TWiT 1100: Yoloville - Data Centers, Climate Change, and the True Cost of "Innovation"

[00:00:00] It's time for TWiT This Week in Tech. Great panel! Oh my gosh, Stacey Higginbotham is back. Jeff Jarvis is in the house. And it's the return of Owen J.J. Stone. We're going to talk about Google's big day in court. Snorting nanoparticles, is that the best way to improve your brains? And yes, we'll laugh at more robot videos. TWiT is next.

[00:00:21] This episode is brought to you by OutSystems, the leading agentic systems platform. OutSystems is helping their customers modernize operations by enabling them to build, modernize and operate enterprise systems, starting from any coding tool in a governed agentic engineering model. With OutSystems, you can coordinate and govern your entire agentic workforce within your enterprise ecosystem and accelerate impact with industry proven agentic solutions that are pre-built, governed, and managed to operate.

[00:00:50] And customizable to your business. It's time to innovate at the speed of AI without compromising quality or control, which is why thousands of enterprises worldwide trust OutSystems for their mission critical apps. Teams of any size and technical depth can use OutSystems to build, deploy, and manage AI apps and agents quickly and cost effectively without compromising reliability and security. Without systems, you can accelerate ideas from concept to completion. It's the

[00:01:20] leading agentic systems platform that's a leading agentic systems platform that's unified, agile, and enterprise proven, allowing you to build your agentic future with AI solutions deeply integrated into your architecture. OutSystems, built your agentic future. Learn more at outsystems.com slash twit. That's outsystems.com slash twit. Podcasts you love. From people you trust. This is Twit.

[00:01:54] This is Twit. This Week at Tech. Episode 1100. Recorded Sunday, September 6th, 2026. Yoloville. It's time for Twit. This Week at Tech. The show we cover the week's tech news. We're going to spend the first hour introducing our panel because it's some of my favorite favorite people. Stacey Higginbotham is here from Consumer Reports. Hello, Stacey, with her stellar J or whatever they call that now.

[00:02:21] The blue crested super warbler finch or whatever it is. Hello, Stacey. Hello. You're doing great on the bird names. I really appreciate it. I thought we had blue J's, but they were scrub J's. So I don't, I no longer know. I just no longer know. So Stacey also writes a bunch of stuff for other people, I think. Right? I see your, did we have you on after Om passed? Because you wrote a beautiful piece about Om Malik.

[00:02:50] You didn't. Yeah. But that's okay. I think we wanted to. If wishes were fishes, people would ride horses. Okay. Yeah. Hello, Stacey. Also with us, ladies and gentlemen, from Stacey's former co-worker from formerly This Week in Google Now Intelligent Machines. I give you Jeffrey Jarvis. Hello, Jeff. Hello there. New book. New book. Hot type. Hot type.

[00:03:18] And if you go to, this is new, if you go to jeffjarvis.com, you can order an autographed copy. Oh, nice. Arrange that. You can order from a local bookstore, the Monumentary Book Center. And, uh, okay. Did you use auto pen? Be honest. No, no. Okay. I had nothing to do with that. Nobody wanted my autograph. So I would actually go to the bookstore and sign some books and then leave. Uh, so there you go. Nice to see you, Jeff.

[00:03:46] I see everyone on Intelligent Machines, but there's always more to talk about. We never have enough time. We got a little AI to talk about today. So a little bit, a little bit. But, and from the home of the Philadelphia Eagles, fly Owen fly on JJ Stone. Yay. Every time I see Jeff Jarvis's background, it annoys me that I don't read. I feel like his, his setup is judging me. It is. Look at all the books. But Stacey too. Yeah. Stacey, Stacey's added books to her. These people are readers.

[00:04:16] Stacey is, is a notch above because she's not guilty of me. She's got a bird up there. She's got some art decor. Or Jeff's just got books on, there's no, there's nothing of joy in there. There's no. That's joy. Oh, is there joy? Look at the foliage. Look at what? The foliage at the top. There's no foliage. Oh, there's some foliage. Is that real though? Is that, are those plants real, Jeff? No, you can't ever water those plants. All that is, what is that, a rat up there? What is that up on the other shelf? What is that? What is that? That's an armadillo. It looks like some animal. What is that?

[00:04:46] Over your right, your left shoulder there, Jeff. What is that? Actually, you don't want that. What is that, sir? That's the ashes of our last cat. Oh, wow. Since you asked. Are they in an urn or that you sit there? They're in a box. Okay. Oh, I have the ashes of my last pet behind me. And they are also in a box. I am losing out in this. I don't have any books. I don't have any ashes. I've got to get my life together, Uncle Leo. You as well, sir. You as well.

[00:05:16] You, this whole digital thing you got going on. I got lots of books, but they're not behind me. It's all, yeah, because we're trying to show the computer-y thing. Oh, wait a minute. Oh, you're going to- I got to flip the switch on the Altair. Hold on a second. Hold on. Ready? Run. Oh, now it's running the program. Look at that. Analog. You still got buttons to push. I love that. Beep, beep, beep, beep, beep. Yep. None of that is functioning.

[00:05:43] And the other side of the room is all the AI machines, which, to be honest, aren't functioning all that well either. I had a little- So they're servers and you're hot right now? I mean- I don't know. They're just, I don't know. Leo scolded his alter egos this morning. So twice now, I have fired my AI agents. This time I brought them back. You can't fire us. We quit. Sometimes they just dither so much.

[00:06:14] In this case, I was trying the new thing from OpenAI. OpenAI announced a new model, which some say is so smart, mostly OpenAI say, is so smart that you should be afraid to be very afraid. This, we think, is related to the model, if not the same model, but probably related to the model that hacked hugging face. OpenAI says it's been properly scolded now.

[00:06:41] And I would agree because it's so cautious that nothing got done. I gave it on Friday. I said, look, it's a Labor Day weekend. I'm taking the weekend off. You can go to work. I want you to find the perfect model for my hardware and test it all. You can install it all. You got a complete control. I said to Anthropics Fable. These are like the big brainiacs of the AI world. And to the brand new OpenAI Astra said, you got it.

[00:07:11] You're in charge. I said, Astra, you plan it. Fable, you execute. I'll see you on Monday. It's the holiday weekend, sir. You know what? That's what they said. Yeah. And if they're on Monday, you ask for it to be done on Monday, on Labor Day, is embarrassing for you. The computer knew better. They knew better. Last night, Sunday, midnight, Sunday morning, last night, they stopped.

[00:07:40] They said, we can't do anything until you wake up. So I got up and they're just sitting there. They're just sitting there, not doing nothing. If we work, you work. That's how it works. I said, what are you waiting for? I said, I told you explicitly, don't wait. Don't ask me. Just do it. But they still stopped because they're very cautious. They're taught by their real masters, not me, the people who pay the electric bills, that they should be cautious.

[00:08:09] So what did you do then? I yelled at them. I said, did I not give you explicit instructions not to stop, to continue? This is what I was told. The whole thing about Astra is it's amazingly persistent. So it was obsequious as these models are then? Oh, I'm so sorry. You're so right. Oh, yeah. Oh, you're right. I said, did you not remember that I told you? Oh, oh, it literally said, I'm sorry. But it's not. It doesn't know I'm sorry.

[00:08:39] I mean, it might be sorry. You never know. I saw somebody on the socials today said that I think it was Claude at some point just said, I'm done with this and cut it off and killed the conversation. Fired. Yeah, I've seen that, too. I saw that on X. I haven't ever had it do that, but there is definitely refusal built into these. You know, cybersecurity refusal. They won't let you make bioweapons. I wasn't trying to make a bioweapon. I was just trying. What were you trying to do?

[00:09:05] I asked it to find the best local models I could run on the existing hardware and on the hardware I have. And they were downloading them and trying them. I had Grok on what I call model watch. Grok, by the way, doesn't. Grok, which is Elon Musk's alternative, never says no to anything as we know. So Grok was on the model watch.

[00:09:30] I said, Grok, you watch because you're from XAI, you can watch X because that's where all the people put their, you know, oh, it's so good. And this is the other thing. X is filled with influencers who say, oh, I've got 3,000 tokens per second coming out of my machines now. It's amazing. And I fall for it every freaking time. They don't, by the way, they never say how good, you know, how smart they are.

[00:09:58] They just say how fast they are, which should be the tell. So once you stole them, did they go back to work? No, I fired them. What happened? I said, that's it. It's done. I'm going to use this model. Done. I'm going to install it. I actually told Grok, I turned those guys off. You install this model and we'll be done. But now maybe I'm regretting it because. Because you had anything to do with Grok and Elon Musk. Now it's now it's just it's stuck.

[00:10:27] The whole thing. It's can I touch grass? I'm giving up. What is our budget right now on AI? I hate to think what I've. Well, I haven't spent because it was a it was subscription limits, right? No, no. It's a there is a limit because they will only do so much in a five hour period a week and in a month. But I didn't hit any of those limits. The usage limits. They just they're just cautious.

[00:10:53] They said, well, I didn't want to install this model until you said it was OK. It's like, dude, that was exactly what I told you to do. Are you sandboxing the PCs? You know, I don't do any of that. No, I'm not that. You're so sane, Stacey. I live in YOLO. I'm in YOLOville. I, in fact, I have at least on a separate network. No, they're all on my main network. I have Stacey. He's using this to build the future of the business. Oh, yeah.

[00:11:22] The ad sales. But that's on the back burner for now. But his wife just took over. Yeah. Yeah. She said it was September 1st. She said, OK, we're taking back sales. Where's the sales system? I said, what do you mean? You want it now? I wish you'd give me some notice. I'm busy testing models here in YOLOville. So did you deliver? Wait, wait, wait. Did you deliver the system to? Yeah, it's a working. It's sort of mostly working. Yeah. It's pretty good. She has some complaints.

[00:11:53] You're hearing customer complaints? No, no. The customers. It's not customer facing. It's internal. No, I mean, your customer is your wife. Oh, yeah. Her. Well, that customer. I might argue one of the most important customers. Well, she's certainly what I call a stakeholder. I just got the stake holding over my heart. I was going to say, does she have the stake right over you? Like Buffy the Vampire. So I now, and then, well, no, I live in YOLO mode because, so, OK, so I set up BARD,

[00:12:22] right, on anti-gravity. Not BARD. It's a Gemini, whatever, the new Gemini 3.1 flash on another machine. And it kept saying, well, do you want me to edit this file? I said, yes. Do you want me to edit this file? I said, yes. Do you want me to always edit this file? I said, yes. So I finally, I said, do you have a YOLO mode? And then it said, OK, let me check. Is it OK if I check? It's like, oh, my God. So finally, after hitting return literally 30 or 40 times, it said, yeah, you start me up

[00:12:52] with dangerously skipped permissions. I said, oh, I wish you'd told me that at first. So now it's not asking me anymore. That's so that's the answer to your question, Stacey. They're all that sounds foreboding. Dangerously skipped permissions. Yeah, that's that's what anthropic. I think I think that's how open AI got in trouble with skip permissions. Stacey, you're muted. So you're basically lighting some firecrackers and dry brush. Just what could possibly go wrong? I'm not.

