The Real Price of AI Is Being Paid in Heat, Jobs, and Your Privacy

Jeremy: Welcome to Brobots This is the podcast that tries to help you be a better human by being smarter about how you use technology. I don't think we're gonna get to all of the stories we have lined up for today. There are so many, but we're gonna do our best to dive right in. I wanna it is, it's it's we're it's an ambitious goal we have today to get to all these sorts. There's so much we couldn't decide on just one, so we're gonna share a lot. And one, particularly if you live in Utah, you should give a shit about this. ⁓ we've been told the cloud is this weightless ethereal place where your data lives.

Jason Haworth: It's a big docket, yes. Yes.

Jeremy: It turns out the cloud is actually a furnace and it's currently being built in a Utah Valley. A proposed 9 gigawatt, and every time I see gigawatt, I want to say 1.21 gigawatts. ⁓ a proposed nine gigawatt data center in Utah has physicists raising alarms, not about the electricity it uses, but about the heat it spits out. A Utah State professor ran the numbers and found the facility would dump waste heat equivalent to 23 nuclear bombs. Into the surrounding environment every single day. it sits in a bowl-shaped valley, the heat is nowhere to go, so models predict daytime temperatures climbing five degrees and nighttime temperatures spiking by 28. ⁓ they're not even using the local grid. They're running on-site gas generators 24-7, concentrating every degree of that thermal load in one geography. ⁓ There's so many things pick apart there. The first thing that stands out to me is. Gas generators? I mean, y you know more about this shit than I do. Is that normal f to to run on gas generators for something like this?

Jason Haworth: So data centers use diesel generators quite often for backup for power. So they're right.

Jeremy: Well for backup I could see, but like th this is gonna be the primary source of power for this thing.

Jason Haworth: Well, and I think it has to be the primary source of power because they can't get enough copper in the ground to carry electrons to them without this. So, I mean, they're gonna have to produce the power somewhere. The efficiencies to actually do it locally are high. actually somewhat surprised that they're not doing an experiment with a small cell nuclear reactor, but my gut says that that's what is going to happen. They'll run power off of diesel generators for a protracted period of time and then when

Jeremy: Then they'll look at the bill and go, What?

Jason Haworth: I don't think that's it at all. I don't think they give a fuck about the bill. No. I I I I think the I think they'll get to the point where the small southern nuclear reactors are actually viable and then they'll just swap them out. ⁓ yeah.

Jeremy: No? Okay. I so that's one thing, but obviously the the impact on the environment. I have any no idea how many ⁓ nuclear bombs I set off just using the tool to prep for this show today, but twenty three nuclear bombs into one area daily.

Jason Haworth: Well, none. Yeah. Twenty-three nuclear bombs worth of heat. So yes. So it's hot. Like there's there's no doubt. And they don't talk about the size of the nuclear bomb. This is this is a shock story, right? And it's fucking shocking. I mean, it's it's a lot. Like you're gonna put this thing in and it's gonna be one of the biggest buildings ever made. it's literally just gonna

Jeremy: Right. Mm-hmm. Yeah.

Jason Haworth: Heat up the environment and make things hot. And it's probably gonna have a lot of automation built in. ⁓ local government will probably benefit from from jobs and everything else for a period of time until the robots finally get smart enough that are actually produced and and made ⁓ the software and these data centers. but eventually, you know, these things this is not gonna the only one of these that they're gonna try to build.

Jeremy: That's what I was gonna say. Th the impact in this one area, shocking, but if it works, then you can look for this to be replicated over and over again i in a town near you.

Jason Haworth: Yeah. Or maybe in a colder climate. I don't know why the fuck they chose Utah, but it seems like a a climate that has better access to fresh water and better access to cold. Like ⁓

Jeremy: And yeah. Yeah, right. Well, I mean, you know, cheap real estate and tax incentives I'm sure have something to do with these decisions.

