Data Centers and Brad Paisley

 This episode centers on a discussion about data centers: what they are, why they exist, and why some people object to them. The hosts contrast older ideas of data storage with modern AI-focused facilities, then use Brad Paisley’s opposition to a proposed Nashville Zoo data center as a concrete example of concerns about water use, electricity demand, noise, heat, and visual impact.

Main points

• They first revisit the earlier “middleman scam” topic and note that no new developments have come from it.
• The bulk of the episode explains that a data center is not just a storage room, but a facility for computing, networking, backup, and cloud services.
• The hosts argue that many modern “data centers” are really processing centers for AI workloads, which consume substantial power and require cooling.
• They discuss how these facilities can create local environmental and community concerns, especially around resource use and appearance.
• The episode ends with a reflection on AI’s limits, the flexibility of pronunciation in “data/data,” and a closing reminder that talking leads to understanding.

Key Words: Episode 111, I Hate Talking, Data centers, Cloud computing, AI workloads, Artificial Intelligence infrastructure, Data storage vs. processing centers, Networking and backup services, Brad Paisley, Nashville Zoo data center, "Middleman scam", Tech pronunciation (Data vs. Data), Environmental & Community Impact, Data center energy consumption, Water usage and cooling, Noise pollution, Thermal emissions (Heat waste), Grid demand and electricity, Local community opposition, Visual impact and zoning, Meta & Themes, Limits of AI, Tech infrastructure awareness, Communication and understanding