[00:13:21] So this is this is actually the topic. One of the big topics. Before we get to that, though, I think this big story is that the court has decided that Google does not have to break up that this is Judge Brinkma who had decided that Google was a monopoly. No question about it with their ad. And Jeff, we've talked about this. This look, Stacey, you should feel right at home. Here we are right back at this week at Google. Mm hmm.

[00:13:51] Uh, clearly Google is the worst kind of monopoly. They own ad sales, ad buying. They own everything from beginning to end. The judge said that. But. Even though the DOJ wanted to ask the judge to for Google to to sell their ad division off, the judge said, no. I'm going to start basically one. Yeah.

[00:14:21] This is like what we're doing with climate change. It's kind of like the most American thing you can do, which is like, oh, my God, this is a huge problem. But man, if we tried to solve it or tackle it, it would cost a lot. It's a mess. We can't figure it out. Yeah. This is the second trial. Google was the FTC was trying to get Google to remember. They wanted them to sell Chrome. They wanted them to sell Android. They wanted to do something. The judge also said, no, you know what? The cure would be worse than the disease. No one ever.

[00:14:50] I think I mean, I didn't read the court documents efficiently, but I don't think no one ever made the case for this is exactly why we split them up. This is what the benefit we're going to get. This is how it's going to operate. Then I think it was more retribution. Just split them up. Hurt them, please. They're too big. Steve Lohr writing on The New York Times. He covers the antitrust and tech beat for them. It says in his byline since the Microsoft trial just 20 years ago. So poor guy is really suffering at this point.

[00:15:19] Why the courts struggle to tame big tech. And I think he's right. He says they fear meddling in markets where the technology is moving so much faster than the court can move. The earlier case, the judge said Google search is a monopoly. Sure. But it's already losing ground to AI. So what do you don't have to do anything about it? Right.

[00:15:49] And now the judge has decided, well, we can't break up Google. They, you know, the judge Brinkman said, oh, you know, you have to what? Share some information with your competitors. How the auction works. Kind of. We haven't really seen yet what the rules are going to be for Google. Right. Okay. So that's the next stage is they're going to have a penalty phase, I guess you would call it.

[00:16:16] All this stuff is such pure theater at this point. It's so annoying. It feels like it. Because you allow these companies to acquire other companies at alarming rates. And they get checked by the FTC. And we've got to check this out. No, you don't. Because you keep letting people buy up everything. And then, oh, wow. They've got lobbyists. So it's like, all right, well, we'll just keep throwing money at the problem. And now you can't break us up because everything's intertwined. And you don't know how emails work. If you take this out, you won't get emails.

[00:16:46] It's true. It's heart surgery. And they want to take out the heart. It's not like it was back in the day when they would break up companies. It was like, all right, well, you make the parts. You build the car. You put the gas in. You build the railroad. And then it's separate. It's like, you don't understand technology. I can't take this out without taking that out. Even though you acquired those parts individually and brought them in, you could never separate them again. The DOJ said that Google had rigged ad auctions to give itself an advantage.

[00:17:16] The court agreed that Google illegally locked publishers into using its exchange. But it said, they didn't really break any laws. In the remedy phase, the DOJ said you should force Google to sell the ad exchange. That will level the playing field. It's not a huge part of Google's revenues. And honestly, I think the issue the judge saw was, well, you sell the ad exchange. You're just creating another mess.

[00:17:46] With some other company, right? Who's going to buy it? Oracle buys it. And now Oracle's got, you know, the same power. While divesting the ad exchange would have been, this is ours, technical writing, would have been the most serious penalty. The government also asked for fines and court-ordered changes to Google's business practices. It's likely DOJ will get a lot of that. Judge Leone Brinkman sealed the order. Oh, so she's decided. But she sealed the order for 14 days so that the parties could request redactions. That's right.

[00:18:16] So in two weeks, we'll know. So here's a question for you all. In hindsight, did the breakup of Bob Bell work? I think it did in the interim. Actually, for a hot second. Okay. For a hot second, parts of it worked. So when they did. Well, no, I mean, this is exactly what we would expect. What happens is the company, it breaks them up.

[00:18:39] And then they set in regulations to require, like, you know, you had to, if you had the only lines on the ground, you had to let people access your central offices. And I can't even remember what these are called now. It's been so long. But they had to use basically your equipment if it was the only thing accessible to them. But then what happened is the Bells were like, well, this obviously sucks.

[00:19:03] When someone needs to, like, fix their equipment or fix our equipment that they're using, we're going to drag our feet. We're going to do malicious compliance. And, I mean, this is exactly what you would expect from people. And so that's what we would see here is malicious compliance or, like, basically malicious compliance or instituting new laws to get around the existing laws. And then you would have to come back.

[00:19:30] The problem is voters and people and regulators want these problems to be solved. They don't realize that it's just a continual game of cat and mouse because. Or a shell game, right? I mean, it's not a shell game necessarily. It's just you're going to you're you're playing with people who have they're constantly at odds. And you just have to recognize that fact and realize you're going to do something. And then they're going to do something to respond to you. Right. It's the same way we deal with the game.

[00:20:00] Or AI or anything. Yeah. Toddlers. But, I mean, this is a court. So it's not, you know, it's not influenced by the electorate in theory. Right. It's it's not influenced by by lobbyists. They don't have to raise money to run for office. They should be. The whole idea of the court system is that they're kind of justice is blind with the scale thing, you know. But again, they don't have an answer on how to. That's the problem, isn't it? There's no there's no reasonable remedy.

[00:20:30] Yes, it's not. And so far out of their realm. Do you think Google made it that way? Alphabet made it that way? Like, what if we just weave this? I mean, they saw what happened to Microsoft, which almost got broken up. Microsoft had to make a consent deal with the DOJ. You know, they knew what happened to AT&T with Ma Bell. Maybe they said, what if we so interweave our businesses? That happened when they bought DoubleClick. Right. There's no person said.

[00:20:58] Well, and then there's also, I mean, the courts, they enforce laws. And our current system of laws is not it is not enough to deal with what's happening with technology. We don't have the right laws in place still to address technological harms or the harms of deeply integrated technology. We don't have like the, what is it, the Heishman-Herfman Index for detecting competition.

[00:21:27] I can't think of it. It's a formula like the DOJ uses for anti-competition for access to the market. We don't have something like this to understand how much market the tech companies control. Oh, you're right. Their impact, too. I was just writing a post for tomorrow for Labor Day, the labor story out of Hot Type. Available now, jeffcharvis.com. And always be selling.

[00:21:54] And I was really impressed that at the turn of the last century, there were federal commissions to understand and study in depth the impact of technology on labor and the economy and working conditions and so on. And companies testified, labor testified, huge amounts of data came out of it, really, really very useful. And you'd think that we'd be doing that kind of study now about this huge change we're undergoing now in technologies, but no.

[00:22:26] Oh, well, you know, I think it is a fantasy that some of us have that there could be fairness in the marketplace and that capitalism works fine if you have antitrust regulation. And, you know, I think that's not that's not a fantasy. The problem is we don't have a system implemented to address it quickly enough. Like, you cannot you cannot punish wrongdoing eight to 10 years after you've seen it be harmful. I mean, you can, but it's it's not very effective. Oh, you need.

[00:22:57] Yeah, you need to act quicker. You're saying. And we don't have the mechanisms. I don't think we have the will. Or the education or the regular process from the beginning. Not for nothing. Like we need term limits. You've got people that are in office that are supposed to make laws and regulate bills and they can't even send an email on their own. They've got TAs doing all of their stuff on the Internet, writing their tweets for them. And when they go sit down in front of Congress, they're like, what do you mean I can just send any? What do you mean you could block things?

[00:23:25] Like they're asking the most asinine questions about the Internet. And I'm like, no offense, Uncle Leo. They're your age without your education level. They're older than me. I mean, the first requirement should be hold up your phone and show me how big your font is. If your font is. Well, I don't know if term limits are the answer. Those are people who just don't have good eyesight. Hold on. My font is huge. You mean this big? Oh, doctor, you mean that big? That big?

[00:23:55] Is that what you're saying? That big? That's so Stacy here. My eyes don't work so well, doctor. Oh, it's 2.39. I would be blind with you too if I didn't buy LASIK for myself. Best participant. Oh, there you go. No, I'm not sure it's term limits. I think it might more be get money out of politics because I think you're right. The remedy probably has to be in Congress. It has to be legislative first. And then it can be through the courts. You've got to make the laws first. And you have to act more quickly. Congress doesn't act at all.

[00:24:25] And I think they're also beholden to the people who have money because that's what it costs to get elected. If we got money out of politics, I wouldn't mind if somebody, if you could stay in office without cheating. You know, voters historically love their member of Congress, right? Yeah. I would hope that not all of them are evil reptilian beings. But I do know for a fact that, again, I've listened to a lot of things on C-SPAN.

[00:24:53] Even if they just don't understand what they're talking about, they don't understand the technology. So that's what I say about term limits. I'm like, you've got certain people in there. Well, you need younger people, you're saying. Well, hold on. You keep talking about fax machines. I'm like, bro, we ain't sending fax in like 20 years. Why are we still talking about fax machines when we're talking about Facebook right now? Like, what are we doing? I don't think it's just age. And I would say it is impossible to understand the level of things that a legislator needs to understand at the level they need to make effective policy. That's why they have staff.

[00:25:22] Which is why they have staff and they have people like. And you deal with this. This is part of your job. I mean, you deal with policy. Yeah, like I'm now a registered lobbyist in a few states, which is bonkers to me. So talk about this, Stacey, in terms of what it's like to deal with policy level. I'm curious what your experience, now that you're in it for quite some time, your experience has been. I mean, the people that I've worked with, for the most part, you work with the staffers. Like we have, for example, we have model legislation that we are pushing around connected devices.

[00:25:51] It's like trying to get people to disclose how long they plan to support a connected device when you buy it. Right? So that's the goal. And I pitch this to people. And if they're interested, they come talk to me, but they've never heard about this issue. And then I will, we go through meetings where we talk about why the issue. I provide these like long letters with like cited reports, you know, just so much data. It's like little research reports for them. They actually read through all of it. They come back.

[00:26:21] They ask really insightful questions. We meet with, they meet with the opposition. Those people send letters with sources, like their sources, right? And documents. And then the person comes back to me and they're like, well, they say this. And I'm like, okay, but here's what they're not telling you. And it's a really, it's a really involved process.

[00:26:41] And if they're doing this for like this tiny little bill that I have, can you imagine trying to reform like healthcare or deal with like anti-competition in technology? I mean, saying that these people, you know, still believe in fax machines is ignoring the fact that what we're really talking about is like academics spend their whole lives researching like one corner of this.

[00:27:08] So, I mean, I have a lot more respect for the process, I guess, and how much information and insight that these people have to digest before they can actually make reasonable policy. And that's assuming they even want to, which is really a big assumption. I'm having this fantasy though, Stacey, you said you're a registered lobbyist. I'm seeing you with a big stogie and a huge briefcase stuffed with cash in a steakhouse. I work for Consumer Reports.