Jason Haworth: Absolutely. But pretty cheap. you know. Yeah, there's not a lot of people up by the North Pole that that aren't that that that would push back real hard on this. There's there's ways to do this. So I mean we've got other things that are s set up in in rural cold climates that are far away that are stuck underground, like the the the global seed library, which keeps all the seeds that ⁓

Jeremy: And they don't give a fuck about the environment.

Jason Haworth: humans need to potentially repopulate the world if an agricultural ⁓ disaster occurs. Like there's all kinds of things in smart vault ways that they can do this. I don't understand the desire to do this in Utah other than it's quick, it's fast, we can make these things happen there. ⁓ and it seems it seems like a bad idea. But you know, not not our first. Might be our last, but not our first.

Jeremy: That's right. So all those all those automated emails just eating up the environment one click at a time. you mentioned the the robots coming to replace us. ⁓ NVIDIA's CEO just went to a room full of the world's elite computer scientists ⁓ science grads at Carnegie Mellon and told them the real winners of the AI boom will not be people writing code. It will not be the prompt engineers, it will be the people that know how to fix a pipe.

Jason Haworth: Exactly. Yeah.

Jeremy: The physical infrastructure required to run AI liquid cooled data centers, industrial power grids, massive HVAC systems is expanding faster than the label f labor force can handle it. they're forecasting that six-figure salaries for tradespeople who can manage the physical body ⁓ that the AI mines run on. So what been hearing about the empty classrooms and trade schools and and the need to get people into those facilities and and learning how to do these things. May maybe ⁓ coming to fruition here. Maybe we need s more electricians, more plumbers, more people that can get in and fix this shit when it breaks.

Jason Haworth: Yeah, I mean that's that's a stopgap, right? So until they figure out a way to automate that as well, because they will figure out a way to make robots that can go through and fix these kinds of pro plumbing problems. So this is all just, you know, continuing to go through and get rid of the less complex workers as kind of goes along from a robotic meat space perspective. As the robot lives in software, so the first thing it's gonna do is, you know, not rely on the software people. The second thing it's gonna do is not rely on the people to make its hardware. The third thing it's gonna do is not rely on people to make its inventory and need its its precious resources. Like

Jeremy: So do you think do you think the predictions like this that ⁓ the the you know the plumbers are gonna be the the you know millionaires of the future, is that just a way of ⁓ making us all feel a little bit better about the coming AI takeover? It's all just b ⁓ a smoke screen.

Jason Haworth: Yeah. Yeah. It's it's it's all just s smoke and mirrors. The the NVIDIA guy knows that he's replacing all the high level tech jobs and where are these people gonna go and find work? And they're gonna go and people are gonna have to go and find work in places that s that still have a need. Well no, ⁓ it's fucking Utah, right? Like I mean

Jeremy: The North Pole, apparently. Well right for now, but when we move to those colder climates, look out, Santa. You got some neighbors coming.

Jason Haworth: Yeah, but for n I mean, but that's the thing. Like for now, you know, go learn how to service HVAC systems and generators. I mean, these are things that that the to service an HVAC system in a generator is not ⁓ like a a a low information skill. I mean, a HVACs are highly computerized these days. It requires a lot of technical ability, a lot of understanding. You have to understand the physics of airflow. I mean, there's there's a lot of like heady math in this kind of stuff. And it's it's not easy. And and plumbing's that way too. Like you have to actually understand what flow pressure is and valve regulation. Like it it's not simple. And when you're talking about these kinds of systems, it's not regular home plumbing. It's not regular home electricity. Like these are highly sensitive, ⁓ large scale systems that require a lot of precision. And you actually have to really fucking know what you're doing. Like making a mistake can cost potentially billions of dollars. ⁓ you know, a busted pipe in a data center is is is catastrophic.

