58when you talk to people, one of the fears is the data center that isn't near a big city,
05:03the data center that's out in the middle of Tennessee, and whether that becomes a stranded asset.
05:07You're kind of the answer in terms of connecting all of this.
05:11How big of a concern is that?
05:13I mean, do you think that the connectivity between some of these more remote data centers
05:17is feasible over a longer term with a bigger load?
05:20Honestly, we already cover a lot of the remote data centers.
05:23And the ones that are being built are looking at all of the constraints, energy, fiber, etc.
05:29And we're right there in partnership with them as they build those plans.
05:32So our coverage is an asset.
05:34It gives us an advantage in that marketplace, and we're leaning into it.
05:38It's how we're winning so many of these contracts.
05:40From an energy scarcity perspective, there's a couple different things.
05:44We have a programmable network that allows companies to move data anywhere very quickly in real time.
05:51So we are actually energy's best friend because we can actually pivot workloads
05:58to markets where there's more energy available or where the pricing is more favorable.
06:03The Citadel versus Citrini 7.
06:06I just want to go to this because it really speaks to what you're talking about.
06:09The whole Citadel comment was that the amount of energy that would be required
06:13for the kind of productivity and human replacement that Citrini was talking about
06:17totally undermines the whole argument.
06:19And that we're so far behind it that it just seems absolutely implausible.
06:22At that point, just getting human to do it is much easier.
06:25I mean, how big of a constraint is the energy picture right now?
06:28I think right now the problem is that the workloads and data are trapped in markets
06:34where there are constraints on energy.
06:36But if you look at all of the energy that's available across all the markets in aggregate,
06:41there's plenty of energy.
06:42So the question is, can you move the processing to where the energy is available?
06:48And with a programmable network from Lumen, the answer is yes.
06:52We've explored some of these issues with you before.
06:54I'd love some deeper insight on this particular topic.
06:56The difference between this moment compared to, say, the Internet of 30 years ago.
07:01Back then, the telecoms weren't the people who made the massive money.
07:05The massive money was made at the application layer with some of these big tech firms
07:08that came along, like the Amazons, the Metas of this world,
07:10that produced the kind of things that took advantage of the infrastructure
07:13that was built for them.
07:14Is this different with AI?
07:17It's different for a couple of reasons.
07:19Number one, there was a huge overbuild at that time.
07:22And every project looked like a great project for telecom companies.
07:26They were also in the utility mindset where they were charging rents.
07:31And it was kind of like, oh, there's going to be this huge traffic boom.
07:33We can't wait to get all these orders.
07:35This is different because it's not just about pipes.
07:39It's not just about bandwidth.
07:41It's about the capabilities to connect that bandwidth to the application layer.
07:46It's how everything talks to each other.
07:48That's the essence of our strategy,
07:50is this digital platform actually makes the network programmable,
07:55just like compute and storage were made programmable in the world of cloud.
07:59We've promoted network to the same level,
08:01and now it's become an incredibly valuable asset.
08:04Manage your energy.
08:06Manage your workloads.
08:07Push your data from anywhere to anywhere.
08:08And data applications and people are spread everywhere,
08:11so you need that fiber.
08:12You need that fiber.
Comments