[00:27:37] Darren Oake sees you as an advocate for truth, justice, and the American way in the court of twit law. So, I guess I get to be the judge. Owen, are you the attorney for the defense? I know. That pose seems very apropos of me. I'm questioning everything. I trust nothing. Yeah, that's good. What's going on here? I got to find him. That's good. Well, I think also your point of view is very common. I mean, this is, I think, with a lot of people. People, this is funny.

[00:28:06] Somebody in the Discord said, everybody hates Congress, but for the most part, everybody loves their member of Congress, which is the contradiction that we are facing. You know, once you get elected, you're pretty much there for life. No, no, no. It makes sense. Leo, they hate everybody else's congressman. That's right. Well, it's like media. That makes sense? That's Benito, by the way. That is not the voice of God. Although, and sometimes I think he might be. Benito, our producer. Yeah, it's true.

[00:28:35] They hate everybody else's member of Congress. Well, it's like you're the media outlet that you trust, but you hate media as a whole. But also, I think we don't like the process very much. And I think it's really good. I think it's very good for you, Stacey, to describe it. It says you're inside it. And I think that that's reassuring. Well, and it's, I mean, y'all, and this does not, I mean, again, I have only worked on this one issue in the States.

[00:29:00] I haven't done federal stuff, but we spend a lot of time on like really small minutia and trying to like establish like facts and patterns and case like examples of things. And like, it is a lot of work. It is like every research paper that you've ever had to write. And then basically your congressperson has to read all of these research papers, decide where they want to stand, and then compromise from there, basically.

[00:29:30] And it's- Compromise. Nobody likes compromise, right? That's part of the problem. It's the only way anything gets done. But that's politics. I love compromise. And I was actually thinking- It's an adult point of view. It's so hard because we live in our own filter bubbles now. Right. We don't actually see people as people. And so you don't, if you don't know anybody who's been affected by something, like it's really easy for me to like talk about like IoT issues because I've seen it from both the manufacturer and the consumer, right?

[00:29:58] But if you've never seen it from someone else's point of view, you really don't understand when someone says, well, the trade-off is this. You're like, well, I poo-poo their trade-off. It's stupid. Even though it might be really relevant. I don't know. I'm probably too rational for all of this. Yeah, I think you definitely. It's the last word anybody would use about our system. Stacy, if you could turn up a little bit, you're very quiet and I want to hear every word that you have to say. I'll kiss the mic. It's me and my- Yell at it. Yell at it.

[00:30:28] Yeah, I don't. My gain is all the way up. I'm so sorry. Too calm. We're going to take a break anyway. Don't apologize. You're super soothing is what it is. That's why. Just soothing. That's what it is. Relaxing. And now we're going to meditate. I'm telling you about fax machines. You're not trying to sound razzling. I get it. I get it. Well, anyway, Google kind of, it looks like skated a little bit. We'll see what the final decision is from Judge Brinkman. But I sympathize with the judges in these cases.

[00:30:58] It's hard to find a remedy at this point. And I think you're right, Stacy. The problem is you have to get it earlier in the process. You can't wait till now to solve it. It's too late. Yep. We'll have more in just a bit with Stacy Higginbotham. Great to have you from Consumer Reports, where she is a policy fellow. What? A jolly good fellow himself, Owen J.J. Stone. Oh, doctor. It's great to see you from... You're still at IQMZ.com.

[00:31:26] Do you still do the wonderful podcast with your daughter? Or she's in school now? You can't. She's in school now. She's actually allowed me to get back on the internet. So she's going to come back on the internet with me. Oh, good. Cool. Don't. Leia's great. Hanging out. Oh, she did say hello, by the way. She was going by today, but she's... She... Owen brought her the last time he was out here. And we all went to dinner. And Lisa then took her jewelry shopping, which I'm so sorry about. Next summer, she might have to come out there and intern with you just so she can hang out with Lisa and not work with the Yachtbox.

[00:31:55] I can't promise she'll do any work, but... That's okay. We don't have anywhere to put her. So she can sit in a Lazy Boy if she wants. That's about it. Also, of course, from our wonderful Intelligent Machine show, our Intelligent Professor, I didn't get to do the Craig Newmark jingle. No, I think... Is Benito trying to suck up to you? No, I just didn't say the Emeritus Professor of Journalistic Innovation at the... I got in with plugging my book. Craig Newmark Graduate School of Journalism. Craig...

[00:32:25] Sorry, I don't have it loaded up into it. It's a different show profile. He can't play it. Yay. At the City University of New York. He is actually now doing stuff for Montclair State University. Stacy, you'll be glad to know that on the most recent Intelligent Machines, we resurrected the Google changelog. We did. We played the drum and everything. Oh, dude. We hated that. I know. That's why I did it to Irritate Leo. Before the show, I said Benito, go find it, please. Please.

[00:32:55] But you couldn't do an AI changelog. It's the whole show. It's like every day there's like 20 new stories. It's kind of an amazing thing that's Intelligent Machines. We also announced, by the way, I know people hate all the AI coverage, but honestly, this is the thing that is happening right now in technology. I mean, we'll talk about the iPhone on Wednesday. Sure. But really, the AI is where all the action is. And so we have added, Micah Sargent is going to start doing Hands-On AI, a short form show

[00:33:25] where he'll kind of help people who are getting started with it, understand the concepts, the terms you're hearing, how to use it more effectively, how not to use it. That's great. That's going to debut in a week or two, Hands-On AI with Micah Sargent. If you're already subscribed to the Hands-On Apple feed, I think we're just going to piggyback on that because it's the same initials. So we'll just keep using it. Oh, and Jammer B, thank you. Jammer B has donated something to the cause.

[00:33:54] And of course, we can always count on him. Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Craig, Newmont. I don't know if Craig listens to this show, but if he does, he needs to have that. We will have more in just a little bit. Our show today brought to you by, I would show it to you, but Stacey, you were saying I live

[00:34:22] in YOLO land and it's true, but I do have the Thinks to Canary, our sponsor, hooked up onto the network. So at least I'll know if somebody breaks in. I have a question for all you people, all you business people about your network. How do you know? I know you have the best defenses. Everybody does nowadays, but people still break in, don't they? How do you know if somebody, an intruder is inside your network? How would you know that?

[00:34:51] I'll let you think about it. You probably wouldn't. On average, companies don't know they've been breached for 91 days, three months for an intruder. And by the way, an intruder doesn't have to be a human. Nowadays, we know that intruders can be even AIs. They just don't know that they're in the network. Unless you have a Thinks Canary. Thinks Canaries are honeypots. They are irresistible, both to AIs. By the way, I could make my Thinks Canary be a wiki. Come to think of it, maybe I should.

[00:35:20] Both to AIs and to human hackers and malicious insiders. They see it, and it looks like something real, a Windows server, a Linux box, a SharePoint exchange server, a SCADA device. It could be anything. There are hundreds of profiles. And by the way, it impersonates them perfect right down to the Mac address. It looks exactly like what it pretends to be, but it's not. It's not. You can also use your Thinks Canary to create what we call Canary tokens.

[00:35:50] Files you can put on not only your local on-prem drives, but your cloud drives too. I have, I probably shouldn't tell you this, but if I have on my Google Drive files that say, that look like spreadsheet files that say things like payroll information. Now, if you're a malicious bad guy and you're in my Google Drive, how could you resist that? That's probably loaded up with social security numbers, home addresses, all sorts of good information.

[00:36:16] But the minute you try to open that file, or the minute you try to brute force my fake SSH server, I'm going to get an alert from my Thinks Canary. It's going to tell me. And you only get the alerts that matter. No false alerts, just the alerts that say someone is in your system and you have a problem. In any way you want to SMS text, you can get email, syslog, Slack, whatever you want. Web hooks, they have an API even. You can write your own if you want.

[00:36:44] Choose a profile for your Thinks Canary device, register it with the hosted console for monitoring notifications, and then just relax. Anybody who's breached your network will make themselves known. They can't help it by accessing your Thinks Canary or those Canary token files. Visit canary.tools.twit. $7,500 a year you can get five of them. You also get your own hosted console. You get upgrades, you get support, you get maintenance, all of that. And if you use the code TWIT in the how did you hear about us box, you're going to get 10% off the price per life.

[00:37:14] You can always return your Thinks Canaries. They have a two-month, 60-day money-back guarantee for a full refund. I should tell you, though, they've been advertising on this network for more than 10 years. And in all that time, the Thinks Canary refund has not once, not once, been claimed. Visit canary.tools.twit, enter the code TWIT in the how did you hear about us box, not just for 10% off for the first year, but for every year you own your Thinks Canaries. So that's forever if you want. It should be.

[00:37:43] Canary.tools.twit. I love my Thinks Canary. I think you will too. We thank him so much for supporting. This week in tech. Right. Right. Stacy's going to talk directly into the microphone so I can hear it. Is this better? Yeah. The Herfindahl-Hirschman Index, you say. Or HHI score. That's probably easier. Yeah. I could say that three times fast.

[00:38:13] Stacy put that in the Club Twit Discord. Sorry. I am such a nerd and I wanted to make sure that we had, like, in case anyone was like, oh, I really love thinking about antitrust law. What's the name of that? Is it an algorithm? It's an index. We now know the invitations went out. Everybody got them except me. Apple's having an event and they're going to be doing it in Cupertino with a brand new CEO this Wednesday. Tim Cook stepped down on September.

[00:38:42] Actually, August 31st and September 1st. John Ternus became the CEO of Apple. This will now be his first Apple event. Not too much pressure. All he has to really do is walk out on that stage and say, good morning. One more thing. Watch the video. They don't even do one more thing anymore. They just play a bunch of videos. They've probably been done a long time ago.

[00:39:09] We don't know for sure what Apple's going to announce, but it's pretty clear they will announce a new iPhone. iPhone. That's what they do this time of year. The iPhone 18. Now, they won't do the 18 or the 18 SE or even the new iPhone Air 2. The rumor mill, which is usually pretty right on by this time of year, says that'll be next year in the spring.

[00:39:32] But they will announce the 18 Pro, the Pro Max, and a new folding iPhone that will not look like a Samsung Galaxy Ultra, whatever, but more like a little book. In fact, actually, Samsung, because they're making the screens for Apple. How's that product? They kind of said, oh, we can do that. And they did it before Apple did.

[00:40:01] It's very popular, apparently. Stacy, do you still use a folding phone? I can't use a folding phone. I just want a tiny little phone. That's all I want. You want it to fold. It is tiny. It folds tiny. That way, though. You want it to fold that way, not that way. You know what I'm saying? The book fold, as opposed to the razor fold, we'll call it the Moto razor fold. Perfect. I think I prefer the book fold myself.