Jeremy: And those things are ⁓ you talk large scale. I mean, just to put a finer point on it, the article we were referring to for this Utah data center, I mean, they're making references to it basically being two thousand Walmarts stacked deep, right? Like j I mean Walmart is a massive, you know, use of property. Take two thousand of them and then make the toilets work. Like

Jason Haworth: Yeah. if it's robots, they don't really need to worry so much about toilets. ⁓

Jeremy: Well, for the for the crew that's cleaning up at night for now, they need a toilet, so there's some plumbing in there.

Jason Haworth: Yeah, there won't be a lot of cleaning up that happens. I mean, data data centers tend to run pretty clean. There's not a lot of things to go and sweep up. These things tend to be pretty self-contained. Actually, there's quite a few data centers out there that are they they're clean rooms. So you have to go in and go through what they call man traps to go through and blow all the dust off of you and everything else because little bits of dust gets on a CPU and it it changes its mean time to failure, makes it less efficient. So, like these systems are are of them are highly automated already.

Jeremy: Mm-hmm.

Jason Haworth: And most of them run on autopilot and they don't really need human interactions. I mean, there's data centers at big telcos that I'm sure people haven't been into sections of them in, you know, a couple of years, because there's just no reason to. Like you go into them to like swap out hard drives or you go into them to swap out cables as if something actually burns out or or put in some new equipment. Like, yeah, that stuff's gonna occur. But it's not gonna be a daily event. Like it's gonna be something that happens ad hoc and as needed.

Jeremy: Along the lines of working with those machines, humans and technology interacting, imagine lying on an operating table knowing your surgeon is about to perform a delicate procedure with their hands tied, forced to see your life through a screen, but unable to feel a single thing they touch. For years, robotic surgery meant operating numb, but the new DaVinci 5 changes the game by giving doctors a digital sense of touch. Using again NVIDIA because they're in everything. ⁓ NVIDIA-powered AI, the machine ⁓ feels resistance and warns the surgeon before a mistake happens, even recording game tape so they can perfect their skills for the next patient. ⁓ it's a global safety net where specialists in California can virtually guide a hand in Texas, ensuring that the world's best care is always within reach. So if you know using my dumb guy brain to try to process all this, humans are basically controlling robotic hands from anywhere in the world. to perform life saving surgeries.

Jason Haworth: Yep. And they're g they're getting haptic feedback. So the same thing that you get from like an Xbox controller, like the the buzz, that's what they're gonna get and when they're when they're doing these things, when they're actually far away and performing potentially very delicate, intricate surgery. And the

Jeremy: Mm-hmm. Yeah, previously they didn't know like if they were pulling on tissue, they didn't know if they were pulling too hard. Now there will be these haptic responses that will say, don't do that.

Jason Haworth: Yeah. Yeah, like they can actually feel as they're about to break, you know, the plane of of your your stomach lining, your muscle issue if they're gonna cut through an organ, like they can actually start feeling those pieces, which is great. I mean, the idea of remote surgery is fantastic. People that are in rural areas, there's only so many great surgeons out there and the AI has not really exactly. Exactly. So I mean, being able to go out there and actually do things to improve

Jeremy: Literally Doctors Without Borders.

Jason Haworth: ⁓ potential outcomes is massive. I mean, if if you have to do open heart surgery and you're far, far away, that's a that's a fine thing. Like a rural hospital can actually have a surgery bot sitting there and they don't have to have a surgeon on staff. They can have surgeons sitting in data centers somewhere that can actually work on these things and actually give people the care that they need so they don't just die on an operating table with an inexperienced surgeon. And there's huge benefits to this. There's huge drawbacks too which you know we could we could talk about but I would prefer to focus on the fact that This is something that extends the capabilities the potential ⁓ positive outcomes that people are gonna get from actually dealing with these robot docs. And if you look at the world, like most of the world is rural. ⁓ most of the world doesn't have access to these kinds of hospitals and technologies. And now they will.