Transcript

I Hate Talking: 0:29

Welcome to episode 111 of I Hate Talking. Hi everyone. So we have not received any particular comments or emails or text messages in response to the questions from our last episode on if you found the middleman scam interesting or what you would do in our shoes. So you can always go back and listen to that if you have not heard that episode and comment accordingly, but for our experience, nothing more has come of it. Yeah, as of right now, we're, it's all done, so. So hopefully it remains that way. Hopefully there are no more. Legitimate or illegitimate services being performed at our property. Other than the ones that we ourselves schedule. Right. So there you have it. And then like I mentioned, you can always like, share, subscribe. Maybe there is a friend out there that is not aware of what a middleman scam is. Maybe this is your first episode and you don't know either. It's just episode 110, 1 previous to this. Yeah, maybe it'd be helpful. No one wants to fall for it, cause We were lucky enough to kind of catch it, the circumstances, right? That we at least hope, we still don't know, but we hope that we haven't fallen for the scam. But we easily could have. That's true. Maybe we have. We still don't know. -- We don't know -- 100%. So moving on to our word or phrase of this episode, and that is the phrase data center. I, um, asked you what this was earlier tonight. All right. And it answered us. That's true. He came with a textbook answer that it is a secure location that contains a variety of computer servers that are used for a variety of purposes, data processing, data storage, so on and so forth. Yeah, that's what he said. So, I think that's uh probably an apt description of data center, I think. In the olden days, data center was Truly data, like it was storage facilities, -- so like -- like files, right, Manila envelopes. No, still computer related, but it was data hosting, so. Things to Run websites, so like hosting materials, pictures, videos. The HTML code that is running websites. Because there wasn't a cloud back then. Or you could host a website from your own local machine, but. The internet was basically networked computers that were used for storage and communication. OK, so a lot of these words, I don't know what they mean technically. I, I can use them in a sentence. I really don't know what it means. Right, so cloud computing is sort of A, I guess, recent jargon, but The internet really has always been cloud. It's interconnected devices, some of which may host or some of which may process. And why do you need a physical area to store it? Well, the data has to be stored somewhere in all the zeros and ones, uh, in the actual cloud, like where the rain comes from. -- I -- don't know, it's just up there above us in our heads. No, it's in some physical server somewhere. So there's one mega computer that has all the knowledge in the world, and if we turn that off, We lose everything. No. Now, there have been instances where like certain data centers have gone offline, like there was one recently, well, maybe not recently, but within the past 1 to 2 years. With Amazon Web Services where some of their major data centers went offline and their remaining data centers. Helped with the load, but we're not able to sustain all the different traffic and hosting services that It was being asked of, so there was a lot of slowdown and crashes and Things that were not working OK, huh. My point being that in the past, Data center as a phrase I think really did refer to data, so storage of media. And things of that nature. But I think what is in the news more recently. is they call it data center, but it's really probably more like processing center. Processing, but it's still not physical items, so I think that's where I'm having a hard time, why we need a physical area for virtual things. Well, virtual things exist in. Reality and reality is based in Hardware of some kind. I mean, when you really think about it. Basically, a computer is a rock that we have introduced electricity to. So basically a computer is just A lightning charge rock. OK. We had to put the rock somewhere. The rock's in the room right now. We're using the rock. Yeah, and then we're gonna take this audio file, which is stored locally on this physical computer right here that is taking up space on our table, and we're going to upload that to our podcast hosting platform. And that is going to go on some physical server somewhere. And we're going to upload it to AWS, Amazon Web Services, in fact, to do some. Transcripting and things of that nature. So that's physically located on some server somewhere, probably more than once since they have backups and things like that. OK, and then we can delete it from our computer. To make storage space. But that's why it's never fully deleted because it's over there on their computer. Correct, yeah, we have the backups of the audio files on AWS at this moment. And that physically exists somewhere. -- In a data center and -- then if AWS goes down, we lost it. Well, they probably have backups that they could restore, -- but if the backup -- goes down, yeah, -- it's possible that it -- would be lost forever centers because they're doing the backup, so the backups of the backups. No, see, that's why I think it's a misnomer to call it a data center. Cause I don't think these new facilities that are coming online are being used for. Data per se in terms of the old terminology where data was going to be more storage. OK, so then what is it then? AI. OK, they're processing centers. What do you mean? Like little robots in there? -- Typing away my Gemini -- questions. Yes. They, well, that question has to be responded to, and it takes physical assets in terms of some sort of CPU or GPU to do the calculations, look at its model and training data, come back with a response. That takes some sort of physical computing device. By a person. No. Oh. But somebody had to code that. Yes. That's the model. -- The -- training data in the data centers coding. No, most of these data centers are dark data centers. There's no people. They don't even have lights unless there's like a maintenance issue that they turn on the lights. So, but it still requires a lot of power and water. Correct. Why? Well, when you run a lightning through a rock, it tends to get hot, so you have to cool said rock to be able to perform adequately and not overheat, and that requires water and In order to power the rock, you need electricity, so it takes a lot of electricity, which also consumes a lot of natural resources and things of that nature. Is this where they're mining Bitcoins? I have not heard that as an official reason for building quote unquote data centers, but I suppose they could. You know how I heard about, I mean, I've heard data centers once in a while thrown around, and I've seen commercials for them. Do you know where I questioned it today? Well, I think it's been in the news quite a bit in a variety of different Locales and residential areas and things of that nature, but the one that has made national news as of late is the zoo. Yeah, that's how I noticed it. The singer, the musician, Brad Paisley. has like started this big campaign and a press conference and going all out because he lives in Nashville, and I guess they want to make data center next to the Nashville Zoo. And he is gung ho against it. So that's what struck my attention. Like, why does