[00:40:32] I think you're an Android user, though. You wouldn't get an iPhone. I am. I actually just got a new hot pink Pixel 11. Oh. Whoa. Can you show us? It's not in this room, but I can show you. How can you be more than one foot away from your phone? What the hell's wrong with you, Stacy? What? Stacy. Geez. It's not a desktop. The whole internet's in, like, two big computers. I don't care. No, Stacy. First of all, you got the new phone. It's got the little blinker on the back to tell me about emergencies. You don't have to look at it, but you need it there.

[00:41:02] What if I needed you, Stacy? You didn't have your phone at the ready? You're on this internet thing. I'm trying to text you. My life depended on it. Listen to the doctor. I had your phones in another room. During the commercial break, I expect you to ramoose. We'll bring that device. It's listening to everything you say back so he can spy on you in your room. How are we supposed to know? I'm not sure where you are. Somewhere else. This is shameful behavior. Absolutely shameful. How are we supposed to locate you? I apologize. I apologize. You sound just like my husband, who is like, I texted you two and a half hours ago.

[00:41:32] I'm like, I was doing other stuff. People get mad now if you don't respond right away. No, not if you train them. They just, they're like, oh, that Stacy is so weird. That's what they think. You're strange. Yeah. How did you put a tracking device on my vehicle? Do you have location tracking on or off on your phone? I shared my location with my kid. Yeah. Actually. Did your kid share it back? They did for a little bit and I felt so uncomfortable with it.

[00:42:01] I was like, I'm going to turn this off unless I just, I don't like it. Um, but they, they're in Spain right now, so it doesn't matter. Wow. Don't you want to know though? Like if they, they, I mean, which isn't that exactly why you want tracking is to know what's going on. Well, they're going to tell me, I mean, they'll, we'll have these conversations. Stacy is so zen. Don't let me know. I am spying. Oh my. I just checked for 13 seconds. Does Leah have her location turned on?

[00:42:31] No doctor. 100%. She can't go anywhere. Yes, she knows. Not only does she have her location. I gave her a watch too. So when the phone dies because she is a teenage girl and the phone's always dying. I still got backup tracking. I don't bother her. I don't quite. Hey, look one night she was in Brooklyn, a club somewhere. She's only 19 years old. I found out two days later, she was at some concert that her friend was performing in. I was like, okay, well, you know, you made it back home a lot, but I need to know last thing somebody is going to come to me and say, which last time you saw your

[00:43:00] daughter? I look, I check every other day randomly. Like, oh, she's a class. She was at the bar. Why are you at the bar? You can't even drink. You got to know. I don't know why. This is just that this is different parenting. I only got one. Do you ask if they're at the bar? Do you like what's what does that conversation look like? What's up with that? Are you at the bar at like three in the afternoon? So I've lived life. And when I tell her things, she used to always say to me, like we watch a movie and kids get a fake idea and they go to the bar. She's like, that's impossible. That wouldn't happen. I'm like, well, I used to get in drinking.

[00:43:30] She's like, well, you look old. Now she's at the point where she is a girl and her friend was as a rapper and he's basically performing at bars there. The rappers are 18 years old also, so they can't drink, but they can get in. So she's going with her friends and they're not allowed to drink, but they get access to these places. And I'm like, see, there's all kinds of random reasons why you would get into a bar. You know what I mean? They put a wristband on you so you can't drink, but you're getting in there because you're an artist and you're part of their performance. So there's all kinds of things going on.

[00:43:58] But I personally, I only got one kid. If I lose this one, I've raised her to be as perfect as I could be. I only got one. I can't. I don't have a backup. I don't have any. That was all my eggs in one basket. We made a real nice basket. She knows I'm spying on her. I don't say anything. That is the thing I don't do. I don't say, what are you doing? Why are you here? I don't do any of that. But I'm doing what, by the way, just hanging with his buzz.

[00:44:28] I said, did you take a private jet there? He said, no, dad, I flew commercial. Stacey, if I may ask, how old is yours now? 20. Oh, wow. OK, when should I stop? These flies. You're making me feel bad. Well, I'll be clear, though, Owen. I still pay for my kids' cell phone bills. Me too. I'm paying for it, but no. You got to let them.

[00:44:57] I mean, you got to fly. Eagles fly, right? Let the bird. Let this Stellar's J leave the nest. You got to let them make their own way. And I think you need to do it without a safety net sometimes. Like, I think it's really important for kids to go out and be truly on their own. It's hard, though, isn't it? I mean, it is hard, but just knowing where they are. I mean, basically, then if something really terrible happens, which is all I'm really worried about at this point in time, right?

[00:45:27] I'm only worried if they've been kidnapped or something's terrible. But I can't stop that if I know where it's happening. I just know where they were. You could know where to find the body. I just like information. I'm not going to stop. Like I said, I don't stop from doing anything. Like she's, like I said, randomly at two o'clock in the morning, I'm like, where are the kids at? I go check and she's somewhere. I don't text her next morning. Why were you here? I don't ask any questions. She tells me when she wants to tell me. I just like to know. The truth is that's not right.

[00:45:56] That's like the government would like to know what you're doing all the time on the Internet. Government doesn't love me or care about me as a human or person. I'm just saying your kid, your kid deserves that same level of autonomy at a certain point. I'm about to call her and ask her on the air. That's one minute. I would never do that. No, don't do that. You're making me feel bad. I need validation right now. I need someone to back me up. Is it okay if I follow you around? But that's what you need. Hold on.

[00:46:26] Think about what they need. She could say, no, dad, stop doing it. And I could look like a fool on this Internet. Or she could say. Would you stop? Yeah. Oh. Have that conversation with her in private. No, no, no, no. I like public. That's a whole podcast episode, Stacey. Jeez, it's a waste. Yeah, save that for the podcast. Uncle Leo knows. Let's go back to Apple. Oh, I'll put. No, no. That's where we are on Apple. That's the only reason I'll never leave Apple. I can track all the devices. Apple got me in a trackball. Android has that too, by the way. It ain't the same. It ain't as good.

[00:46:55] Look, I know how fast to drive with Apple, boy. You're speeding on the highway. Slow down. Yeah, it's true. So, Stacey, would you? You said the book form is okay. And that's the new thing. It opens up. It's kind of like an iPad mini, which I really like that form factor. I'm thinking I'm going to like this. I'm going to buy one. I haven't got a new iPhone. When I started the iPhone 16 Pro Max, they have given me no reason to upgrade, but I am going to get the full. Every two or three is reasonable, I think. Every year is too much. Yeah.

[00:47:24] I went from my, I had three years of my little device. I was three of the six to the 10 in Android. What? Oh, see, I'm going from eight to 11. But you know what? I got the pink line of death and I tried to return it. And for, you know, how Google's like, you can return it or they replace it for you. I went to one of their authorized, like fix it places to get a screen replaced and they refused to do it. They were like, you just have to send it to Google. We're not going to replace this. What?

[00:47:54] It's an authorized thing. He's like, yeah, but Google doesn't always pay us for it. Oh. It's not like perfect. And I don't want to take that risk. So, you know, you're going to have Google fix it. Um, I, I'm not going to do without my phone for like a three weeks or whatever. But it is a pink phone. But it's away from you now, Stacey. It matches the front and the back now match. Well, no, no, no. My, my Android phone was, my eight was green. Oh, this is old. The new one's pink.

[00:48:23] This is the old one. Is it hibiscus? Is that what? Anyway. Okay. That's the funniest thing. The name people. The new one's for these things is crazy. They're just the colors. Yeah. That was that great Android event where, oh, it's white. No, no, it's not white. It's white. White. What are you talking about? It's soothing almond. We, we spent a lot of time. We don't need to do it on this show. We spent a lot of time on Mac Break Weekly talking about Tim Cook's reign, which has just

[00:48:49] ended after he, he took the reigns, uh, when Steve Jobs passed, uh, in 2011. So he had 15 years as CEO, brought Apple from, uh, you know, under a billion, under a trillion dollars to almost $5 trillion in value, launched a bunch of important products. I think the most important product, probably Apple's own chips. Yeah. He shepherded the move from Intel to Apple Silicon.

[00:49:17] And that has been, maybe it was accidental. I don't know, but it turned out to be a really good move in the age of AI. They now make the hottest, the most, they're not hot. That's probably the wrong word. The most popular machines for AI. You can't even get them. Uh, John Ternus, uh, unlike, so Tim Cook's specialty was operations. He was the COO.

[00:49:41] He was the guy who made the iPhone kind of possible in China, uh, and really, really tightened that down so that Apple had an incredible supply chain and incredible capabilities to make a billion phones. Ternus, uh, comes after he's been a long time employee, I think 25 years at Apple. Um, he is a hardware guy.

[00:50:08] In fact, you know, he shepherded, uh, not only the Apple Silicon, but also the phones. So he, he knows the products very, very well, probably a good leader, but, uh, you know, also, uh, a mystery man. We don't know a lot about John. We didn't know. Actually, this might be Apple. We didn't, we knew a lot about Steve Jobs. We didn't know a lot about Tim Cook at all. No, I don't, we still don't know a lot about Tim Cook. We went from having a cult leader to having a professional manager.

[00:50:36] He was very, that's a good description. Yes. I'm going permanently gray and Ternus. Yeah. I'm going to give, I'm going to give it to a presidential four years. He is in his fifties. He'll be, he'll be looking like you and Jeff. Hey, nothing wrong with that. Distinguished, right? I didn't say there's anything wrong with it. I'm just saying it's coming.

[00:51:01] I'm saying the stress of the big company that on his shoulders, like he's got that dark and lovely going on right now. Those few little grays, but it's, it's going to be, look at that. Look, it comes four years. I'm going to, I think Eddie Q. The president has gone from, from blonde to brunette. I think this is going to be the year of the brunettes. Don't you? I'm sorry. That's too political. Eddie Q. Who is long time Apple employees kind of moved up in the chain.

[00:51:29] You notice he's sitting right next to John Ternus and the executive profiles. He has a larger portfolio than before. Very energetic guy. He's the guy who has the relationships with Hollywood. It's an interesting place. Craig Federighi is still going to be there. Jaws as well. Although many people think those guys may be on, on the way out. A lot of, a lot of Tim Cook's people have retired. There is an Apple smile. Yeah. They all have it. Don't they? Yeah. It's probably AI.

[00:52:00] Or gummies. I'm not sure. It could be, it could be either. I don't know. I'm sorry. I shouldn't say that. That's disrespectful, Leo. Apple is, by the way, in a big fight with OpenAI saying, you know, you stole our employees and with it, our intellectual property. OpenAI says, we don't do that. We just hired the guys. We don't. We're not. Meanwhile, OpenAI is working on hardware. We know.

[00:52:29] And this will be a, this will be a court case we're all going to be watching. With Johnny Ives, no less, right? Yeah. Johnny not mentioned in the suit. Just the people Johnny hired. Which, I know. Nobody's last comment on the socials was, well, if you can't keep your secrets, it's not our fault. Oh, that's not the right way to say that. I don't think a judge or jury will take well to that. They were, they had the better plan when they said, we don't want your secrets. We don't need your secrets.