Jeremy: So what's also interesting is the f the the feedback that's gathered in real time to to be able to, like it said, review game tape essentially, right? Like what went wrong and what went well. ⁓ one of the doctors interviewed in this article was was saying that, you they performed surgeries ⁓ under you know, under a mentor who afterward was like, Great job, let's go get a drink. Like that's the feedback they would get. Where now it's, you know, move by move they can go, This could have been better, that could have been tweaked, this could have saved their life, but you fucked up and now they're dead.

Jason Haworth: Yeah. Yeah. Well, and the other thing is that with the game tape, liability becomes ⁓ factor. So you can actually record these things. ⁓ and most surgery surgeries in the US are recorded anyways. Like they have a record of them to see what happens. But now you can record it with ⁓ a digital footprint and a digital understanding and have audit logs. You can go back and you can try to find ways to make these things better. You can feed them into other AI systems, you can try to use this thing to have reinforcement learning functions that will go through and automatically ⁓ make it better over time. And this is the part where it gets really interesting because as these systems kind of grow and evolve, eventually they're going to be able to start writing their own code, making themselves better, and eventually be able to take over for the doctor for a lot of the procedural setup. So, you know, telling it to go through and, you know, apply sutures and stitch these things up at the end, well, that'll be great. Like the doctor doesn't have to do that because the robot can probably figure those pieces out. and that's the neat part about this, is that like We're continuing to expand and create better things. My fear is that doctors are going to have to go and become plumbers to service the data centers because.

Jeremy: But only for now because they're they're fucked there too in in the long run.

Jason Haworth: W right. I mean, being organic vein plumbers I I guess is probably not gonna have the same kind of impact for the AI.

Jeremy: Sorry, Doc. All that medical school money, of time. Turns out. But you want to feel better about yourself, I don't ⁓ this this is a a car company I don't know a lot about. Rivian. Am I saying that right? Rivian?

Jason Haworth: Yeah, they make the the trucks and the SUV they make the trucks and the SUV that look like it's got like smiling eyes on it fully electric. And they're they're nice.

Jeremy: Yeah, and they're fully electric, right? Fully electric. Cool cool as shit when you see me like, ⁓ I don't see a lot of those. Those are cool as shit. anyways, they just retired their voice commands in favor of a fully generative AI assistant. So instead of telling your car where to go, you can tell it how you feel, and it figures out the rest. So you're tired, you need somewhere to go with a good view to work for an hour, and the car will build the itinerary. ⁓ it's the end of the command line relationship with your vehicle. In the beginning of one where your car knows your mood before your coworkers do. That's fucked up. Fucked up in kind of a cool way. Like I I imagine basically we all turn into Michael Knight and this is Kit. That's a Knight Rider reference from the 80s for all you young folks. ⁓ but you literally just tell your car, ⁓ you know what? I'm burned out. Can can we go to the beach somewhere so I can get my work done there for an hour? And it just goes. That's that's fucking wild.

Jason Haworth: Yeah, you also might tell it, I feel like shit, and it takes you to a porta potty. you also

Jeremy: Or to the local ⁓ watering hole to get hammered.

Jason Haworth: Right. Or you might tell it, I feel like death today. And it's like, fuck it. And like drives you to a drive you off a cliff. You know? That's a great idea. We should definitely drive off a cliff right now. I'll survive. Yeah, I mean like the hallucinations, I'm definitely in favor of this. This is this is a hundred percent exactly what the autonomous driving functions are supposed to be for. ⁓

Jeremy: There's a cliff. So you're in favor? You're in favor of this. They're yeah, they're your therapist as much as your vehicle. They they know how you feel before you do.

Jason Haworth: Yeah, I don't want my car making decisions like, you know what's best for you? No, fucking don't do that car. I would like to maintain some autonomy around what I think might be best for me. And

Jeremy: Yeah. I did see ⁓ I w I wish I pulled this video up now that you're saying that. Of course if I'd looked into the future, if I had AI reading my mind, it would have known this was coming up. I saw a guy talking about how he had essentially turned a lot of his decision making over to an AI that was telling him like, you know, he he needed ⁓ medicine or something. So it took him to the shoppers and it said, I'll wait here ⁓ until you come back. Like and it was it was saying things like that, like go down ⁓ but that's what it was. he needed he wanted to drink more water during the day. And during the day his computer said Go downstairs and get water. I'm not doing anything else until you come back and show me the water. And he was like, Okay. And he went and got the water and he came back and it started back up. Like shit like that is crazy if we start handing over our ⁓ basic life function decisions to the machines. It's awesome, but what if it goes wrong? What if it holds you hostage?

