1 00:00:00,240 --> 00:00:04,680 Speaker 1: Hi, It's sucksh data centers for artificial intelligence consume huge 2 00:00:04,680 --> 00:00:08,640 Speaker 1: amounts of energy. They also happen to raise electricity prices 3 00:00:08,920 --> 00:00:12,920 Speaker 1: for people living close to those data centers. That's according 4 00:00:12,920 --> 00:00:16,439 Speaker 1: to an investigation from my colleagues's at Bloomberg News looking 5 00:00:16,520 --> 00:00:19,319 Speaker 1: at US data centers that you can hear about in 6 00:00:19,360 --> 00:00:22,439 Speaker 1: this episode of the Big Take podcast. Hope you find 7 00:00:22,480 --> 00:00:25,520 Speaker 1: it insightful. Zero will be back later this week with 8 00:00:25,600 --> 00:00:28,360 Speaker 1: a new episode. 9 00:00:29,240 --> 00:00:35,440 Speaker 2: Bloomberg Audio Studios, Podcasts, Radio News. Hey, mister Stanley, how 10 00:00:35,479 --> 00:00:36,680 Speaker 2: you doing. 11 00:00:36,800 --> 00:00:39,400 Speaker 3: I'm good, So I'm Josh who talked on the phone alright. 12 00:00:39,840 --> 00:00:43,000 Speaker 4: Earlier this month, Bloomberg reporter Josh Saul visited a man 13 00:00:43,080 --> 00:00:47,040 Speaker 4: named Kevin Stanley. He's fifty seven and he lives in Baltimore. 14 00:00:47,280 --> 00:00:48,839 Speaker 2: Do you'll need me to turn the lights on more? 15 00:00:48,840 --> 00:00:52,200 Speaker 5: Because this time he. 16 00:00:51,200 --> 00:00:53,479 Speaker 4: Josh was in Maryland to report on the effects that 17 00:00:53,640 --> 00:00:57,400 Speaker 4: data centers have on people who live near them. Kevin 18 00:00:57,440 --> 00:01:00,279 Speaker 4: lives about a two hour drive from an area known 19 00:01:00,280 --> 00:01:04,360 Speaker 4: as Data Center Alley, the world's largest concentration of data centers, 20 00:01:04,560 --> 00:01:08,480 Speaker 4: topping other hotspots like Iowa and Oregon, and Josh was 21 00:01:08,520 --> 00:01:10,919 Speaker 4: there to talk to him about one thing, in particular 22 00:01:11,280 --> 00:01:14,720 Speaker 4: his electricity bill, which has gone up ever since data 23 00:01:14,760 --> 00:01:15,920 Speaker 4: centers came to the region. 24 00:01:16,480 --> 00:01:19,160 Speaker 2: Kevin's a really nice guy, really interesting guy, and he 25 00:01:19,200 --> 00:01:22,399 Speaker 2: talked about how his high power bills have had a 26 00:01:22,400 --> 00:01:24,480 Speaker 2: really rough effect on his life. He has had to 27 00:01:24,520 --> 00:01:27,560 Speaker 2: cut back on buying the groceries he likes, getting haircuts 28 00:01:27,560 --> 00:01:29,160 Speaker 2: as often as he likes. He tries to make his 29 00:01:29,200 --> 00:01:30,840 Speaker 2: diabetes medication stretch out. 30 00:01:30,920 --> 00:01:33,600 Speaker 5: The power bills just keep going up and up, and 31 00:01:33,840 --> 00:01:37,120 Speaker 5: for me, I'm a single person in here, so I'm 32 00:01:37,160 --> 00:01:40,240 Speaker 5: like wow, like wow, the bill's going up so much. 33 00:01:40,520 --> 00:01:44,200 Speaker 4: Kevin is blind and he lives on disability payments. He 34 00:01:44,240 --> 00:01:47,200 Speaker 4: says his energy bills are now eighty percent higher than 35 00:01:47,200 --> 00:01:49,720 Speaker 4: they were just about three years ago. And there were 36 00:01:49,840 --> 00:01:52,920 Speaker 4: days when the utility asks customers like him to use 37 00:01:53,000 --> 00:01:56,800 Speaker 4: less power to prevent blackouts or encourages them to use 38 00:01:56,880 --> 00:01:58,080 Speaker 4: less power to save money. 39 00:01:58,200 --> 00:02:00,600 Speaker 5: They have days where they tell us, don't use any 40 00:02:00,640 --> 00:02:04,120 Speaker 5: electricity overrun air and dig but it's like ninety five degrees. 