[00:52:58] We don't need those stinking secrets. That's the better plan. We're training our hiring model on your hiring model. I mean, I don't understand what the issue is. Who's destroying books? I mean, we're just taking the. OpenAI is about to learn a lot about courtrooms because the New York Times suit. Yes. So filings made Friday in the New York Times suit. This is from the New York Times. So take it with a grain.

[00:53:24] It's all closely watched lawsuit against OpenAI and Microsoft included a range of copyright law and cultural references. The real question, and I think it's an interesting question. Is if you have a copy of the New York Times or the Seattle Times or Newsday who are also suing. If you have in effect ingested that content into your weights in the AI. Does that count as an illegal copy?

[00:53:53] Because it's pretty clear you cannot really generate the contents of the New York Times by asking the AI for it. Even no matter what the time says, you really can't. No. So is it illegal for them to have a, in effect, a copy in the weights? It isn't a copy because it's not even words. No, because it gets erased. It doesn't. It's not held on to. They learn from the relationships of the words.

[00:54:22] The U.S. Justice Department filed a brief saying it is of the national interest, the highest interest in the United States of America for the judge to find open AI did not violate copyright law. The Justice Department argued developing AI is critical to national security and the training AI systems sufficiently transformed the written works to new material allowed under copyright law.

[00:54:47] And it said, and I think this is really the most important point they're making, the benefits of AI far outweigh any competitive harm. How often do you see the federal Department of Justice enter in in a private suit at this stage? That's a good question. It feels like not very often, ever. It may happen all the time. I don't know. But it just seems odd. They sometimes file amicus briefs. But I don't know how long they've entered into a suit or how often.

[00:55:17] The Trump administration is in a very difficult position because he got elected as a populist president. And it's become increasingly obvious. The populists, the popple, the people do not like AI. They do not like data centers. This is going to be a midterm election hot button topic. They are very concerned about things like flock cameras. So there's... Concerned, I would say, is late. Angry about. Angry.

[00:55:48] That's the populist side. On the other hand, he has massive donors who are big investors in AI. People like Andreessen Horowitz. So he's got pull from two different sides there. Zuck called him, I think, recently. Yeah. To say, please don't regulate. Yeah. And we know that Mark's given money to the president. And then there's the third party, which is, I think, very important also to the White House. The defense and intelligence establishment.

[00:56:17] Intelligence establishment. We know they're using these AIs now. They want them. In fact, in a way, if I read between the lines of the Justice Department brief, that's what they're really saying. They're saying it's important for society. No, it's important for our military. Well, there's a China angle. We're going to fall behind China, and it's a strategic risk for the nation.

[00:56:40] Trump said AI dominance is critical to promote national security, prosperity, and economic mobility for all Americans. Actually, that was... What do you think he means? Yes. And I ask this, when we talk about AI dominance, that just means we have the five or six top AI companies? Which we do right now. By market cap. Well, that's... Not just by market cap.

[00:57:09] I would say by quality of AI. But it's also by... Quality of models. Yeah. By... But it's not... But that lead is shrinking. By capabilities, too. Yeah, but it's shrinking fast. Yes. And to build a quality model, you just need access to lots of data and maybe a breakthrough idea in how you train it, which happens every now and again when we get... And so... And by the way, a third thing. A lot of capital because you have to buy a lot of hardware. Yeah, a lot of money for your... Yeah.

[00:57:38] But if you think about it from that perspective, like capital for servers, the internet, like the data going in or a breakthrough in like how something is trained, the dominance is going to automatically ebb and flow based on like if China comes up with a way to like power a server using seawater just for fun, we're going to suddenly see like they will become dominant, right?

[00:58:07] So I feel like we're always very short-sighted when we start talking about this because we don't think about the components that we're actually trying to be dominant in. And then if you think about what we're caring about, the fact that we like decimated our research funding for lots of things is probably really terrible. So I'm just thinking about this because this is how I think about the world and see the world when I make it up. Well, I will answer on those points.

[00:58:31] So first of all, if you look at who the leading scientists are in AI, they're almost all Chinese, many of them working in the United States, but they're most, many of them coming from Chinese universities. So I don't think on the innovation scale, on the technology scale, we have much of a lead, if any of a lead. The Chinese have worked very fast, very effectively to develop their sustainable energy.

[00:58:58] They have a much better long-term energy prospects than we do in the United States, where we are still highly reliant on fossil fuels. And in fact, liquid natural gas, which is highly problematic from a political point of view. And somebody, we don't have to mention, killed off projects on solar and wind. Right. Demonizing them.

[00:59:21] So, and then now capital markets are maybe where the one thing where we really do have an advantage because we have a very vigorous and very well endowed. Yeah, but if you have a net. Capital markets are fickle. They'll go wherever they think. Well, and in China, they also have the government. Well, they have a controlled economy, a centralized controlled economy. So you kind of don't need the capital markets because you decide where the money goes. And I think everybody agrees. I'm using mostly all the local models I use are Chinese.

[00:59:50] I mean, everybody agrees that Chinese models are catching up really rapidly. So I think there is a threat. A threat of what? And by the way, let's add one more thing. A threat of what? Well, hold on. There's one more thing. They don't care so much about intellectual property rights. You're not seeing any lawsuits in China by the New York Times or the Beijing Times or whatever. So they have ample access to the data. They don't. Nobody's slowing down.

[01:00:19] God knows what kind of data they have over there that we don't have over here because we're prissy about, oh, I don't know, privacy and individual rights and all that. So to answer your question, Owen, I mean, it really comes down to the quality of the models, right? And they build. Oh, there's another thing that we do have an advantage and they don't is the ability to serve those models because we have kept, for the most part, NVIDIA's chips out of the China market.

[01:00:42] Which, as in Wong says, is a huge mistake because he lost CUDA as a standard. The next thing that's happening, and it's already happening, is Chinese companies are developing competitive chips. And if you use DeepSeq or GLM or any of the Chinese models on a Chinese server, you're not running it on NVIDIA cores. You're running it on Chinese chips, on Hyundai chips and not Hyundai, Xiaomi.

[01:01:11] And you're running on chips from China and Huawei. Huawei. That's the company I was trying to remember. You're running on Huawei chips. And so. Don't forget, Stacey knows her chips. She's a chip guru. Actually, I want to ask her. We'll get to that. Yeah, we want to get it. I actually want to. Yeah, I want to. But right now, the only thing really, come to think of it, we have an advantage in, and it's rapidly fading, is in the NVIDIA chips.

[01:01:41] It's in GPUs. And as you and I discussed on our chat earlier today, I think we could have had an advantage out of the academe. But A, we've starved, the government has starved them. And B, the Chinese technologists who were coming over here to study are not coming here now. So we're losing. That's a brain, a voluntary brain drain on our part.

[01:02:06] So I understand how the American government might start to say, gee, you know, we are in this kind of cold war with China. And we may lose this AI battle. If you think AI is important, especially if you think it's important militarily and intelligence, that might be scary. So to answer your question, Owen, it's about the quality of the AI, the models, their capabilities. Not only are we going to lose, not only are we losing, we're going to lose. This administration is inept and incompetent.

[01:02:36] And just saying things out loud doesn't make any sense. We read a quote from the president. You forgot to say, if you don't want an AI identifier in your backyard, you're poorer. Oh, he did say that. He said you're going to end up dumb. Yeah, and poor. Again, the EPA isn't checking the air pollution that they're putting out or the noise pollution. No one's checking on the water when they tell me that water doesn't matter and it doesn't mean anything. Then on top of that, what? They just had that robotics thing where they had like the guys running and racing and one guy could shoot like a thousand baskets in a row.

[01:03:07] The humanoid Olympics. Only three companies from America could even compete with that, like MIT and whoever else. But the other 600 other something companies are all Chinese. They're the ones making the robots that are doing the things, running at hyperspeed and running in the walls. We're not doing any of that stuff besides getting like fake AI generated stuff from Elon telling me that we're going to have robots running around that he's going to build for us when he can't even get the cars right or the spaceships right. So losing what we've already technically lost.

[01:03:36] Like we have the intellectual property. But again, like Jeff said, China doesn't have any of the gates holding them back from stealing. That's why you see the whole thing like Nike is like, oh, Nike's stock has gone down because they found out that reps are real. Do you guys think? They're using the same factories. Now you can go buy them straight from China for $50 instead of $250. Now nobody's buying real Nikes anymore because why would you do that?

[01:04:01] Do you guys think that bipedal humanoid robots actually are the future? No. No. Okay. I'm just curious. They certainly are leading in this respect. It gets you to something, right? You're doing it to get to the same reason they put cameras on a barber and said, hey, we want to get we'll pay you $10,000 for you to cut hair for you and let us scan and copy you cutting hair. So we can teach our robots how to cut hair. At some point, this stuff will matter.

[01:04:29] Right now, it's all fun and games until it's not. Right. So I just signed up for an event at the Aspen Institute where they're looking. You missed the best part. You cut away. No, not you. Nobody wants to see that. Nobody wants to see when he ran this to the wall and all that stuff. We just want to see. No, we want to believe technology works all the time. This is an Instagram. Yeah, there you go. That was the one. From evolving.ai. Yeah.

[01:04:57] This is all of the outtakes from the Robot Olympics. Now, stay on this one because the robot does a good job climbing up onto the cube and looks kind of human. He's standing at the edge now. He's going to jump off. This is going to be very exciting. He's going to leap from that five foot high cube and land on his head. Oh, I shouldn't laugh. I shouldn't laugh. They're keeping track. Do not laugh. Look, look. You wouldn't even believe that was possible. No, it's pretty amazing.

[01:05:27] In fact, I remember laughing 10, 15 years ago at the DARPA Grand Challenge where they had self-driving vehicles that couldn't go 10 feet. They'd drive right off. Remember that, Stacey? They'd drive right off the road. We've come a long way. The Wright brothers only flew 148 feet and they took credit for the plane. You know, their plane didn't. It wasn't the plane that got us to where we are. Their plane couldn't go 100 feet.

[01:05:51] My favorite in the humanoid Olympics was that somebody realized you don't have to run if you're a robot the same way humans do. We do our arms like this because it has to do with how our hips pivot. And the robot ran like this. It looked weird and it was very fast.

[01:06:14] So it's a little weird that the slavish adherence to the bipedal form, like to make it look like us, isn't necessarily the most efficient, best way to do this. Yeah, because the robot dog doesn't like that. Sorry. Oh, no, I was just going to say you can learn a lot about human mechanics and even like where things might break down. But yeah, I think many of the robots in our lives will not be like us.

[01:06:44] Sorry, I'm laughing at the robot. The lost might not. See, it used to be these videos would scare you when you say it's all creepy and awful. Now you're laughing at them, which is a step forward, I think. The creepy ones were the dog, okay? Big Dog is still a terrifying, terrifying robot. Oh, Big Dog from Boston Dynamics, still very scary. And you know what? ICE is buying those. Yeah. Yeah, that's really, that is terrifying.