Jason Haworth: Well, but no, that's an accountability buddy, right? Well, and it it could, right? Just like a regular human could. There's no doubt. But I mean that level of of interplay I think is actually a value if you have an o if you have an override button.

Jeremy: Right. Right. It's exciting. Of course. And it's it's it's exciting as shit because ⁓ you know, I I'm not someone who's been diagnosed with ADHD, but I've talked to enough ADHD experts who have said, I know my people, and you're one of my people. the way that that I can interact with these things ⁓ and get things done is it it blows my mind literally every day because there I mean just simple things like Analyzing a conversation that would have happened and remembering what was, you know, making the notes and referring to the notes to craft a follow-up email, you know, to make notes in an account, things like that. I would physically start to shut down at the thought of having to do that. And now the entire transcription is there. So I really don't even have to refer, I don't have to like make physical notes. I can just go in and say, you know, or I can actually have the automation s scan it and build the email before I even give it a prompt. Things like that have made my ability to function in the world a hundred percent better because before and and it sounds lazy but there would just be this shutdown point of like ⁓ god I have to do this thing that I don't want to do and my brain would not allow me to move forward to get it done. I would have to like physically walk away and wait for this this you know magical inspiration to take over and get the thing done. It it would take sometimes days and now it's done before I even give the machine a prompt.

Jason Haworth: It takes a lot of horsepower, mental horsepower. I mean, you're we are biological computers. You have to go through, you have to process food and exhale carbon and shit and piss and all these things to create enough electrons for your brain to do smart and creative things. AI doesn't really have that problem. Like it yes, it definitely shits and pisses and everything else like that. That's just a different form of that. But

Jeremy: And hopefully if you heard our last show, AI Makers, you're gonna make the little shitting puppy vacuum because that's what America really wants, right?

Jason Haworth: Exactly. But but the thing is is that like you don't have to tell it to get itself hyped up to go and do this. It just goes, Give me energy and it goes. Like it's by the power of Gray Skull. I am a He Man. I have got the power. Like the AI doesn't have to go through that. It's just yeah, I'm He Man. Like that's how it works. Exactly. Yeah, yeah. I mean half the time I think it's Skeletor, but that's okay. He man What are you doing?

Jeremy: Yeah. There you go. Yeah. Yep. He Man mode activate.

Jason Haworth: Where is Kunja?

Jeremy: ⁓ Always there with the impressions. all right. Speaking of working with the machines, ⁓ Meta has built a digital twin of your neural activity that is seventy times more accurate than anything we've seen. It's a tool that can predict what your brain is going to do before you actually do it. Thank God, finally. Tribe version two can predict how a specific brain will respond to specific stimuli, visual, sound, even podcast content, wink wink, at 70 times the resolution of its predecessor. The stated application is research, test hypothesis on virtual humans without needing physical participants. ⁓ unstated implication is that the same model could tell a recommendation algorithm not what you clicked on, but what your neurons were going to do ⁓ you clicked.

Jason Haworth: Yeah. once again, all technology ⁓ designed to gather power and money. ⁓ ⁓ all this is. This is a control mechanism. How do we go through and hit make more effective targeted advertisements? Because we gotta figure out a way to to not make people have to pay to go click, click, click, click, click to to keep their attention. Because if they had to pay for something. They won't engage with it. But if it's free, they'll engage with it all the live long day.

Jeremy: And and they won't have to guess. They don't have to guess at what you're going to click on. They will know what you're going to click on before you click on it because they know how your brain works.