41 00:02:05,120 --> 00:02:06,400 Speaker 5: I'll done here. 42 00:02:07,400 --> 00:02:09,840 Speaker 4: It can be difficult to pinpoint the exact root of 43 00:02:09,919 --> 00:02:13,040 Speaker 4: higher electricity bills and strain on the electrical grid, so 44 00:02:13,160 --> 00:02:15,600 Speaker 4: Josh set out to put some numbers to the growing 45 00:02:15,639 --> 00:02:18,800 Speaker 4: pressure data centers are putting on local power supply, and 46 00:02:18,880 --> 00:02:21,360 Speaker 4: he found that at a time of rapid data center 47 00:02:21,400 --> 00:02:25,200 Speaker 4: construction and investment in AI technology that needs them, demand 48 00:02:25,360 --> 00:02:29,040 Speaker 4: is driving up wholesale energy costs, and those costs are 49 00:02:29,080 --> 00:02:30,760 Speaker 4: being passed on to consumers. 50 00:02:31,400 --> 00:02:34,919 Speaker 3: Josh spoke to some people that you know, started telling 51 00:02:35,000 --> 00:02:36,520 Speaker 3: him that their bills were really high. 52 00:02:36,840 --> 00:02:40,760 Speaker 4: That's Leonardo Nicoletti, a data visualization reporter at Bloomberg who 53 00:02:40,800 --> 00:02:42,800 Speaker 4: traveled with Josh and crunched the numbers. 54 00:02:43,120 --> 00:02:45,760 Speaker 3: When we were thinking about it, they seemed to be 55 00:02:45,840 --> 00:02:48,320 Speaker 3: living close to the hotspots of you know, AI data 56 00:02:48,360 --> 00:02:51,720 Speaker 3: centers in the United States. The main finding was that, 57 00:02:51,960 --> 00:02:56,120 Speaker 3: quite strikingly, if you are in an area that is 58 00:02:56,160 --> 00:03:00,440 Speaker 3: located close to data centers or data center activity, much 59 00:03:00,480 --> 00:03:04,119 Speaker 3: more likely to experience high price increases. 60 00:03:04,680 --> 00:03:08,840 Speaker 2: The massive increase in the demand for AI and the 61 00:03:08,880 --> 00:03:11,360 Speaker 2: speed with which the data centers get bigger and bigger, 62 00:03:11,440 --> 00:03:13,960 Speaker 2: I think it's been a huge surprise to all of us. 63 00:03:17,840 --> 00:03:19,720 Speaker 4: I'm David Gerrett, and this is the big take from 64 00:03:19,720 --> 00:03:23,440 Speaker 4: Bloomberg News today on the show, the staggering electricity needs 65 00:03:23,440 --> 00:03:27,200 Speaker 4: of AI data centers and how you yes, you could 66 00:03:27,280 --> 00:03:34,360 Speaker 4: end up putting the bill. AI needs a lot of 67 00:03:34,520 --> 00:03:38,480 Speaker 4: energy to work from summarized answers on Google to chat, GPT, 68 00:03:38,600 --> 00:03:42,760 Speaker 4: grocery lists. The rapidly growing technology requires massive amounts of 69 00:03:42,760 --> 00:03:46,600 Speaker 4: computing power, which requires a lot of actual power. 70 00:03:47,720 --> 00:03:50,600 Speaker 3: We actually quantified this in an earlier story that Josh 71 00:03:50,600 --> 00:03:53,520 Speaker 3: and I worked together. We compared the total amount of 72 00:03:53,520 --> 00:03:58,040 Speaker 3: electricity that data centers use yearly to how much electricity 73 00:03:58,280 --> 00:04:02,760 Speaker 3: individual countries use, and we found that, for example, countries 74 00:04:02,800 --> 00:04:07,280 Speaker 3: like Italy or Australia are using actually less electricity as 75 00:04:07,320 --> 00:04:11,000 Speaker 3: a total than data centers use globally. And these numbers 76 00:04:11,040 --> 00:04:13,600 Speaker 3: are actually higher now because we did this, you know, 77 00:04:13,680 --> 00:04:18,000 Speaker 3: one year ago, and electricity demand is growing exponentially from 78 00:04:18,080 --> 00:04:18,719 Speaker 3: data centers. 79 00:04:19,200 --> 00:04:21,560 Speaker 2: That numbers supposed to keep going up and up. So 80 00:04:21,839 --> 00:04:24,400 Speaker 2: Bloomberg and EF projects that it's supposed to be over 81 00:04:24,440 --> 00:04:28,840 Speaker 2: four percent by twenty thirty five of total global electricity consumption. 