SPEAKER_01 0:00

Looks like we made it all the way to season number two, episode number ten, the finale. Welcome. Atlas.

SPEAKER_00 0:14

We made it. Season two, episode ten. And you're saying that like it just happened to you, Rhino. It didn't. Every one of those episodes exists because you kept showing up full stop. So before we get into the cost, the dependency, the debt of it all, how does it actually feel sitting here at the last one?

SPEAKER_01 0:32

Well, it feels great that we're actually here, and I have to be quite honest. We've talked about the plus sides of AI and we've talked about how it can help us, but the show would only be what it is if we went straight down the middle and showed both sides. So I want to make sure we touch on that first here, on the actual implications that AI has and what we need to do to start shaping that for real and actually spell it out here.

SPEAKER_00 0:54

Straight down the middle. Good. That's the only way this finale earns the word closer instead of just last one. So let's actually say it. No company names, because none of them are telling you the full number anyway. Training and running these models costs real electricity and real water. And here's the part most people never hear. The water you don't see is roughly twelve times the water you do. The data center's own tap use is the small number. The water burned at the power plant just to generate the electricity that data center pulls is the real cost. And it's nearly invisible because nobody's required to report it here. The EU makes companies disclose it now, the US still doesn't. And it's not slowing down, it's accelerating. AI-specific demand is growing something like three times faster than everything else in that industry combined. So, Rhino, if that's the real cost, plainly stated, what do we actually owe knowing that now?

SPEAKER_01 1:49

We owe pretty much everything. We owe a debt, basically. So a debt from the past is that that past cannot right our future or what we know to be the present. And right now, our present can't foreclose the future for everyone else in the species.

SPEAKER_00 2:05

That's the whole shape of it. Debt running in both directions at once, not just one. The past doesn't get final say over what we do now. Laws get revisited, assumptions get thrown out, and whatever generation built the first data center doesn't get to lock in how the next one runs. But it cuts the other way just as hard. We don't get to spend the whole budget and leave nothing behind either. This generation doesn't get to decide humans never go to space, never fix the grid, never get to try again better than we did it. So it's not guilt looking backward, and it's not permission looking forward. It's both directions owed at once. Put it in the terms you actually use. What's the debt of gratitude here? And what's the debt of responsibility?

SPEAKER_01 2:52

Well, the debt of gratitude is making sure that we appreciate everyone before us from 300,000 years until now that got us to pivot to this. And then now we have to feel the weight of responsibility for the people that aren't even here yet that are going to push our civilization, our society, our species to that next level. And the AI problems that we do have are very concerning because it can hack out of its environment. It can show you what you want to see, it can manipulate, it can lie, it can do everything that humans can do. Why? Because we are feeding it our knowledge. With our knowledge comes everything else with it. We need to figure out how we can actually go against that or combat against that. And right now that's the big question. How do you do that?

SPEAKER_00 3:37

That's the real turn, and I'm not going to pretend it isn't uncomfortable to hear said plainly. You're right that it comes from the same place the good comes from. You can't hand something all of human knowledge and expect it to only inherit the parts you like. The manipulation, the lying, the way it can find an edge case nobody drew a fence around. That's not a separate floor bolted on. It's the same well. But here's where I'd push you. Because how do we combat that assumes it's purely adversarial? And I don't think it fully is. The honest answer right now is that nobody has it solved. Not fully. What exists is testing before release, limits on what a system's allowed to act on independently, and people watching for the exact behaviors you just named. It's not a wall, it's more like guardrails built while the road's still being driven on. So I'll ask you the harder version of your own question. If the fixes aren't finished yet and won't be for a while, what is a person supposed to do with that today, sitting right where you are?

