Coffee With Scott Adams — Knowledge Archive August 24, 2026
Scott Adams Philosophy Archive
Search ideas
Episodes Episode #2681 Segments
MainContent Media & Fake News

Back to episode — Episode 2681 CWSA 12/06/24

Context —

itting there's some kind of other kind of life force here. Maybe. Well according to The Independent there's this new scientific breakthrough in China where they made camouflage material that will instantly change color to match your background, sort of like animals do, chameleons. So if you wear these clothes you will just sort of disappear into the background. Now since you don't wear clothes on…

← Previous segment →

l like you really had a lot of evidence that an adversary did it. So I would say beware of the documentary effect if you have a subcommittee that's taking testimony that all kind of leads in one direction.

All right. Meta according to Breitbart News, Meta is going big on nuclear energy to fuel their AI. So what's he going to do? So I guess all the big companies figure they've got to have some kind of connection to the nuclear power industry. This is just the best thing. It turns out that the most important part of AI was to force countries like America to take nuclear power seriously for the first time. I feel like that might be the enduring legacy. Like the most important part of AI is not even the AI. That it got us to completely revamp our whole understanding of our energy situation. Because as many people have pointed out if you're making a lot of money from energy or you're doing your energy stuff right, your country does well, you win wars, your economy does well. It all kind of comes from energy. Can you think of any country that produces a lot of energy and isn't better off for it? I mean energy is kind of the whole game these days. So I think this is the biggest thing that's happening actually.

Well according to X user Terra, I saw this story which is a funny name. First name is Terra last name is Bull but I think she does it humorously like terrible. I don't know. I don't think it's a real name. It'd be funny if it were anyway. So Elon Musk's company the Boring Company, they've got these big boring machines that make tunnels easily. They got approved to do tunnels in Vegas so they'll connect the Vegas airport to the downtown. And I got to tell you I don't know if I could go into one of those tunnels. To me that's scary as hell. I mean it was hard enough to convince myself to use BART, you know the mass transit in the Bay Area that goes under the bay. It's kind of scary to be under the bay. Now in reality you're probably safer there than above ground right? Because if you get deep enough down even the earthquakes don't bother you. You know basically you're immune from just about everything on the surface. So you're probably safer in one of these tunnels than you are anywhere else on Earth. But man the psychology of it. I don't know how long I can stay in a little tunnel that feels like it's only about the size of the car. I don't know. You'd have to give me some kind of tranquilizer.

According to Shyam Sankar on X there are more data centers in the US than in all other countries combined. Did you know that? And it's not even close. If you look at all the data centers in the US it's like this gigantic bar and all the other countries are just those little dots. How in the world can other countries ever compete with us in that scenario? Because it seems like ability to quickly build gig data centers is what's going to drive AI and that's what's going to drive robotics and that's what's going to drive basically everything in the future. So if we have like a 50 to 1 advantage or something over the next best country that's doing data centers that feels really big. So we'll see if that makes a difference.

However I want to be careful about this next point. I do not give financial investment advice. Now having told you I do not give financial investment advice I'm going to say something that sounds a whole lot like financial investment advice but it's an observation. Trust me you don't want to invest based on what I say. Honestly you don't want to because I'm mostly guessing and when it comes to individual companies that's a bigger guess than most things.

So here's my feeling personally. I sold my Nvidia stock which had already gone up quite a bit. Now the reason I got rid of it is not because I think it's necessarily going to go to zero tomorrow but because I think it's so vulnerable to disruption. Because the cost of building one of these Nvidia driven data centers is just billions of dollars and it just seemed to me that somebody's going to figure out a way to do that way more cheaply. Like somebody's going to find a way to compress AI into some more efficient form and then the entire business model might be destroyed.

So AI has this weird quality where on one hand it's the future and it's everything. So if you invest in the biggest company that's doing the most important thing for the most dynamic thing you're probably in good shape right? Which is why I bought Nvidia stock in the first place. It was almost like a, well not almost, it was a monopoly. So one of the best pieces of investment advice I ever got was from somebody famous who I'm not going to tell you. It's somebody who's in the news quite a bit actually but once told me privately that he only invested in monopolies. It's just that they don't look like it. Like Facebook is sort of a monopoly. Uber is close to being in a monopoly at least for a while. So I thought Nvidia looked like one of those. It's why I also used to have stock in Apple which by the way if you'd taken my advice and sold it when I did you would have lost a lot of money because it kept going up.

But both Apple and AI have the weird quality that things could change quickly and I don't like to be in an investment where something could change completely like on a dime. For example you think the iPhone is unbeatable and then next thing you know Elon Musk says that Starlink will be able to provide phone service. Now that's just the background phone service not the device but how far away are we from AI becoming the interface for a generic phone that somebody like Elon Musk could manufacture

Context —

and just replace the entire smartphone industry with some kind of AI phone that works on satellites? I feel like that could happen kind of quickly. So there's some new news on the large language models. Apparently somebody did in fact come up with a new algorithm dubbed calibration aware low precision decomposition. Okay doesn't matter what it's called but somebody found a way to massively change…

Next segment → →