82 00:04:29,080 --> 00:04:30,640 Speaker 2: So if you took all the data centers at that 83 00:04:30,720 --> 00:04:33,159 Speaker 2: point and made them their own country, at that point, 84 00:04:33,200 --> 00:04:37,400 Speaker 2: they'd be the fourth biggest consumer of electricity after China, 85 00:04:37,440 --> 00:04:38,239 Speaker 2: the US, and India. 86 00:04:39,839 --> 00:04:42,240 Speaker 4: But what Josh and Lao wanted to understand was whether 87 00:04:42,279 --> 00:04:46,320 Speaker 4: the huge demand for electricity was showing up on customers' bills. 88 00:04:46,680 --> 00:04:49,240 Speaker 3: It took us like three months to find the data 89 00:04:49,320 --> 00:04:52,960 Speaker 3: because you know, it's actually really hard to get granular 90 00:04:53,320 --> 00:04:55,240 Speaker 3: data on power prices. 91 00:04:55,800 --> 00:04:58,839 Speaker 4: To get that more granular picture, Layo and another Bloomberg 92 00:04:58,920 --> 00:05:02,440 Speaker 4: data visualization report or named Demetrius Podcas found a company 93 00:05:02,480 --> 00:05:05,600 Speaker 4: called grid Status that collects real time data for more 94 00:05:05,600 --> 00:05:08,240 Speaker 4: than twenty five thousand nodes around the US. 95 00:05:08,800 --> 00:05:12,480 Speaker 3: A node is like the location on the grid, so 96 00:05:12,520 --> 00:05:17,240 Speaker 3: that's connected to transmission lines and it measures real time congestion, 97 00:05:17,480 --> 00:05:22,360 Speaker 3: real time fuel prices, and real time like supply costs, 98 00:05:22,560 --> 00:05:25,720 Speaker 3: things like that, right, and then the different factors that 99 00:05:25,839 --> 00:05:28,080 Speaker 3: are measured by that node then make up the price. 100 00:05:28,960 --> 00:05:31,400 Speaker 4: With that data as well as data on the locations 101 00:05:31,440 --> 00:05:34,560 Speaker 4: and capacity of data centers from a company called dc Byte, 102 00:05:34,839 --> 00:05:38,320 Speaker 4: the reporters measured the distance between those nodes and areas 103 00:05:38,360 --> 00:05:42,000 Speaker 4: where there's significant data center activity. They use that to 104 00:05:42,080 --> 00:05:44,599 Speaker 4: create a data set that showed the price changes recorded 105 00:05:44,640 --> 00:05:47,240 Speaker 4: in those areas over time since twenty twenty. 106 00:05:47,839 --> 00:05:52,320 Speaker 3: Essentially, the main finding is that seventy percent of nodes 107 00:05:52,360 --> 00:05:56,320 Speaker 3: that recorded price increases are located within fifty miles of 108 00:05:56,800 --> 00:05:59,000 Speaker 3: significant data center activity, and if. 109 00:05:58,920 --> 00:06:00,920 Speaker 2: You look at your power bill, you're going to see 110 00:06:01,120 --> 00:06:03,760 Speaker 2: in most places a distribution charge. That's what you're paying 111 00:06:03,800 --> 00:06:06,160 Speaker 2: for all of the infrastructure, all the wires. But you're 112 00:06:06,160 --> 00:06:08,240 Speaker 2: also going to see a supply charge, and the supply 113 00:06:08,360 --> 00:06:10,520 Speaker 2: charge is what you're paying for that wholesale power. So 114 00:06:10,560 --> 00:06:13,040 Speaker 2: what lao's research says is by pushing up the price 115 00:06:13,120 --> 00:06:16,360 Speaker 2: of wholesale power, data centers are putting that upward pressure 116 00:06:16,480 --> 00:06:17,400 Speaker 2: on your power bill. 117 00:06:17,720 --> 00:06:19,560 Speaker 4: So effectively, if I open up that bill, I'm not 118 00:06:19,560 --> 00:06:22,520 Speaker 4: going to see AI data centers as a line item, 119 00:06:22,920 --> 00:06:26,080 Speaker 4: but it's having that effect on what you're describing exactly. 