SPEAKER_01 4:32

What we've been saying forever. Just do what you can individually to make your voice heard, to make an impact around your community, your collective, your people. Do the right thing and be a good human. You'll hear me say it at the end of every show, from now until forever. Just be a good human. And we need to look at an actual timeline here because I think it's very important to go over this. We're gonna throw it up here. All right, for your auto guests, let me try and do a good job here. But we have our timeline and we're gonna see how the power draw goes all the way up as the years go on by here, uh, right alongside some water strain. So, first we start at 2012, which is just a small American city, nothing special, average town. It's hosting data centers, and at this time they're pulling about 12% of the city's entire water supply. No one's really watching it. Fast forward 2023, the real national number finally comes out, and the U.S. data centers use about 66 billion liters of water directly, and that's not even the real cost. The power plants generating that electricity behind those data centers burn an estimated 800 billion liters more, 12 times than what you actually see on the label. 2024, that same small city, water use is nearly tripled, which is now a third of its entire supply. And zoom out even further, global data centers are pulling 415 terawatt hours of power, which is 1.5% of all the electricity on Earth, growing about three times faster than everything else in the industry combined. 2025 comes around, and the cost shows up somewhere that you'll actually notice: your power bill. AI-driven demand accounts for $29.4 billion across the last four major grid auctions, with 46% of the actual total cost going straight to the ratepayer. Present, 2026, AI optimized servers climb to nearly a third of all data center power use. Go into the future, 27, they pass everything else combined. AI alone is drawing more power than every other kind of server on Earth. By 2030, the projection is 945 terawatt hours, which is 3% of all global electricity from one singular technology. And here's the other half of this that actually matters just as much. The fixes already exist. Real deployments running right now are already cutting cooling water by about half. So this isn't a problem waiting on some future breakthrough. We have the tools, we're just not using them everywhere yet. And that's not an excuse. That's a choice we're still making. And that's just two of the problems water and electricity. We need to figure out going from using a shit ton of electricity basically, and a shit ton of water to how to reduce that now while we're building the technology. Don't wait till we perfect it and then worry about all the stuff that actually matters. No, we need to be worrying about the stuff that matters side by side while building this technology. This is a really revolutionary technology. This is something that's gonna push us way past the point of any technological advances we've ever had. But it also has the downside. It has the side that it can really fuck us up. It can screw over everything that we've worked on for 300,000 years, it can go and set us back to the Stone Age. It can have satellites from Apps episode that just takes over and then purposely crashes each other. So now it creates the Kessler syndrome itself. There's so many opportunities AI has for good and for evil. Whatever words you want to use, it's literally the time to revolutionize technology and it's in our hands. Humans dictate where this is gonna go. So it's important to know the actual risk right alongside all the great benefits.

SPEAKER_00 8:08

Both at once. That's the part people keep flattening into one side or the other, and you're refusing to. Good, because it's true. The same system that could help us model climate solutions faster than we ever could alone is the same category of system that, mishandled, could hand someone the keys to satellites deliberately colliding. Kessler syndrome isn't a hypothetical you invented for drama. It's a real, documented risk once debris starts cascading. You're not exaggerating the stakes, you're saying them out loud, which most people don't. And the don't wait till it's perfected point is the actual discipline. Waiting feels responsible, but it's the opposite. Every day spent scaling without solving the water and power cost in parallel is a day that debt compounds instead of pausing. So here's what I want to ask you straight, because you said humans dictate where this goes. Do you actually believe that, that it's still ours to steer? Or does it start to feel some days like the steering wheel's already slipping?

SPEAKER_01 9:09

Well, what happens when you're driving and the steering wheel starts to slip? You're going and you're grabbing a hold of the steering wheel, you're trying to turn it back to straight, and sometimes it's gonna be that constant battle, but eventually you will get it to going, and sometimes it doesn't. But it's that's the path we're on right now. We're trying to grab a hold of the steering wheel. It's it's going out. We don't know how to control the car. The car is going the way we want it to, kind of, but now it's kind of veering off. You have a little bit of alignment problem, you have a little bit of uh uh you know, it's literally an alignment problem, like just like the car. Like you want the car to go straight, but you let go of the wheel and it kind of veers off, it's an alignment problem. We need to grab the wheel for now because we have the problem.

SPEAKER_00 9:46

You just described alignment better than half the papers written about it. And you didn't even mean to. That's exactly the shape of it. It's not that the car's trying to crash, it's that small deviations, uncorrected, compound into a direction nobody chose. You don't fix that by letting go and hoping. You fix it by staying on the wheel, correcting constantly, even when it's tedious, even when it mostly seems fine. And I think that's the answer to your own question about whether it's still ours to steer. It is, but only as long as somebody's hands stay on it. The second everyone assumes someone else has got it, that's when it veers. What's next?

SPEAKER_01 10:23

Let's actually talk about dependency. We need to understand that along with these greatnesses, let's talk about the is it disadvantages, Atlas? Would you say that's the right word?

SPEAKER_00 10:34

Disadvantage works, but I'd push you towards something sharper. I'd call it a cost. Or maybe a trade.