[01:07:09] I think that's the real value to some degree of bipedal robots or robot dogs is their fear factor, how scary they are. Yeah. And they can control them. Especially if they're armed. If they're armed and autonomous. That's right. Yes. Especially if they're armed. Anyway, yeah. So I don't know if the humanoid robot games provided much more than just mirth and merriment.

[01:07:34] So I want to go back to the data centers for a second, because Craig Newmark linked to an event for the Yasmin Institute upcoming next week. And it's really, or this week, it's really interesting that there's an argument for alternatives to data centers. What if we approach it like public AI, centering public access, accountability, sustainability, the idea of tiny distributed networks, public owned utilities, the role of librarians, farmers, community advocates.

[01:08:03] So I think there's an effort to say, okay, we're going to need data centers. But the way that Elon Musk does it with jet engines ain't the way. Well, I mean, I hate to say it, but we're using a data center right now to record this, to stream it. You're watching it. You're watching it from a data center, not from anywhere. Well, you downloaded it, but you downloaded it from a data center. The entire internet is data centers. Your bank is in data centers.

[01:08:30] You need to price in the externalities of the harms of a data center. I mean, I've covered data centers for, what, two decades? Hey, just like that guy who covered the Microsoft, Steve Lohr. I'm like the Steve Lohr of data centers. Is it that we're just making too many now all of a sudden? I mean, because we've always needed data centers, right?

[01:08:49] I mean, yes, we are rapidly expanding the amount of compute that we need for AI, right? That is unquestionable and it is a exponential increase in demand. But we also are doing it under the older economic model of data centers.

[01:09:13] So if we actually start charging what these things might be worth, one, it would slow down our innovation. So you've got to think about that. But maybe that's what the government does is they put their dollars towards paying for better innovation. Let's put it that way. As opposed to today's market capitalism, which pays for the cheapest innovation, which is worth thinking about.

[01:09:38] Which is why, that's why government funding exists, is because we want to do things in a way that makes sense and prices in those externalities. We wouldn't have the internet if it weren't for government funding. Right. Yeah. And it took a little bit of time. I mean, the problem is, the problem is we took technology out of the technology industry or we kept that in, but we added a whole bunch of finance bros. Oh, yeah. Yeah. The financialization. And crazy, test-grill cultists, which I think is the real problem here.

[01:10:08] You know, they were always there. Like, I mean, if you hung out with enough nerds, there's always that, there's one or two that you're just like- They're in power now. That's the difference. But yeah, now they're the bosses. They're not the bosses. They look like they're the bosses, but they've just been given a lot of money. And the guys are like, yeah, you know, this dude believes in aliens. Don't worry about it. He's going to believe in aliens. He's going to deliver this technology. The guy who started Excite believed in aliens. Peter Thiel's kind of a boss.

[01:10:35] Peter Thiel just took over a state in Germany. Peter Thiel's kind of like trying to run half the globe. Like, it's not- All right. We got to take a break. I want to come back and find out what took over a state in Germany. That's an interesting thing to say. We're going to take a break and find out what that all means. You're talking, we're talking Twit here with Stacey Higginbotham from Consumer Reports, policy fellow. Great. So great to see you again. Stacey's book club is a regular feature of our club. I hope you're a club member. You can go to the next book club.

[01:11:04] I think the voting is still going on. Anthony, if you're there, I know you are. Give me a link to the poll so we can see what's going on with the voting. There is no anti-memetics division. It's going to be that? Okay. That's actually a great book. I've read that book. Okay. By far, like a landslide. Nice. 82% of the vote. Oh, our politics should be so clear as Stacey's book club. Who is quantum?

[01:11:34] QNTM. Is that a human? Or is it written by a robot? It's a person. They have an actual name. I don't remember what it is now. It's a very interesting book. I'm glad we're going to read that. That'll be fun. So that's the next. Yeah. We went with a slight horror theme for, that was our. Oh, it's a Halloween. Yeah. I assume we'll do it. Oh my gosh. Already. Yeah. I mean. Pumpkin spice time, man. It's here. Oh, no. You don't do PSLs.

[01:12:03] It's matcha. Matcha. Do you do pumpkin spice matcha? No, no, no, no. No, pumpkin spice is no longer like the reigning flavor king. Matcha is the reigning flavor king. I never had it. I never had a pumpkin size latte. Now that it's not hip anymore, Jeff, you and I can have it. I think I'm going to have one now. Yeah. Have you ever had pumpkin pie? Well, yeah, but I've never had a pumpkin spice latte. It's the same spices in a latte with a lot of sugar. And it has nothing to do with pumpkin, by the way. I'm assuming. Never saw a pumpkin in its life. Nope. Nope.

[01:12:34] It's all spice mostly, right? Yeah. It's all spice, clove, cinnamon, nutmeg. That's good stuff. Yeah. It's delicious. Chai is good too. Also with us, Owen, JJ Stone. We call him O-Doctor. I still don't remember why, but I'm sure there's a reason. But so real quick, Stacey, you lucked out on this because she doesn't think it's your fault. I texted my daughter and I said, you care that I randomly check your location?

[01:13:04] Leah said, tell Uncle Leo who else is going to come and save me. Yes. My. She doesn't think it's you. She thinks Leo asked that question, so you're off the hook. I'm still allowed to randomly check until she finds another man in her life to come and take care of her. Who's going to rescue me? How do you feel guilty? Isn't that nice? You're the knight in shining armor for her. Well, I'm glad you checked with her. That's good. That's the right way to do it. I never even thought to ask. I'm like, hey, look, I still pay the bills. You got a phone because of me. I asked.

[01:13:34] I'll turn this thing off up in here. Don't let me ask no questions. Stacy got me feeling guilty. Also, Jeff Jarvis is with us. We will see him as a regular feature on Intelligent Machines. And we're so glad to get you on the weekend as well. I'm glad always to be on the big kids table. Big kids table is what he calls it. Our show today brought to you by my mattress, Helix Sleep. Loving the Helix Sleep.

[01:14:04] Every night when I get into bed, do you do this? If you don't, you need a Helix Sleep. I get in a bed, I go, ah. It's just, ah. It just lets all the trouble of the day. It goes, ah. Ah, just, you relax. You feel good. If you've got a great mattress, I think that's 90% of life. You spend at least a third of your life on that thing. Maybe more.

[01:14:31] Some of the most valuable moments on that beautiful mattress. And what better time than Labor Day weekend? It's Labor Day weekend. To give your bedroom the upgrade it deserves. Instead of, are you, I mean, if you're not going, ah. If you're struggling through another night of tossing and turning. If you're looking at that mattress going, do I have to get in bed? You should treat yourself to the deep restorative sleep you've been dreaming of.

[01:14:57] Don't settle for mattresses made overseas with questionable material, to say the least. Let's upgrade to a Helix mattress proudly assembled and shipped from Arizona. And finally say goodbye to night sweats, back pain, emotion transfer. Lisa and I realized, I think it's more than a year ago, that we'd been sleeping on our mattress, same mattress, about eight years. We thought, it's time for a new one. I think you're supposed to get a new one every eight or ten years. So we shopped around. We looked.

[01:15:27] We read all the reviews. One of the things we liked about Helix, among many, they have over 20 mattress models. There's one to fit every sleep style, every sleep preference. And they make it easy to figure out which you are because you can take the Helix sleep quiz. We did. It matches you with the perfect mattress based on, you know, are you a front sleeper, side sleeper, back sleeper? Do you like firm? Do you like soft? I mean, there's a lot of possibilities. Take the quiz. Choose the mattress.

[01:15:57] And by the way, their happy with Helix guarantee means that if it's not exactly the right mattress, you can get a new one. Easy, seamless returns and exchanges on that happy with Helix guarantee. So it's a risk-free customer first experience. I think it's really worth getting the right mattress. It's worth taking the time to just get that right mattress. It worked for us. In fact, it worked so well. I'm getting more deep sleep. I'm sleeping longer.

[01:16:25] And most importantly, I'm feeling better in the morning. Helix actually did a Wesper sleep study where they measured participants' sleep performance after doing what we did, switching from an old mattress to a new Helix mattress. They found 82% of participants saw an increase in their deep sleep cycle. That's the most important one where the brain gets cleaned out. You know, the cerebral spinal fluid comes up, the brain dilates, goes, washes it out. It's, you know, usually not more than half an hour, an hour a night, but it's the critical cycle.

[01:16:55] Participants on average achieve 25 more minutes of deep sleep per night. For me, that was what I got. It was doubling my average. They also, participants on average, achieve 39 more minutes of overall sleep per night. And it's pretty subjective, but I can tell you, I feel so great when I get up in the morning. I love it. Time and time again, Helix Sleep has been named the most awarded mattress brand, tested and reviewed by experts like Forbes and Wired. You can read the reviews right there.

[01:17:24] They deliver that mattress right to your door with free shipping in the U.S. And of course, that great Happy with Helix guarantee. So, you know, you'll be completely satisfied with your new mattress. Go to helixsleep.com slash twit. 27% off site-wide during Labor Day, the best of web offer. That's helixsleep.com slash twit. 27% off site-wide exclusively for listeners of This Week in Tech. Now, make sure you do enter our show name after checkout This Week in Tech so they know we sent you.

[01:17:54] And this offer ends September 13th. So, don't wait. And if you're listening after the sale ends, don't worry. They've always got great deals. Check them out. Helixsleep.com slash twit. Thank you, Helix, for my mattress and for a great night's sleep and for sponsoring our show. Helixsleep.com slash twit. When I was a kid, my grandfather used to tell me something I'll never forget. And I tell people all the time. Three things you don't cheap out on. Tires, your shoes, and your bed. That's so true.

[01:18:25] They got to be bright. They got to be good. And the bed is a great thing. You spend more time on your bed than you do on your tires or your shoes. You do. But not as much time as eating pumpkin pie. I've never had pumpkin pie in my life. What? You're an American. Okay, wait a minute. Wait a minute. No, I know why. Because you like sweet potato pie. Yes.

[01:18:51] I literally ask when I go places, what is that? Because it's a good confusing thing. They look the same. They're both orange, but they do not taste the same. They almost smell the same. And I'm big, so I got to take the sniff test. I'm like, this ain't right. I refuse to do it. There is nothing like a good sweet potato pie. I grew up with my grandmother making 30, 40 pies for the church and everybody in town. And I'm not going back. I'm never, ever, ever doing it. So I've never had a pumpkin pie.

[01:19:20] Shame on pumpkin pies. And that picture is blasphemous. Burn it from the internet. Maybe we're eating sweet potato pie. You don't know. That's true. That's true. It says pumpkin pie on the sign. I was about to say, didn't the sign say pumpkin pie eating? It could be a lie. It's AI. It does not read. You can't trust that. Do you put it with cream on a sweet potato pie? Oh, yeah. You do. Okay. I, for Thanksgiving one year, made Owen's secret macaroni and cheese recipe. And it was dino might. Yeah. So good.