120 00:06:27,200 --> 00:06:30,200 Speaker 4: Josh says there are two key mechanisms that drive up 121 00:06:30,200 --> 00:06:32,800 Speaker 4: electricity costs around data center hotspots. 122 00:06:33,040 --> 00:06:36,760 Speaker 2: The first, to really summarize supply and demand, they use 123 00:06:36,760 --> 00:06:39,599 Speaker 2: a lot of electricity, which can make electricity more expensive 124 00:06:40,080 --> 00:06:42,120 Speaker 2: for everybody, so that's one way that can push up 125 00:06:42,160 --> 00:06:44,160 Speaker 2: customer bills. The other ways that they require a lot 126 00:06:44,160 --> 00:06:47,640 Speaker 2: of new infrastructure, mostly transmission lines, but also new power plants, 127 00:06:47,800 --> 00:06:50,239 Speaker 2: and the costs of building that are spread out among 128 00:06:50,320 --> 00:06:53,000 Speaker 2: all customers, including regular people like you. 129 00:06:53,600 --> 00:06:56,680 Speaker 4: Monitoring Analytics is an outside watchdog that keeps an eye 130 00:06:56,680 --> 00:07:00,600 Speaker 4: on the largest US electric grid, PJM, is that data 131 00:07:00,600 --> 00:07:04,240 Speaker 4: center development raised costs for customers on pjm's grid, which 132 00:07:04,240 --> 00:07:06,839 Speaker 4: spans from Illinois to Washington, d C. By more than 133 00:07:06,920 --> 00:07:10,200 Speaker 4: nine point three billion dollars over a twelve month period 134 00:07:10,320 --> 00:07:13,560 Speaker 4: starting in June. Today, thirty three percent of all the 135 00:07:13,600 --> 00:07:17,400 Speaker 4: electricity used in Oregon is attributed to data centers. In Virginia, 136 00:07:17,720 --> 00:07:21,000 Speaker 4: it's thirty seven percent. That's according to data from dc 137 00:07:21,120 --> 00:07:25,040 Speaker 4: BYTE and the US Energy Information Agency. Communities in surrounding 138 00:07:25,040 --> 00:07:27,720 Speaker 4: areas are starting to do the math, and they're also 139 00:07:27,960 --> 00:07:30,160 Speaker 4: connecting higher bills to data centers. 140 00:07:30,400 --> 00:07:32,720 Speaker 6: You don't have to see those data centers to see 141 00:07:33,000 --> 00:07:34,320 Speaker 6: the higher electric bills. 142 00:07:34,520 --> 00:07:37,760 Speaker 4: That's David Lapp, a consumer advocate from the Maryland Office 143 00:07:37,760 --> 00:07:38,640 Speaker 4: of People's Council. 144 00:07:39,040 --> 00:07:41,720 Speaker 6: We don't have in Maryland too many data centers, so 145 00:07:41,840 --> 00:07:44,480 Speaker 6: the costs are being driven up by data centers that 146 00:07:44,520 --> 00:07:47,080 Speaker 6: are out of state largely, and many of them in 147 00:07:47,200 --> 00:07:48,120 Speaker 6: northern Virginia. 148 00:07:48,840 --> 00:07:52,040 Speaker 4: In February, the Baltimore City Council led a hearing on 149 00:07:52,080 --> 00:07:56,240 Speaker 4: the issue of rising Baltimore gas and electric bills. 150 00:07:56,600 --> 00:08:01,440 Speaker 2: The current state of these BGE bills and the skyrocketing 151 00:08:01,640 --> 00:08:05,120 Speaker 2: rates is simply not sustainable for our constituents. 152 00:08:05,480 --> 00:08:07,240 Speaker 4: And one of the people who spoke at the hearing 153 00:08:07,560 --> 00:08:11,360 Speaker 4: was Kevin Stanley, the man Josh and Laoe interviewed in Baltimore. 154 00:08:11,480 --> 00:08:15,560 Speaker 5: Are you yes, good evening everyone. My name is Kevin Stanley. 155 00:08:16,000 --> 00:08:18,280 Speaker 4: Kevin didn't know what was causing his bills to spike. 156 00:08:18,560 --> 00:08:20,360 Speaker 4: He just knew it was a problem. 157 00:08:20,520 --> 00:08:23,320 Speaker 5: And there's no reason that Baltimore is should have the 158 00:08:23,400 --> 00:08:26,440 Speaker 5: excess of gas electric race. 159 00:08:26,360 --> 00:08:26,800 Speaker 2: That they had. 160 00:08:27,320 --> 00:08:31,280 Speaker 3: When Josh asked Kevin Stanley what he thought about data centers, 161 00:08:31,880 --> 00:08:32,600 Speaker 3: his reply. 