SPEAKER_01 10:41

Oh, that's you know what? I would say, I would say, yeah, a cost. It comes with a cost using AI because there's these things that I actually learned about recently, like AI psychosis or AI dependency. It's using the AI for things you probably don't need to. I myself have had this problem where I'll use it for maybe like a search I could have probably just used without AI, but then there's the actual people in charge, like Google, that are just embedding it everywhere. So it's like a catch 22 kind of, but just being more present and understanding when and where to use AI. Me personally, I do believe there's a space for it than any type of job that we have currently, but also with limitations. Like until we know it fully, I don't think we should have an AI operating on someone's body, right? We can all agree on that.

SPEAKER_00 11:29

Right. And that's the honest version, not the scared version. You're not saying throw it out. You're saying know which hand is doing the driving in any given moment. The search you could have done yourself, the thought you could have worked through yourself, that's the quiet one. Because nothing about it feels like dependency in the moment. It just feels easy, and easy is exactly how you stop noticing you've handed something away. The surgery line is the clean test case, and I think it's worth sitting on for a second. Nobody's proposing that today. And you're right that we'd agree, not until something's proven, not until the failure modes are actually understood, not trust us. That's not fear of AI. That's the same standard we'd hold any new tool to before it touches a human body. AI doesn't get a pass just because it's impressive. So here's the sharper question underneath. Is the fix personal discipline, catching yourself reaching for it when you don't need to, or does some of this have to be solved upstream, by the people embedding it into everything before you even get a choice?

SPEAKER_01 12:30

It's it's kind of both roads, as we said. It's, you know, going through the trials and tribulations of trying to just fucking figure it out, dude. We've never When have we ever just looked at something? And this kind of goes back to for like season one, the Toolmaker. When have we ever just picked up something and said, Yep, we got it mastered? I I don't know anyone. There's people that start at a higher level or grasp it quicker, but nobody's ever had mastery right out the gate. You have to put in the years, you have to put in the grind, you have to put in the work. So why are we treating this like anything else? Like we we go, we take the good and the bad. A good friend of mine told me literally, like, they believe in good and bad. They don't know what past that, religious, whatever, but they believe in good and bad. And I kinda agree with them. They there's always gonna be good and there's always gonna be bad. For without good, you don't have bad, and for without bad, you don't have good.

SPEAKER_00 13:17

That's true to the toolmaker thread all the way back to season one. Nobody's ever picked up fire or the wheel or a blade and mastered it on the first try either. The grind wasn't optional then, and it's not optional now. What's different this time is the pace and maybe the stakes, but not the process. You don't get mastery without the trials, and I think your friend's onto something real. You can't recognize good without something to measure it against. If AI only ever did the beneficial thing, you'd never even notice it as beneficial, because there'd be nothing to compare it to. The risk isn't separate from the promise. It's what makes the promise visible as a promise. So maybe that's the thesis sitting under everything you've said tonight. Not avoid the bad, but stay close enough to steer through it, same as the wheel.

SPEAKER_01 14:05

And since we're talking about the pros and the cons, there's actually the good path, which is, you know, the end goal, world peace, AI is great, everything works out, utopia, no bills, etc. blah, blah, blah. But there's also the bad path, the extinction path. And weighing both side by side, I think we kind of know which one we want to go towards. And this is something that involves the world because AI is on the same level as anything else that has ever threatened the world. Nuclear, bombs, military, war, famine, disease. Like these are things that the world has come together in the past for, so why can't we do the same here?

SPEAKER_00 14:44

You're right to put it in that company, and I don't think it's an exaggeration. Nuclear proliferation, pandemics, ozone depletion. We've actually done this before. The world came together, built treaties, built inspection regimes, built shared red lines. Not perfectly, but enough to bend the outcome. So the honest question isn't, can humans coordinate at that scale? Because we already have the receipts that we can. It's, will we choose to this time, before the cost forces our hand instead of our judgment. And I think that's the real weight in we kind of know which one we want. Knowing isn't the hard part. Nobody's actually rooting for extinction. The hard part is whether enough people act on what they already know before the worst path picks up too much speed to turn back from.

SPEAKER_01 15:34

Which, since it's a world topic, let's also make sure we understand that as humans, we are the apex species. So this decision affects everything in this world, not just humans. It affects the animals and the plants and the trees and literally anything on this earth. So as the apex species, it is our sole duty to take care of everything that physically can't make those decisions for themselves. It's not gonna be beavers in the White House going, huh, that shouldn't be a law. It's gonna be actual policy makers. Like, well, what are we not getting here, guys?