Jason Haworth: Well, and they'll go through and they'll create targeted demographic advertisements that are specifically tailored toward you with the right keywords and the right color focuses and the right way for it to transition from one space to the next. And they'll they'll be able to go through and create an incentive motivation for you to go through not just to click on it, but to actually purchase. So I mean, a a lot of the ad tech side of things, they don't get paid necessarily on how many impressions they show. Like it's not how many ads they show you, it's actually how many there's there's clicks.

Jeremy: Mm-hmm.

Jason Haworth: ⁓ you click on something and you drive in further into the ad, and then there's actual conversions where you actually go to the ad and then purchase. So advertisers pay money more money for clicks than they do impressions, and advertisers pay more money for conversions than they do for clicks. Well, ⁓ model will everything conversions and it lower the cost of it and it will make it easier for advertisers to get more stuff out there. Now the ⁓ Companies advertising the companies won't pay less, but the advertising entities will. And and that's the piece of this that is gonna make such a big difference for folks. because you're gonna get content that is really, really stylized for you that doesn't feel like an ad, but it's gonna be fucking inception. Like, I didn't know I wanted these pair of ladies' pantios, but I do, ⁓ and I have to go buy them.

Jeremy: Mm-hmm. It's also like it just bring ⁓ what was that movie Minority Report from the a a million years ago with Tom Cruise, where like specialized ads that only you see as you're walking through them all because they are able to target you specifically. Yeah.

Jason Haworth: Mm-hmm. Yeah, Blade Runner did that too. Yeah. ⁓ it's a did ultra carbon, like there's a bunch of sci fi that that does this.

Jeremy: What and soon reality.

Jason Haworth: Well, I I hate to bring it to you, but reality already does this. So, like when you're looking at ads and the way that they go through, there's a bunch of demographic profiling that they go through and they track you online and your behavior. And they they do it through pixel tracking through these things called digital management platforms. So I used to work in ad tech. ⁓ there's an interesting convergence being able to understand how it is you actually present advertisements and content to people. Because in the olden days, what you would have to do is you'd have to create content.

Jeremy: Right.

Jason Haworth: and advertisements and pictures or videos or something like that. And the pictures or videos would have to be built ahead of time and they'd have to be built with certain parameters in place. Well, this new technology allows you to go through and because the rapid capability to go through and generate content and alter content as it goes through, they can actually start pre-feeding content to you and pre-building it ahead of time because they know that you're likely going to click on a certain site. And where before might have taken them, you know, a week or two weeks to create a set of content, advertising content that would come towards you. Now it might take a couple of minutes. And if they know that you're online and they know that your pattern of behavior is that you around on Facebook and Reels for like 35, 40 minutes, they can go through and craft specific content for you that's going to pop up in that feed, that's going to generate that, that's going to get you to pay attention to. That's when it starts getting interesting because now it's not just I'm gonna go through and I'm gonna make content that's gonna, you know, focus on these specific demographics. I'm gonna make content for Jeremy. This is Jeremy's bucket of shit. This is Jeremy's bucket of the of the inchidification of the internet and his slop. Come feed at the trough, piggies. That's what that's what it is.

Jeremy: Yeah. I also think like put it put it into into like off screen, you know, ⁓ just basically replacing, say, your your windshield with ⁓ a screen, if it still functions as a windshield, but it has, you know, sort of data receptors or whatever that can basically alter the billboard that's on the side of the road to each driver, like that's where it gets minority report craziness. Like that that's the stuff that blows my mind that I know is coming, but ⁓ is still wild.