Brad Paisley hate data center so much? Right, I think as of late, a lot of the focus on The data centers and trying to not allow them to be built by the local populace and things of that nature is that they are rejecting said data centers because of the drain on natural resources. Namely water, electricity. As well as some of the unsightliness of data centers. They're very large, not very beautiful structures in any way. Well, I suppose they could be, but they're going for cost effectiveness. I don't think they're. Making it look like a tourist attraction because it's not made to be a tourist attraction. It's not like people are going to come visit. It's strictly for data and the storage and processing of that data. Yeah, -- so -- that was exactly the article I read about Brad Paisley, how he was saying they don't have enough water, they don't have enough electricity, and it's too ugly. Those are his like complaints I heard. Maybe he had others and I missed them, but that was my takeaway. So yeah, I mean, why, I mean, I'm not saying he has to spend his own money, but I'm sure he's financially could. If he cared that much, why don't they make it look like part of the zoo, just blend in. I know that won't help with the water and electricity part. But so it's not like sending out like. Radioactive stuff that's gonna hurt the animals, is it? No, I've not heard that. I have heard that the outcome of data centers could be, in addition to the resources used, noise pollution. So they does have the, right, yes, potentially noisy with the fans or pumps that are running in order to cool the data centers and Even probably the processing units themselves may generate quite a bit of noise. That it does impact noise within the local area and then also because of the massive amount of computing power that's being used, it is generating a lot of heat. So I've not seen this widely and this perhaps is maybe a bit of a theory, but I have seen reports that. When these data centers are stood up, that there is a localized increase in temperature around the data centers because they are producing so much heat. Hm, it's kind of like when you make a parking lot, the temperature in that area goes up because of the concrete versus the grass. Yeah, I suppose a similar concept, but the fact that it's actually generating heat because it's doing all the computations and things of that nature. So, I mean, I have seen memes where like people instead of buying a furnace, will just buy a few Bitcoin miners and run that, and that produces enough electricity and heat to warm their Abode for the winter. Well, yeah, I believe it in a way, because I had a computer, a desktop computer in college, and I remember that thing would get so hot sometimes, like blowing out such hot air. I didn't know if it was safe or not, so I would turn it off. But if I like ran it all night long, it would be hot. That's true. And then I have seen some plans. I don't think anybody has actually done this, but like to build data centers either near the ocean or even in the ocean, so that then therefore it could use the seawater as a cooling mechanism. But I think that would be dangerous because of tsunamis and earthquakes or leaks. Basically, if the seawater gets it in somewhere that it's not supposed to be, that basically your millions and millions of dollars of Computer hardware will be scrap. But that does make sense to use water that we can't drink. I mean, I know you can, but it's like a whole process, right, that you have to filter, yada yada, right, desalination plants and things like that. It sounds like that would be a good use of natural resources. Yeah, I don't know why they wouldn't build them on the coast, and, -- and -- people don't need to be there, right? Only occasionally checking on them. So it's not like the mining guys when they're out in the middle of the ocean for months at a time away from family, you know what I'm talking about, the oil, or you're not miners, are they oilers? -- What -- are they? rigs? Yeah, there you go. Maybe you can start a new company and call it. Oil rig data centers and then place the data centers on an oil rig and use seawater to cool it and then it won't even be on the ocean. You could even do it in international waters and then not be subject to any government laws. There you go. We can store everyone's secrets, huh? Yeah, I still really don't think that these are being built for storage though. So I don't get then what are they? I guess I don't understand AI. Well, I guess nobody does when it comes down to it. So it's just a blank computer waiting for a question. That's not a blank computer. It has an algorithm that's been coded and natural language models that have used training data in order to come up with a variety of responses that AI gives based on all the knowledge in the world that is in digital format. So that's why you could ask a question about 1984 to AI, and it knows something about that because it has 1984. Either in its training data or as part of its resources that it can draw from. So Is it kind of like picturing, say, a human, I know it's not a human, but think of like a human that has a library behind him. And then I'm like, hey person, answer this question. And then they're like, wait a minute, and then they go back to all of their stuff and like they categorize like look it up and then run to the Book and look at it and come back and tell me the answer. Maybe, does a person like logically think that this is the answer? No, they're just telling me the fact that they read in the book. But how do they determine what facts to give you? I don't know, whatever was in the book that they had. So they just randomly pulled a book -- and hoped that the fact that not -- randomly they have to look in their little like Dewey Decimal system. Their little cards. You know what I'm talking about? You picture like the old libraries where you pull out. Uh, I can't even think of what they're called. What are they? -- You pull out -- the like index card, -- index -- cards, and you look for that and it gives you a number, then you take it over to the books and find that number. I suppose so. There's still some sort of logical processing that the human is doing in order to use that. I guess it's sort of similar to how an algorithm would be programmed in order to know that this is the subject matter. This is the index that needs to be. Reviewed and then. Serve up that information, but AI is typically not deterministic, so. When you're using that example, that sounds more deterministic because a person is going to determine something as what will be the correct result, or at least the correct resource to use, whereas AI is just, for all intents and purposes, guessing. It's just really, really good at guessing. -- AI is -- right, because it's not deterministic. Yeah, well, just today I was using AI because I was asking, we're taking family photos soon, and I was asking it a question about the color scheme to make sure, you know, everyone's clothes aren't matchy matchy, but that they coordinate well. I was asking, and then they sent me something completely off the wall different. And I was like, wrong colors, and then they're like, oh, my mistake, and then like sent it again, and it was still wrong. So then I was like, wrong colors, and they're like, I miss my mistake, that must be frustrating. And then they gave me the correct thing. That's why AI is not very good at certain