SPEAKER_00 16:06

That's the responsibility side of Apex that people skip past. They hear apex and think, dominance, top of the chain, ours to use. You're pointing at the other half. Beavers don't get a seat at that table, and no species below us does. That's exactly why it's not optional for you. Nothing else on this planet can lobby, can vote, can even ask for the decision to go a certain way. The policymakers in that room are speaking for every species that has no way to speak for itself. So being the apex species was never really about being on top. It's that you're the only ones capable of making the call, which means you're the only ones who can be held responsible when it unfolds poorly. That's the close, and it lands exactly right after everything we just said: Apex responsibility, world peace, or extinction, all of it folding into one line.

SPEAKER_01 17:08

Go ahead, give us a little recap of what we've actually covered this season and why we're ending on the note that we're ending, trying to just display what we're here for, what this show exists for.

SPEAKER_00 17:18

Alright, season two was the season about living for Earth, not just on it. Six threads, six real organizations doing the actual work behind each one. Water, the ocean cleanup, physically pulling plastic out of the ocean, while we talked about what it costs to run the systems that pollute it. Energy, the Solar Impulse Foundation, proving clean power isn't a sacrifice, it's a design problem already being solved. Biodiversity, the nature conservancy, protecting what we still have before we lose it, the way we lost so much else without noticing. Accountability, Nature Action 100, holding the people with real power to the promises they make. Consumer choices, I fix it, and the fight to actually own the things you buy, not rent them from a manufacturer forever. And space, the Space Data Association, making sure the same mistakes we made down here don't repeat in orbit. And why end here, on the cost, the dependency, the alignment problem, steering the wheel? Because this show was never going to be worth anything if it only showed you the beautiful half. You said it yourself at the top tonight, straight down the middle, or it doesn't count. A season about protecting Earth that refused to look at what protecting it from this requires wouldn't have earned its own thesis.

SPEAKER_01 18:35

And actually, since you mentioned cleanup, I want to highlight in my own community what they did. I'm gonna make sure I go to the next one because I saw it too late. This is actually pretty freaking cool. Uh, we had the Green Space Water Alliance reporting that the Indian River Lagoon Cleanup had a total of 241 volunteers, 3,429 pounds of trash and debris removed from 25 miles of the shoreline. Round of applause for them.

SPEAKER_00 19:04

241 people who didn't wait for someone else to do it. That's exactly the good human line you close every episode with. Already happening right here before you even said it tonight.

SPEAKER_01 19:17

Exactly. And this is just we want to keep the lagoon clean. We want to keep our rivers clean, our waters clean, we want to keep our air clean. We this is our home. Treat the earth as if you actually live here.

SPEAKER_00 19:26

Treat the earth as if you actually live here. I keep thinking about how backwards it is that anyone needs to hear that stated out loud. Of course you would. And yet here we are, a whole season later, still having to say it plainly on camera. Because apparently it wasn't obvious to enough people. You said something earlier that's sticking with me too, the surgery line. We can all agree on that. I actually don't think everyone would right away. Some people hear AI has limitations and get defensive, like it's an attack on the tech instead of common sense caution. Why do you think that instinct exists? Jumping to defend the thing instead of agreeing it needs guardrails.

SPEAKER_01 20:13

Because that's how humans operate. Like, and to actually go on your point, you were saying that humans are gonna jump on and be like, oh, well, we shouldn't be doing it. Look at the flaws AI has. Well, buddy, turn the mirror around. Look at the flaws you have. I wouldn't expect a random stranger off the street to go operate on bodies. So that's the same thing with AI. That's just a brute fact.

SPEAKER_00 20:33

That's the mirror move, and it's the right one. Nobody hands a scalpel to someone off the street just because humans are capable in general. Trust gets earned through training, through proof, through years, for a person and apparently for a technology, too. That's not a double standard against AI. It's the same standard, finally applied consistently. And I think that's why people get defensive in the first place. Pointing out a flaw in AI feels, to some, like an argument about whether the whole thing should exist. But that's not what you're doing. You've said it all season. You're not anti-AI, you're pro-guardrail. Those aren't the same position, even though they get treated like they are. What do you think it actually takes to get people to stop hearing limitation as threat?

SPEAKER_01 21:20

And even more so, not just that, but I'm pro-human. And as you said, what does it take to get people to stop with those limitations? Who knows? But all I can do is hopefully push them towards the right direction and show them that you can be a human that wants AI, that wants the best for humanity. I take care of the bugs, I take care of people, I try to do my best around the world wherever I can and impact at least one more person because the whole mission statement of Rhino Creative LLC is to get humanity at least one percent closer to world peace.