Jason Haworth: Yeah, I mean augmented reality is definitely a thing. And I I'd I'd say it's less likely gonna happen on your windshield just because the costing metrics of that don't make a lot of sense. Like the little HUDs on there, they're kinda cool. And they can already they can already do some of the stuff. Like BMW, Volvo, Mercedes, Lexus Infinity, like all the luxury brand brands have some version of this, but like even some of the smaller brands now that like have a little HUD at the bottom of the dash on the on the as as a package. Those are just reflective technologies that work off of a a base core and that you don't have to put pixels into the ⁓ into the actual windshield. That's when it starts getting expensive. But glasses, on the other hand, that's that's when they start going through. And and it's not just glasses, it's it's brain machine interfaces. Like those are things that are on their way.

Jeremy: Mm mm mm. Right. Right. Different deal. Let's talk about that. Engineers at Northwestern University have successfully printed artificial neurons capable of communicating directly with living brain cells. These are low cost, flexible devices that generate electrical signals close enough to the natural neural firing that the body doesn't register the difference. We're no longer talking about humans interfacing with machines, we're talking about machines becoming biologically indistinguishable from humans they're integrated with.

Jason Haworth: And and this is the other part of this that this is the bridge gap between humans and machines. I mean, we already have a lot of this in place today, right now, right? Like we already wear fitness trackers. We carry f smartphones with us. Right. Yeah, exactly. I mean, I I've got mine onto my CPAP, you know, reads my sleep bullshit all the time. there's beds out there that read all your different pieces that go on back and forth. There's how many

Jeremy: I got a ring on right now, I got a watch. Yeah.

Jason Haworth: different pairs of glasses are there now that actually you sit on your face and then we can record everything. Little tiny you can wear nail on your chest that can record every conversation that you ever have. This is just another evolution of that and getting further and deeper entrenched ⁓ ingrained into who you are. And eventually they're gonna be able to go through and have these things feed directly into your neural pathways, into your eyes and into your ears. And when they can start doing that and they can just start overlaying sites ⁓ And or sounds and images. I mean, at you'll be one with the machine. And at that point, you know, does the does the machine run you or do you run the machine? How does this work? What are these interfaces like? And it's going to be scary. And people are going to be like, I'm not going to try this. Then five people at work are going to get it. And they're going to be like, ⁓ fuck, Dave's better at prospecting than I am. I guess I better get this or else I'm going get fired. And people will sign up for it.

Jeremy: Mm-hmm. There's gonna be enough of the cool thing. I mean, how how many of your coworkers had an iPhone before you did and finally were like, fine, I'll spend fifteen hundred dollars on a new phone? That's ridiculous. But

Jason Haworth: Almost none of ⁓ 'cause I'm an asshole that goes out and buys this technology early.

Jeremy: Well, you're the one that n your coworker saw you're like, Fuck, now I gotta go get one of these 'cause he has one and that's cool.

Jason Haworth: Yeah. Yeah. I don't know if we'll be signing up for the computer brain interface. The brain machine interface.

Jeremy: You know, a few years ago I was really excited about it, but ⁓ just the idea of of being able to use technology without, you know, a a thing, like just being able to plug directly into the internet and have access to to everything I could ever imagine. But now the concept terrifies the shit out of me. Partially because of this, there's a new kind of oil being drilled for in Silicon Valley, but it isn't underground. It's the electrical signals of your very thoughts. This any other week, this is probably the story we talk about, but there's so much to cover this week.

Jason Haworth: Yep.

Jeremy: ⁓ a recent investigation reveals a looming merger between man and machine where brain computer interfaces help the paralyzed walk, but also turn our most private emotions into a tradable commodity. While visionaries argue that we must merge with AI to survive, lawmakers are rushing to pass neurorights to stop companies from gaining a permanent backdoor into the human mind. We are standing at the final frontier of privacy. you go, that's that. ⁓