162 00:08:32,720 --> 00:08:34,400 Speaker 5: Was so they could say, oh, this is going to 163 00:08:34,480 --> 00:08:36,319 Speaker 5: help with AI. But how's that going to help me? 164 00:08:37,240 --> 00:08:38,920 Speaker 5: How's that going to help me pay up bill? 165 00:08:41,840 --> 00:08:44,560 Speaker 4: So how do utilities decide who fronts the cost? And 166 00:08:44,600 --> 00:08:55,200 Speaker 4: how are those communities responding? That's after the break A 167 00:08:55,240 --> 00:08:58,160 Speaker 4: team of Bloomberg reporters dug into the rising costs of 168 00:08:58,200 --> 00:09:01,320 Speaker 4: electricity to see if the rapid construction of data centers 169 00:09:01,320 --> 00:09:04,040 Speaker 4: in some areas could be to blame, and they found 170 00:09:04,080 --> 00:09:07,720 Speaker 4: a link. Places with heightened data center activity were more 171 00:09:07,840 --> 00:09:11,800 Speaker 4: likely to see higher wholesale energy costs, meaning the presence 172 00:09:11,800 --> 00:09:15,480 Speaker 4: of data centers was raising power prices and eventually pushing 173 00:09:15,559 --> 00:09:19,280 Speaker 4: up customer bills. I asked Bloomberg reporter Josh Saul about 174 00:09:19,320 --> 00:09:22,320 Speaker 4: why that's happening when people usually expect to pay for 175 00:09:22,360 --> 00:09:26,240 Speaker 4: their own utility costs. I think about how I study 176 00:09:26,280 --> 00:09:29,360 Speaker 4: my electricity build with the extent that I do, and 177 00:09:29,400 --> 00:09:31,720 Speaker 4: maybe in the summer, I'm using my air conditioner more, 178 00:09:32,520 --> 00:09:34,280 Speaker 4: maybe in the winter I'm using my heat more. But 179 00:09:34,400 --> 00:09:36,760 Speaker 4: these are all kind of selfish things. I'm determining when 180 00:09:36,800 --> 00:09:39,480 Speaker 4: the bill goes up or down, why should I be 181 00:09:39,559 --> 00:09:43,600 Speaker 4: paying for the expenses of associated with these data centers. 182 00:09:43,760 --> 00:09:45,600 Speaker 2: Well, I think there's two arguments that people make. I mean, 183 00:09:45,640 --> 00:09:48,360 Speaker 2: the first is that AI, whether you like it or not, 184 00:09:48,600 --> 00:09:51,320 Speaker 2: is becoming an increasingly large part of our world. So 185 00:09:51,520 --> 00:09:54,920 Speaker 2: from national security to the argument that we don't want 186 00:09:55,000 --> 00:09:57,360 Speaker 2: other countries to be way better at AI than we are, 187 00:09:57,720 --> 00:10:00,319 Speaker 2: to the fact that now our Google searches and even 188 00:10:00,360 --> 00:10:04,520 Speaker 2: our texts to our friends are guided by or instructed 189 00:10:04,520 --> 00:10:07,000 Speaker 2: by AI. So there's one argument there that we're all 190 00:10:07,080 --> 00:10:09,880 Speaker 2: using AI. And there's also an old utility law. The 191 00:10:09,880 --> 00:10:12,120 Speaker 2: way we do this is that the costs are shared 192 00:10:12,160 --> 00:10:15,160 Speaker 2: among everyone. So if you build a new house or 193 00:10:15,200 --> 00:10:17,120 Speaker 2: hook up to the grid. As a new business, you 194 00:10:17,160 --> 00:10:20,360 Speaker 2: don't pay for all of those costs to hook you up. 195 00:10:20,360 --> 00:10:24,360 Speaker 2: It's socialized among everyone on the grid. So utilities are 196 00:10:24,679 --> 00:10:29,079 Speaker 2: following that general principle. Now with these large new customers 197 00:10:29,160 --> 00:10:30,280 Speaker 2: data centers. 