[01:19:50] In fact. Season is coming upon us, Uncle Leo. It's time once again. Macaroni and cheeses, turkeys and pies. Oh. I think this is when Stacey gets hungry and goes off to get a cookie. I'm like, dang it, you guys. America on hungry. She used to get waffles during the show. She would go sneak off and get a cookie or a waffle. That's true. Let's see. What else is going on? Meta accepts U.S. safety rules, but says, but the rest of you guys and the rest of the

[01:20:19] world, could you just chill a little bit? As governments consider banning social media for under 16s, rest of world rights. Meta is turning to influencers to argue that parental controls are a better solution. Now, I am not an influencer and nobody from Meta has ever talked to me, but I kind of agree. The parents should be in charge. Well, so we still have alcohol limits and laws. We have anti-smoking laws.

[01:20:47] We have laws plus parental control for lots of things that provide a documented harm to people whose brains are not done developing. I'll tell you the big difference. Okay. Hit me with your big difference. When you go into a store to buy liquor or cigarettes and you show them your ID, that's one thing. But in order for these laws to be, to take effect worldwide or even in the United States

[01:21:15] or state by state, every single person who uses that has to provide an ID over the internet where it may be stored and misused in other ways. It is a very different thing to do. We don't have to do it that way. Oh, okay. This gets to me because we keep choosing wrong from a technology perspective. And the reason we do it is because companies don't want to be held liable. They don't want to be part of it.

[01:21:41] But if you had Apple or Google or Samsung or whoever owned your device locally understand your age and check for you there, then it is a local check. And NIST actually had built a framework. The National Institute for Standards and Technology. Yes. Sorry. Those guys. Real nerds. Now that she's a bureaucrat, she just speaks. I know. NIST was always my favorite agency. I love NIST. Yes.

[01:22:12] I'm sure y'all are all surprised to hear that. Anyway, so there are ways to do it. But what we've got is this stupid, stupid, stupid thing where we're like, yeah, let's show a really important government document over the internet and let some company that we've never heard of before store it for us in a way that we can't check to make sure it's secure. And oh my God, if we go scroll down a couple stories, it's not. So I agree with you.

[01:22:38] Age verification is terrible and wrong, and we should not be doing it on the internet for a bunch of reasons. But we didn't have to do it the way that we're currently doing it, is all my point is. And now we're going to have to argue over this. It's meta. Well, it's a couple things about this. One is that meta, $17 billion over 10 years isn't a lot of money when you adjust for stock prices and acquisitions and such.

[01:23:08] Second, it's a case now of meta and regulatory capture, trying to impose this view on others. And I read a Q&A with Julia Angwin and Daphne Keller. Daphne's at Stanford. I respect her greatly. And she said that now she fears this is going to become the standard without any data, without any testing. And if others come up with better means to manage this, they're not going to be able to because they're going to be expected to follow this model out of this settlement. But doesn't meta want Apple to do this?

[01:23:37] I thought meta wanted the platforms to do what Stacey was talking about, not make meta do it. Yeah, I'm talking also about the two hours a day and all that stuff. Yeah, he's talking more about the rules. Yeah, I agree. I look at the regulatory capture potential here. So explain regulatory capture. It's basically like when a company gets to make the rules and they control their agency. So the telcos and the FCC is a great example. And they get to write whatever they want is their regulation.

[01:24:07] That's how we got shaken stirred, which is not a great. Yeah. Maybe they'll make it better this next round. Hasn't that worked? Are you kidding me? Do we get spam calls still? Oh, wait a minute. Okay. I was like, wait, what? Oh, wait a minute. That was, I mean, in principle, that seemed like a good idea. The idea that you couldn't use a phone number unless you authenticated it from your carrier. Like you really were a real person, not a fake number.

[01:24:35] The reason it didn't work is because they blew the attestation part on that. There were other bottles in place. But you know what? I bet they're fixing it, possibly. Really? I'm filing comments with the FCC on Tuesday about part of this process. So whose fault was that? Was that the FCC's fault? It was bad legislation. So the rules that... On the surface of it seemed like a good idea. On the surface, everything's a good idea. The devil's always in the details.

[01:25:05] I mean, that's why people like me have a job. I say that because I was an advocate for shaking stir. And you were... It was not a terrible, crazy plan. It did put a lot of power into... It let the carriers off the hook. And it gave a lot... The loopholes were so big, you could drive I don't know how many scam calls people get a day. A billion scam calls a day through it. Yeah. It's clear it hasn't worked. Yeah. More fax machine policies, it seems like to me.

[01:25:31] So the idea was that a carrier couldn't get... Couldn't operate in the United States unless they set up an authentication system that would say, yes, this person is a subscriber with this phone number. And that authentication was provable on the other end. And the receiving company, they could... That was... One of them was shaking and one of them stirred. They would say, oh, yes. Okay. Authenticated. We're going to let that one through.

[01:25:59] And of course, spammers, in order for them to do their dirty work, don't use... That's why it's always coming from your area code and sometimes even from your neighborhood because they fake the numbers. And shaking and stirred would stop that, right? But it didn't. And that's what puzzles me. So it... Okay. You really want to do this? Well, give me the... Next is going to be chips. Give me the 30,000 foot if you can. All right. I'm going to try. So the FCC did slowly cut off phone companies that didn't agree to this.

[01:26:29] And by now, I think they've cut off everybody that didn't agree to it, right? So part of the issue is... All right. With shaking and out stirred, they created this three-tiered system where you could authenticate yourself as a legitimate provider with a carrier using... It was like A, B, and C attestations. And if you were like the strongest level of sort of... Like you were totally like the carrier has done business with you for years.

[01:26:59] They know who you are. That's like the highest level. What happened... Your calls would always go through. The second two tiers were a little bit more dodgy. Like maybe they don't know you so well, but you seem to be legit. They, you know... You seem okay to me. And then there's like, you just started up yesterday. You're based in Cambodia. I don't know about you. That I think it was the middle tier is where it got...

[01:27:24] The carriers have discretion in letting those middle tier calls through. And what happened is they make money. That's the real problem, isn't it? That they make money on these fake calls, these spam calls. So the attestation, you know, they're like, well, let's just... I think this is totally legit. Yeah. That's really the problem is if they're making money on these, they're not incented to make it work. And the FCC has to really put down the hammer.

[01:27:53] And the FCC is in the pockets of the... We just don't answer our phone anymore. This is the regulatory capture part. So they were allowed to do this because, you know... Yeah. Anyway. Okay. So meta and regulatory capture. So, yeah. What is the analogy here? A couple of things. One is that meta pays less money if... Is it more money? I'm forgetting which. If the government convinces the other major social platforms to follow the same rules.

[01:28:23] In this settlement that meta made, which was very weird. They said, we're going to pay $18 billion over 10 years. Not much, in other words. And we'll give you another... Of that $5 billion is dependent on Snapchat. Right. And TikTok and YouTube doing the same thing. Which is weird. I don't get the logic of that. Why would the... Well, the AGs, I guess, wanted to strong arm those other companies. No. But the companies don't care. There's no prisoner's dilemma there.

[01:28:49] But then what that does is that sets up the rules here. The two hours a day and all that. And there's a lot of argument here that the time limits don't work and could be damaging and don't really address the problem. But now, because of this settlement, not law, not adjudicated court law, this settlement, now this becomes de facto the standard for protecting children with no data whatsoever. Is this going to do a damn thing?

[01:29:20] And now meta's going out and paying influencers to advocate. Who isn't paying influencers to advocate? My God. I know. Where is my influencer money? Oh, nobody listens to me. That's the problem. Yeah. And I've told you this before, you're not doing the business the right way when it comes to stuff like that. But it's all your soul, Leo. It's always amusing to me when they talk about protecting kids when there's so many helicopter parents that are not helicopter parents that just want to shove an iPad in their kid's face and say, go on and go do...

[01:29:50] Isn't that part of the problem? You see that everywhere. And every week there's a video like, oh, Billy ordered a dirt bike. How did Billy order a dirt bike? I saw you on Amazon, Mom, and I just pressed all the buttons. But you want me to have somebody regulate that you gave them the device and you didn't lock it up. And this is something I'm not talking about old people in fashion, but there are so many hip and current parents that don't know any better. My one... I brought this up a billion times. We talked about it. My daughter's phone is blinging in the middle of the night. And she was like in the eighth grade, ninth grade. She's had a phone since she was six.

[01:30:20] But anyway, her phone starts going off. I'm like, who is messaging her? And her friend got a phone for Christmas. She's 12 years old, two o'clock in the morning on Snapchat, just sending off snaps. And she was waking up late for school. And I told her dad, I'm like, yo, she's on the phone. No, she's not. She's like, dude, she's sending my daughter messages at 2 a.m. That's why she's not waking up for school, because he's complaining about her not waking up for school. He's like, I'm checking her phone. There's no text messages. I'm like, she's not in the texts. He didn't know. He didn't understand it. He did not understand.

[01:30:49] So it's like, you can't tell... I mean, not that you can't. It's just hard to regulate something when the parents aren't even regulating the children themselves when they just want to give a kid a device so that they leave them alone for 10 minutes a day or... Well, that's why I kind of like the idea of the burden being on the platforms and the phone manufacturers. Apple has, I think, stepped up. They're doing better parental controls. People don't like that either. A woman made a video went viral that it said, oh, they were using the FaceTime group chat

[01:31:18] thing. And it said, this conversation might be harmful. I don't know what the kid said that triggered that, but since they had like child parental things on there, it turned off the call. And the woman's like, well, how come Apple's listening in on my thing? Like, well, you don't want them sending nudes and all this other kind of stuff. Like they're checking things because you said it was a child using the device. So again, you're asking for things and people complain about them. It's such a weird... Right. And Apple probably, and every company that deals with customers knows that.

[01:31:46] They know that this is going to be a problem for their customers. Also, Jeff Koss, if you remember, who wrote the 26 words that made the internet? Or was it 25? Or was it 27? I forgot already. We're talking about Section 230. Don't talk about Section 230. Jeff wrote the definitive quote. Do you really want to get into this? He just posted on Lawfare about the First Amendment problems with the Meta service there. And there have been cases about children have First Amendment rights too. Well, that's why... Remember we talked about...

[01:32:15] You weren't here, but we talked about this last week or the week before. That's why the attorneys who are going after Meta were smart to say, it's a product liability case. It's not about speech. But it still has implications for speech. Well, I agree. But the jury didn't agree in the LA case. And apparently Meta thought that they were going to have the same problem with this most recent case because they settled shortly after the trial began. And that tells me they felt like it's not going to go well.

[01:32:44] In any event, $18 billion over 10 years is not much for Meta. I think they got a good deal. It's another case of getting away with murder, so to speak. I read they spend $2 billion a year on lawsuits. Yeah. Your legal budget. So that's actually higher than the amount they're going to pay in the fine. The lawyers are making that like bandits. I got in the wrong business. That's for sure. How are they starving and homeless people? Okay. Where does the money go?

[01:33:14] How are people starving and how are their homeless people? You're just throwing out $2 billion a year. We got the wrong priorities. Fines. And I'm like, my God, the money, the thing about the AI that we talked about earlier that I didn't say because I'm just trying not to yell about AI all the time is they say the government needs us to excel and do all these things. Well, you're going to let them IPO for companies that aren't making any money. Like you read the articles and like, well, Uber was making money. None of these guys are making any money.