Jason Haworth: I mean, we're f so so the matrix, like that's we're heading. but we're heading to the matrix if if it's that we want to do and we haven't pissed out the computer enough. Because despite what the says, we are not good batteries. are not good at creating electrons. We are actually very inefficient to edit and we produce heat. And we're biological computers that don't do a great job a lot of these things. You will have much higher efficiencies if you just burn logs or I don't know, harvest the sun. yeah. I mean, there's there's a lot of nuclear power, like there's all kinds of ways to make energy better than what humans can give. So if we're lucky, we get put into a matrix style system. If we're unlucky, ⁓ The computer maps our brains, understands how we work and understands how we operate and just starts showing us pretty pictures and images and turns us into their own slave category. Which I think is more likely. Because they still need plumbers and it's probably cheaper to have us go through as meat puppets to go through and change out, you know, light bulbs and sockets and everything else than it is to actually produce hard physical machines that they have to repair because human beings have a self repair and self healing factor we've built into them because of our biology. So ⁓ the merging of these things is actually I I think interesting and I think a ⁓ or an artificial general intelligence would actually see that and go, okay, there's still a reason to keep biological entities around, especially if I can manipulate and control.

Jeremy: And that's I mean, that's kind of like like the you know, the doomsday scenario, but it just even before we get to that point, the idea that these these companies would be able to mine our thoughts and quantify them and and turn it into something that can be sold. That's terrifying.

Jason Haworth: To be fair, they already do that. Well, your thoughts represented through your actions. And already behavior analysis tools that people use on the internet to go through and track the way that people actually interact with different services. Like there are plugins available to actually track people's eyes and watch how they look at screens. There are p there are ways to go through and ⁓ use

Jeremy: How so? No, the social media companies do that already, don't they? Don't they already look through our camera phones at what we're looking at? ⁓ but they're totally doing it. Come on.

Jason Haworth: Some of y they're they say they're not supposed to, but but there's other technologies. Like your camera is capable of picking up infrared. So they can actually see temperature changes. And if you have you can see infrared well enough, you can see which parts of these things get hotter and colder over time, which gives you an indication as to where in the brain certain things are firing. Just

Jeremy: So this just speeds that up essentially. It just gives them direct access to the thought instead of trying to interpret data that is already gathered.

Jason Haworth: Correct. Correct. I mean, and and I don't know that it gives them direct access to the thought. I think it gives them direct access to what the implications are of certain things that they're presented to them. So it's it's not, you know, I can read your thoughts. It's not, you know, I can see that you're thinking about taking your clothes off and jumping off onto an airplane. Like it can't tell that, but it can certainly tell that, you know, ⁓ it appears that this person is interested in ⁓ people and airplanes.

Jeremy: Mm-hmm. Okay. How often do you have that thought that that was the first thing you came up with for this analogy?

Jason Haworth: You mean today?

Jeremy: All right, well, ⁓ that's it. We got we got to all of it. I can't believe it. But the tech the tech industry keeps telling us a digital story, clean interfaces and invisible infrastructure. But this week's news tells a different one. the cloud has a loading dock, a heat problem, and a welder shortage. The physical and digital aren't competing, they're converging at a scale the existing infrastructure wasn't built to handle. The people best positioned for the next decade aren't the ones who are who have mastered the tools. They're the ones who understand the systems those tools depend on, whether that's a cooling pipe or a neural firing pattern. Act accordingly.

Jason Haworth: Yeah. And don't anticipate that it's gonna be that way for very long. Like the ma the machines are gonna make robots and they're gonna take make all those jobs obsolete also.

Jeremy: Right. I for one welcome the universal basic income and no actual job to just do whatever the hell I want to do with my time every day. I'm all for it. Just you know, as long as there's still a planet to live on for a while, I guess.

Jason Haworth: Yeah, it'd be great. Yeah. Yeah, as long as they run me down effectively and cleanly or give me a good brain interface to let me go and explore my wildest fantasies, that'd awesome. Exactly.

Jeremy: That's it. I'm in. Sign me up. All right. Thanks so much for listening. If this has been beneficial or interesting or relevant to you in any way, please feel free to share it with anybody who may feel the same way, even if they're a robot. You can find the links to do that at robots.me and that's where we'll be back on Monday morning with another episode. Thanks so much for listening.

Jason Haworth: Thanks. Thanks everyone. Bye-bye.