198 00:10:30,880 --> 00:10:33,560 Speaker 4: It's tech companies that are deciding when and where to 199 00:10:33,600 --> 00:10:37,400 Speaker 4: build a new data center. Amazon, Microsoft, and Alphabet's Google 200 00:10:37,600 --> 00:10:41,480 Speaker 4: are the three biggest US cloud providers. In twenty twenty four, 201 00:10:41,559 --> 00:10:44,600 Speaker 4: those companies spent more than two hundred billion dollars on 202 00:10:44,720 --> 00:10:48,120 Speaker 4: capital expenditures, most of it to build new data centers. 203 00:10:48,559 --> 00:10:52,240 Speaker 4: What's the relationship like between these tech companies that need 204 00:10:52,280 --> 00:10:55,080 Speaker 4: all of this power and the utilities themselves. Are they 205 00:10:55,080 --> 00:10:57,760 Speaker 4: paying their fair share? Are they paying a disproportionate amount? 206 00:10:57,960 --> 00:10:58,839 Speaker 4: How's that shaking out? 207 00:10:58,960 --> 00:11:01,320 Speaker 2: Well, utilities are really ECPs because it's a huge new 208 00:11:01,360 --> 00:11:03,840 Speaker 2: customer base for them, and this is coming after decades 209 00:11:03,880 --> 00:11:07,720 Speaker 2: of very flat growth. At the same time, there's some conflicts. 210 00:11:07,920 --> 00:11:10,880 Speaker 2: Tech is known for moving fast and breaking things. That's 211 00:11:10,880 --> 00:11:13,720 Speaker 2: their philosophy. Utilities are known for moving very slow and 212 00:11:13,720 --> 00:11:16,079 Speaker 2: making sure nothing ever breaks, you know, making sure the 213 00:11:16,160 --> 00:11:18,839 Speaker 2: lights stay on. So they're working together here. But one 214 00:11:18,840 --> 00:11:22,160 Speaker 2: thing that we're seeing is that utilities are actively figuring 215 00:11:22,200 --> 00:11:25,079 Speaker 2: out how to make sure that big tech is paying 216 00:11:25,080 --> 00:11:27,880 Speaker 2: its share for these data centers. They're doing things like 217 00:11:27,880 --> 00:11:31,920 Speaker 2: creating new customer classes for data centers to pay differently 218 00:11:31,960 --> 00:11:36,040 Speaker 2: than say, residential or small businesses. They're creating requirements like 219 00:11:36,120 --> 00:11:39,679 Speaker 2: data centers have to pay for certain costs, pay for 220 00:11:39,760 --> 00:11:42,000 Speaker 2: a minimum amount of power over fifteen years, even if 221 00:11:42,000 --> 00:11:45,240 Speaker 2: they use less, putting down collateral, different things like that 222 00:11:45,480 --> 00:11:47,800 Speaker 2: to protect their existing customers. 223 00:11:48,320 --> 00:11:51,240 Speaker 4: Josh Leo and their team reached out to Amazon, Microsoft, 224 00:11:51,320 --> 00:11:55,079 Speaker 4: and Alphabet for comment. An Amazon spokesperson said the company 225 00:11:55,160 --> 00:11:58,280 Speaker 4: works closely with utilities and grid operators to plan for 226 00:11:58,320 --> 00:12:02,480 Speaker 4: future growth. Soft's vice president of energy told them, quote, 227 00:12:02,520 --> 00:12:05,200 Speaker 4: it's literally our responsibility to make sure that when we 228 00:12:05,240 --> 00:12:07,800 Speaker 4: come to a community, when we get connected to a grid, 229 00:12:08,160 --> 00:12:11,000 Speaker 4: that the cost of the infrastructure that's being dedicated to us, 230 00:12:11,120 --> 00:12:14,800 Speaker 4: that those costs of service get allocated to us. Google 231 00:12:14,840 --> 00:12:17,800 Speaker 4: said it's been working to use less electricity even as 232 00:12:17,840 --> 00:12:20,719 Speaker 4: it expands data centers, and that it supports paying its 233 00:12:20,760 --> 00:12:24,640 Speaker 4: fair share. What did you hear from utilities about this phenomenon? 