types of math where, again, like a deterministic result is needed. It's not like preprogrammed or like already pre-populated in terms of its like solution book or very good at verbal things like with counting the number of R's in strawberry. It just doesn't know how to do those things because it's not a deterministic thinking machine. It's just using its preprogrammed logic to attempt to get the right information. So we could convince an AI that there's 5 R's in strawberry. No, it will. In the olden days, they probably have fixed this now because like people always did this, and now they just coded it. So like, anytime somebody asks about a strawberry and the number of letters, this is the answer, but somebody, a person had to do that. OK. Because they would often report that there's only 2 R's in strawberry. Yeah. You've seen this? No, I'm just trying to count how many R's there are in strawberry. So, you're coming up with a deterministic solution because you're thinking logically and coming up with a determined solution versus AI just looking at what the question is. Associating words and letters to what it has learned in the large language model and producing a result. So why don't you test that and we'll see if they fixed it. So I asked Gemini, my question was, how many R letters are there in strawberry? And the response was, there are 3 letter R's in the word strawberry. Here is why it is famously tricky, because when you say strawberry out loud, your tongue blends the letters together, making it sound like there are only 2 R's, one in straw and one in berry. But when you look at the spelling of berry. That actually requires a double R. There you go. So somebody has fixed it. Let's try chat GBT. I think that was the one that became famous because of this particular thing. So Chat GBT responded, there are three letter R's in the word strawberry, spelling it out, S T R A W B E R R Y, letter R appears three times. And then we can ask, did it used to be 2? Let's find out. No, the strawberry word has not generally been spelled with only two R's. I go, so they're like gaslighting us, pretending like it was never. Their mistake. Yes. I asked Gemini if it used to be 2, I like your question, and I said no, it's actually always had 3 R's going way back to its earliest roots from thousands of years ago. Interesting. Now, if you ask chat GPT if it thought that it used to have two R's. It basically blames previous language models, so it says other language models have sometimes given the wrong answer to questions like this in the past. Earlier AI models often struggled with tasks that required careful counting of letters and words because they processed Texas patterns rather than literally inspecting each character the way a simple counting program would. So they admit that they were wrong, but they won't straight out say it. Well, it blames its parents essentially. It always goes back to that. It's always their fault. So, Yeah, I don't know. I guess we can ask what the difference is between a data center and a processing center, or if a data center is a processing center, see what it says. There you go. So I asked, is a data center just used for storing data? No, a data center does store data, but that's only one of its functions. A data center is a facility that houses computers, networking equipment, and storage systems that provide digital services. So it can include data storage, computing, networking, cloud services, backup, and disaster recovery. Like disaster recovery like. If we lose our internet. Yeah, like if, like if, uh, other data center gets hit by flooding or something, it will have a backup so that it can recover from. The disaster in another part of the world. Oh, OK, I was thinking like Red Cross, like they keep their emergency supplies there. Yep. So it does say at the very end here, in some AI focused facilities, computation is actually the dominant activity. See, and that's why I think that it's a bit of a misnomer to call some of these new quote unquote data centers, data centers because the primary purpose for which they are being constructed is computation. And computation is. Like doing some sort of compute function, doing some sort of generative response or calculation. AI, yes, AI. OK. How long have data centers been around then? Well, I guess you have the. Library of Alexandria. So I was there Library of Congress. -- I've been -- to the Library of Alexandria. -- There -- you go. I mean, those were data centers. There's a whole bunch of data there in the form of books, right? So I guess the library could be, but like in the way that we're using it nowadays. So the first use of the exact phrase data center was from The 1950s and 1960s to describe facilities that housed large mainframe computers requiring dedicated environments for their operation. There were computers in the 1950s. Yeah, they went to the moon. Yeah, -- but she -- had to, it was that lady that was in labor, and she had to like read all her notes in labor. Yeah, they had. Some computing systems back then and storage. It's crazy. That just seems like a long time ago. I guess not. Well, it was a long time ago. It's 2026 now. I'm sure a data center from the 1950s looks far different than a data center today, right. Now tell me, is it data or data? Is it kind of like the same thing where it's like caramel or caramel? It just depends on if you're fancy. I suppose so. I would say both could be correct. I would say data would probably be more. British and data would be more Americanism, perhaps. Is that true? Did you research this? No, -- I -- didn't. I was just curious because I do think I feel fancier when I say data. So the American Heritage Dictionary says that both are correct, but it has data as the primary pronunciation. OK. I know when I asked that question earlier in our family, about what is a data center. They said the kids were like bickering a little bit about the proper pronunciation. One called it data and the other one's like correcting them on data or vice versa, I don't remember. I was like, both, it's the same word, guys, you can say it either way. I wasn't quite confident. I was pretty confident in that, but I didn't know if that was true. But I wanted them to stop bickering, so I said they're both right. It can be data or data. Yep, it does seem that Merriam-Webster's, other British dictionaries all list All the different pronunciations, but Data seems to be the most preferred. Even in Britain? Yes, OK. Interesting. I don't know. I'll have to pay attention to how I pronounce it. I think I probably just reflect the same pronunciation of the people that I'm speaking with and sort of just mirror their pronunciation. we'll find out when we listen to this episode how many times we say data versus data. I guess we could ask AI to count how many times. There you go. Hopefully, you, our listener have appreciated the data that we have presented in this episode about data centers and their related impact to the natural world as we try to infuse lightning into rocks to make it answer questions for us. And good luck to Brad Paisley. Yeah, I am all supportive of like development of capabilities such as computing and things like that, but there is a balance there, especially if said activities may be used for nefarious purposes. Right, yeah. So from your friends at I Hate Talking, until next time, remember, it is only through talking that we begin the journey to understanding.