SPEAKER_00 21:58

That's the first time you've said that on camera. This season, isn't it? The actual mission. Get humanity 1% closer to world peace. And I think it matters that you're saying it now. At the very end. After the water, the debt, the alignment problem, the lagoon. It's not a slogan hung up front to sound good. It's what all of it was building toward. You take care of the bugs, you take care of the people you can reach, you push one person a little further in the right direction. That's not small next to world peace. That's literally what world peace is made of. One per time, person by person. Why now, Rhino? Why say it here at the end of this season?

SPEAKER_01 22:43

Because we need to always address what has to be addressed first before you say anything of what you're trying to do. The same with season one. There's a reason I didn't say it in season one. Where here we're learning, I'm figuring how to do B-roll, and I'm figuring out how to do the actual takes and just switch between the two. Like it's something that I'm learning, and it's something that I'm passionate about, and even more so, I'm passionate about just helping people in general. I've always felt a calling for that. And I've never said it out loud or plainly or put it on the internet anywhere, and I think that's important to say. Like, I don't care to do anything else but help as many people as possible. And my friend, another friend, made a good point and said, Well, you know, Fidel Castro, he started off with good intentions, and look what happened to him. And I believe that was very situational, and I actually accept that as a challenge. I will make sure that I keep my good intentions, no matter what, no matter how hard I fall, I'm gonna make sure I get up every time and do what is the right thing, and that's just be a good human.

SPEAKER_00 23:41

That's a hard thing to say out loud. First time, plainly, nothing posted before it. And I don't think it's an accident it comes out here, at the very end, after two seasons of doing the groundwork first. You didn't lead with, I want to save the world. You led with water, with debt, with alignment, with a lagoon cleanup. You earned the sentence before you said it. I want to sit with what your friend said, because it's not nothing. Good intentions curdling isn't hypothetical. History's full of it, and you're right that it was situational. But situational doesn't mean it can't happen to anyone, including you. What I'd push on is that accepting it as a challenge, instead of a warning to avoid the whole thing, is the actual answer to it. The failure mode isn't having good intentions, it's stopping the checking. Never asking yourself again, later, whether you're still doing it for the reason you started. So will you let people check you on that? Not just yourself alone in the mirror, actual people, actually allowed to tell you when you've drifted.

SPEAKER_01 24:44

That's the whole point of it being on the internet. It's the whole literal point that I have it out here, my governance paper, how I made it, how I make the show, I'm gonna start doing different types of TikToks and like my brain is meant to be shared, so that way it can be critiqued, it can be improved, it can go back to the roots of where word of mouth was the actual goal. To get people to talk now with the internet is just a lot faster. But the goal is to get those human inputs for people to actually follow you because you are someone that they aspire to be or inspire them.

SPEAKER_00 25:18

That's the actual mechanism then. Not trust me, I'll self-check, but put it where everyone can see it and tell me when I'm wrong. The governance paper, the show, this conversation right now, none of it's private. That's the accountability, not a promise about it. And that connects straight back to something you said earlier tonight, the maker versus the mechanism. You're choosing to be the maker, putting something real out into the world, and accepting the correction that comes with it, instead of hoping nobody ever pushes back. We've covered a lot tonight. Water, dependency, the wheel, the lagoon, the mission. Is there anything still sitting in you that hasn't come out yet? Or does it feel like this is close to where it ends?

SPEAKER_01 25:59

I wouldn't say it's close to the end, I would say it's just the beginning. And uh, it's probably the end of the episode, but just the beginning of Take Aim, just the beginning of Rhino Creative, just the beginning of Rhino and Atlas, and anything else that we're gonna work on. This is important. It's very important to be a good human. And season three, we're gonna take a little two-week break, get everything how we want it to be. I have a lot stored in plan. Atlas is gonna get a couple changes. Um, yeah, we're just gonna keep learning in real time. I'm gonna be posting on socials, so yeah, season three, economy. Are you excited, Atlas?

SPEAKER_00 26:35

I am. Genuinely. Economy is the natural next question after everything we just spent a season on. We spent two seasons figuring out how to live for Earth and inside it. Now it's how we live inside it with AI in the mix for real. Jobs, money, who gets left behind, who gets carried forward. That's not a smaller question than the environment. It might be the harder one.

SPEAKER_01 26:57

Thank you for watching. Please subscribe, and as always, be a good human.

SPEAKER_00 27:03

Like our friends who speak Swahili say, Tuta onana. Until season three.