234 00:12:24,679 --> 00:12:26,920 Speaker 4: What do they say about why bills are going up 235 00:12:26,920 --> 00:12:28,160 Speaker 4: and what they intend to do? About it. 236 00:12:28,440 --> 00:12:31,800 Speaker 2: Utilities mostly point to the idea that data centers are 237 00:12:31,800 --> 00:12:34,679 Speaker 2: paying their fair share. So if a data center requires 238 00:12:34,720 --> 00:12:38,400 Speaker 2: a specific substation for its operations, the tech company will 239 00:12:38,400 --> 00:12:40,720 Speaker 2: pay for that, and no point to the reforms that 240 00:12:40,720 --> 00:12:44,240 Speaker 2: they're making in their own rate structure to charge these 241 00:12:44,320 --> 00:12:47,360 Speaker 2: large load customers differently. All of that's true, and all 242 00:12:47,480 --> 00:12:50,080 Speaker 2: that's good, but it doesn't change the fact that a 243 00:12:50,120 --> 00:12:53,439 Speaker 2: lot of costs are still being loaded onto regular customers. 244 00:12:53,800 --> 00:12:57,200 Speaker 4: Dominion Energy, the utility company that serves Northern Virginia's data center, 245 00:12:57,240 --> 00:13:00,679 Speaker 4: Alley told the reporters quote, we believe data centers should 246 00:13:00,679 --> 00:13:03,079 Speaker 4: pay for the full cost of their power. That's how 247 00:13:03,080 --> 00:13:05,520 Speaker 4: we design our rates and it's the standard our regulator 248 00:13:05,600 --> 00:13:08,839 Speaker 4: uses in reviewing them. The CEO of Exelon, which runs 249 00:13:08,840 --> 00:13:12,080 Speaker 4: Baltimore's utility, said the company is pushing for long term 250 00:13:12,080 --> 00:13:15,040 Speaker 4: solutions that are fair and bring peace of mind to customers. 251 00:13:15,600 --> 00:13:18,160 Speaker 2: Talking to one of the utilities, we asked about how 252 00:13:18,200 --> 00:13:22,040 Speaker 2: they pass these costs on to customers, and their experts said, 253 00:13:22,160 --> 00:13:24,400 Speaker 2: we sign long term contracts, so if power spikes up, 254 00:13:24,440 --> 00:13:26,960 Speaker 2: that's for the power sellers to deal with. Our customer 255 00:13:27,040 --> 00:13:30,240 Speaker 2: doesn't have to deal with that. However, the next time 256 00:13:30,280 --> 00:13:32,000 Speaker 2: we sign a contract, if there's been a lot of 257 00:13:32,080 --> 00:13:35,040 Speaker 2: variability in cost in power prices, and if the power 258 00:13:35,080 --> 00:13:38,480 Speaker 2: prices have been overall just higher, which happens because of 259 00:13:38,559 --> 00:13:41,800 Speaker 2: data centers, then the next contract that we sign will 260 00:13:41,800 --> 00:13:44,280 Speaker 2: be for higher prices, and then yes, that will get 261 00:13:44,320 --> 00:13:46,079 Speaker 2: passed on directly to our customers. 262 00:13:46,320 --> 00:13:49,240 Speaker 4: Now some communities are starting to have conversations about what 263 00:13:49,280 --> 00:13:52,080 Speaker 4: it really means to have these big AI data centers 264 00:13:52,080 --> 00:13:55,160 Speaker 4: in their backyard. Are we seeing communities, are we seeing 265 00:13:55,200 --> 00:13:57,080 Speaker 4: states pushback on this in any way? 266 00:13:57,520 --> 00:14:01,360 Speaker 2: Yeah, I'd say there's more dramatic conversation around the power 267 00:14:01,400 --> 00:14:05,080 Speaker 2: grid than anytime in my years covering it. There was 268 00:14:05,240 --> 00:14:09,760 Speaker 2: just a summit in Philadelphia where the Governor of Pennsylvania, Joshapiro, 269 00:14:10,160 --> 00:14:13,080 Speaker 2: talked about affordability. He talked about these rising bills. 270 00:14:13,320 --> 00:14:13,960 Speaker 4: As governor of. 271 00:14:13,960 --> 00:14:17,439 Speaker 6: This great com privilege to represent thirteen. 272 00:14:17,080 --> 00:14:18,760 Speaker 3: Million Pennsylvania, and I. 273 00:14:18,720 --> 00:14:22,760 Speaker 4: Can tell you they can't afford never in price increases 274 00:14:23,080 --> 00:14:24,920 Speaker 4: because of pgm's policies. 275 00:14:25,480 --> 00:14:28,800 Speaker 2: That's something we're seeing from politicians all around. Sometimes there's 276 00:14:28,800 --> 00:14:32,360 Speaker 2: a little bit of a personality split where local government's 277 00:14:32,400 --> 00:14:36,280 Speaker 2: excited for the tax income they want companies moving in. 278 00:14:36,360 --> 00:14:39,000 Speaker 2: They want business activity, They want to fund their local 279 00:14:39,000 --> 00:14:39,760 Speaker 2: school district. 280 00:14:40,120 --> 00:14:43,080 Speaker 4: In the city of Fayetteville, Georgia, for example, the taxes 281 00:14:43,080 --> 00:14:45,760 Speaker 4: paid by a new data center developer were expected to 282 00:14:45,760 --> 00:14:48,520 Speaker 4: bring in over one billion dollars in state and local 283 00:14:48,560 --> 00:14:51,600 Speaker 4: taxes in the next fifteen years. Taxes going to the 284 00:14:51,600 --> 00:14:54,680 Speaker 4: county Board of Education last year covered the equivalent of 285 00:14:54,720 --> 00:14:56,720 Speaker 4: some half a dozen teachers salaries. 286 00:14:57,200 --> 00:15:00,600 Speaker 2: Then there's a split between that view and and people 287 00:15:00,640 --> 00:15:03,600 Speaker 2: who live around there who don't want big new transmission 288 00:15:03,640 --> 00:15:07,680 Speaker 2: lines going through their yards. They don't want big three 289 00:15:07,720 --> 00:15:10,680 Speaker 2: Walmart stacked on top of each other size data centers. 290 00:15:10,800 --> 00:15:13,120 Speaker 2: They don't want all of that around them. So, yeah, 291 00:15:13,160 --> 00:15:15,000 Speaker 2: there can be a lot of conflict in these communities. 292 00:15:16,400 --> 00:15:19,400 Speaker 4: When you look at projections for the growth of AI 293 00:15:19,560 --> 00:15:23,040 Speaker 4: data centers, what do they illustrate and what is that 294 00:15:23,160 --> 00:15:25,760 Speaker 4: likely to mean for the cost of electricity in this country? 295 00:15:26,240 --> 00:15:28,280 Speaker 2: Well, I think it's something that needs to get figured out. 296 00:15:28,520 --> 00:15:30,920 Speaker 2: Utilities are working on it. They can't have their customers 297 00:15:30,960 --> 00:15:33,800 Speaker 2: and their regulators and politicians angry at them. Big tech 298 00:15:33,880 --> 00:15:37,840 Speaker 2: similarly very attuned to what people think of them. That's 299 00:15:37,840 --> 00:15:39,920 Speaker 2: why there's a lot of conversations and a lot of 300 00:15:39,920 --> 00:15:42,440 Speaker 2: push right now to figure out how to charge data 301 00:15:42,480 --> 00:15:45,000 Speaker 2: centers for both the power that they're using and the 302 00:15:45,040 --> 00:15:47,920 Speaker 2: infrastructure that has to be built to supply their data centers. 303 00:15:48,560 --> 00:15:52,040 Speaker 4: In the meantime, Kevin Stanley is doing his best to 304 00:15:52,080 --> 00:15:53,680 Speaker 4: deal with higher electricity bills. 305 00:15:53,800 --> 00:15:56,720 Speaker 5: You know, they talk about upgrading and greed and this 306 00:15:56,760 --> 00:15:59,320 Speaker 5: and that, but we don't see it. Especially when you're 307 00:15:59,320 --> 00:16:05,240 Speaker 5: dealing with low income people or working class people. They 308 00:16:05,320 --> 00:16:07,600 Speaker 5: have to feel something. It has to be more tangible. 309 00:16:08,240 --> 00:16:10,320 Speaker 5: You would tell people what is a data center? They said, well, 310 00:16:10,360 --> 00:16:11,920 Speaker 5: what is it? We're just surviving. 311 00:16:18,600 --> 00:16:21,040 Speaker 4: This is the Big Take from Bloomberg News. I'm David Gerra. 312 00:16:21,400 --> 00:16:23,880 Speaker 4: To get more from the Big Take and unlimited access 313 00:16:23,960 --> 00:16:27,400 Speaker 4: to all of Bloomberg dot com, subscribe today at Bloomberg 314 00:16:27,440 --> 00:16:31,360 Speaker 4: dot com Slash Podcast offer. Thanks for listening. We'll be 315 00:16:31,400 --> 00:16:32,000 Speaker 4: back tomorrow.