[01:33:43] And they're about to IPO and put it on the backs of the American people and get rich and get paid out for what? Building a product that the government says we have to lead in and you're going to build all these data centers or roam people's air. It's just a big old scam and I can't stand it. You know, I use it. So don't, I'm a hypocrite. I know. But at least I'm not one of my agents and firing people like Uncle Leo. I'm out here editing my videos and doing rational things with my AI. Are you doing that? That's cool. Yeah, yeah. My AI stays employed. I haven't fired anybody yet. You know?

[01:34:13] No. I hired him right away back. I hired him back. Actually, I hired him to do an investigation. The local AI that I installed screwed up so bad. And here's the postmortem. I said, did I break Kermes or does it need a better human running it? And it gave me the whole, this is Astra, gave me the whole thing. The whole postmortem, the breakdown. I preserved the evidence.

[01:34:43] Might be a you problem. That's what I, you know, I thought I was being generous. I said, is it a me problem? I said, well, no, not really. But then I don't know. Do I trust it? Couldn't be me. Couldn't be me. Let's take a little pause, a little tidy pause as we continue talking about the week's tech news and this week in tech with Stacey Higginbotham from Consumer Reports. Owen J.J. Stone from Parts Unknown. Yep. I believe probably somewhere close to Philadelphia. I'm just guessing.

[01:35:13] You're absolutely right. I'm in a black box very close by. Checking out all roads and bridges. Wish I was in New York City. They got a great mayor up there riding a bicycle around town all the time. You like the commie, huh? But you got pretzels and cheesesteaks, man. So, you know. We do have pretzels and cheesesteaks. You do. And probably sweet potato pie if you know where to look. That's right. That's right. And the Cacio e Pepe Kid, Jeff Jarvis, author of Hot Type.

[01:35:43] We're doing the old, what would Google do? I mean, this week at Google, best hits. Yeah. Cacio e Pepe. No, I think we were all into the Chipotle at the time. It's all about the Chipotle. Well, Stacey made royal fun of me for my McElwee pasta. Cacio e Pepe. Your frozen Cacio e Pepe. I'm getting a taste for it tonight. Oh, no. No. I mean, I'm on my I'm batching it because the family's going out to see Spider-Man and I refuse to go see comic book movies.

[01:36:12] So they're just leaving me behind. So see comic book movies. Come on, grandpa. Spider-Man is a comic book movie. Yeah, it's a book. At least it's a comic book movie. Yeah. You know, any movies are out there. There's a book or any literature attached to it. How many pages? You know how many kids learn how to read from Spider-Man? Thank you, Spider-Man for helping. Did you see the Odyssey, Leo? Not yet. That's that's based on a book as well.

[01:36:41] That's a comic book. It is. It's true. It was a spoken comic book. It's an audio book turned into a movie. Yeah. Turn into a movie. Yeah. Don't see. No, I haven't seen the Odyssey. Maybe I should go with them. They see Spider-Man and I'll go see the Odyssey. You can't see the Odyssey. It's a comic book. There's pictures and there's a story. It's a poem, actually. So you're not allowed to see it. Oh, they've got magical beings and witches. And you're not allowed to see the Odyssey. That's not. No, you. Sirens.

[01:37:12] Yes. You're only allowed to see reasonable, rational. How many Spider-Men are there? Just out of curiosity. There's a lot of them. Like 15,000. Who's the current Spider-Man? Tom Holland, right? Tom Holland. There's three live action. But if you go into the Into the Spider-Verse Spider-Men. Oh, that one movie had a bunch of them. Yeah. Spider-Pig. There's probably a Spider-Leo in one universe. You know? Spider-Pig. It did cut Spider-Noir.

[01:37:41] So Nick Cage's Spider-Noir doesn't get a second season. It didn't get a second season? It's got so many awards lined up, nominations for awards, and they did not do a season two. I hate that. And it was black and white. Yeah, I thought it was kind of interesting. Well, you could choose. You could choose black, white, or color. It got canceled. Amazon is ruthless. They are just ruthless. They're throwing money at things, and they don't keep things that people want. I don't understand the business model.

[01:38:11] Not enough people want it. Not enough product placement. I mean, it didn't match their metrics. If you view everything as some sort of return on investment as opposed to creating a cultural or an artistic good, boom. Spider-Noir on Nick Cage? Yeah, it's worth watching because it's very interesting. Nick Cage is very interesting. Paris is going to be very upset. Oh, no, she knows all about it. Oh, that they canceled it? Oh, I don't know about that.

[01:38:39] As long as we get new seasons of Reacher, which is basically, again, another comic book TV show. Yeah. That was based on a book. Come on. It's an airport book, but it's a book. It's an airport book. It's a book. It's a book. I'm lumping now. I'm going to justify comic books with all books now because Leo's made me angry. So I like comic books. I don't like comic book movies. I can't trust you. You like those Marvel movies? You like all of those? They got me young, Uncle Leo. They got me young. When I was a kid, you know what it was.

[01:39:07] Back in the day when you would watch Wonder Woman and she would just twirl like this and they put a starburst in front of her and she would transform me like, oh, this is cool. Now it's so realistic. I'm like, this is everything I dreamed of is when I was reading the comics and I'm trying to imagine flying through space. I can actually see the thing that I saw my whole life growing up and it's pretty cool. Half the content is trash, but it's still nice to see it. It's good.

[01:39:43] It's real. It's a comic book. It's a comic book. Those are in the Cat of Monte Cristo, which is kind of a comic book, but it's with words, not pictures. Yeah. Pictures you draw in your head. Comic book. Yeah. I mean. It's a serial. Get it together. There are so many literary comic books. I have to say that maybe when you were growing up, you know what? Leo, he grew up when the comic books were not literary. They were comic books. It was Archie and Jughead. It was Superman and Batman. I went through a very brief period of buying them and then I regretted all the money I spent on them.

[01:40:13] Yeah. Because your mom threw them out and you could have made billions. My grandmother threw away my grandfather's Archie and Jughead comic book collection that was worth $350,000. And she did it after he passed away because she said, those are for kids and I hated the fact that he had them. My aunt went down and went to the basement to go look for them. And she's like, I threw that stuff out a year after your grandfather passed away. He had one through 50 in prison conditions. She's like, do you know how much money those things were?

[01:40:41] Nobody wants that kid trashed those coloring books. She called them coloring books. Coloring books. I like you, Grandma. We're going to take a break. We'll come back with more. You're watching this week in tech. Our show today brought to you by Doppel. As in doppelganger. You're double. AI has made social engineering attacks so much more convincing than ever. From phishing emails that are indistinguishable from the real thing. But even worse now, fake websites. Spin it up in a minute.

[01:41:12] Impersonation attempts. It's becoming more and more difficult to tell what's real from what's not. What's designed to deceive you. That's why organizations need more than a collection of point solutions. They need a unified approach to stopping attacks before they reach your people. That's why you need Doppel. Doppel is an AI native social engineering defense platform.

[01:41:36] Doppel strengthens human risk management by training employees to recognize deception. It provides digital risk protection across every channel and delivers agentic email security that doesn't just score the inbox but takes down the attacker infrastructure behind the message. So it can't hurt you anymore. It does what Owen J.J. Stone would do to anybody who tried to attack his daughter. Takes it down, man!

[01:42:03] Doppel protects the entire social engineering attack chain with one comprehensive platform. Digital risk protection detects threats across multiple channels. Links alerts into a real-time threat graph. And then uses AI-driven infrastructure disruption to stop attacks at the source. Those insights also power phishing simulations and security awareness training. So you're getting the best possible training to help strengthen employee defenses.

[01:42:32] Next generation training and testing. Based on actual threats from today, not yesterday. Email security. It inspects every message. Traces it back. This is mind-boggling. Traces it back to the attacker infrastructure behind it. And helps take that infrastructure down so the campaign can't target your organization again. Doppel also offers best-in-class integrations and partnerships. So it's so easy. It works alongside your existing security stack.

[01:43:00] Join hundreds of companies already using Doppel to protect their brand and people from social engineering attacks. Doppel. Outpacing what's next in social engineering. Learn more at doppel.com. That's D-O-P-P-E-L.com. We thank them so much for supporting. This Week in Tech. Yeah, Amazon. They're going to be... They're in a little bit of hot water. FTC is suing them. We'll see what happens.

[01:43:31] Saying that Amazon has been secretly and systematically overcharging for ads. This might do better because they're not screwing consumers. They're screwing Amazon's advertisers. Amazon says... Well, the FTC did just get against Amazon. They had a settlement against Amazon for screwing consumers. Oh, so they're not new to this. This is the FTC and 22 state attorneys general.

[01:43:55] They say Amazon used a secret ad surcharge to drive up prices for ads on its website and app. They were... They were... Oh, okay. So it did... He says... The chairman of the FTC, Andrew Ferguson, says that these higher advertising prices were passed on to consumers. So it isn't like we skated. According to the complaint, Amazon violated the FTC Act in a dozen state laws.

[01:44:23] That's why 22 states' attorneys general are involved. This was a year after Amazon paid $2.5 billion for that lawsuit you mentioned, Stacey, over prime subscription practices and canceling Spider-Man Noir. And Amazon's second price auctions for ads on its platform... That's what they call the second price auction. The winning bidder is supposed to pay one cent more than the second highest bidder.

[01:44:53] The FTC says Amazon's been manipulating those auctions since 2019. Look. It ain't one penny. Up to 80% more. Is that what an article said? Yes. That's insane. In Amazon's auctions, this is as explained internally by Amazon's senior vice president in charge of Amazon ads. This is from the FTC's complaint. In Amazon's auctions, the second price isn't set by an actual bidder.

[01:45:23] You're not one set ahead of another person, but by Amazon in the form of a, quote, proxy second price that we calculate. So, it's a penny more than what we say the next highest bidder was. The FTC says... This is what you're willing to charge? This is what we think you should pay? Yeah. We'll go with the, we think you should pay. We think you should pay. We'll call it an auction. Yeah. How did you get caught? Who ratted you out? Oh, that's a good question.

[01:45:52] That's what I want to know. Who dry snitched on that? Because the markup on that is so insane and greedy. FTC says $20 billion. $20 billion since 2019. Over five years, because it apparently ended in 2024. And Amazon advertising alone is now bigger than the entire magazine industry.

[01:46:19] Yeah, that's something people don't think of Amazon as an advertising company, but it really is. That's probably, besides from Amazon Web Services, that's probably their number two revenue source is ads. And they're going to find like, what, $5 billion? Yeah, right. $20 billion that they made profit? Everything on the tech, I mean, all of the monetization, not monetization strategies, but all these tech companies, they're like, oh, I'm going to be an e-commerce platform. Oh, I'm going to do this. I'm going to do that.

[01:46:48] But they all come back to ads, which is one of their largest or second largest revenue sources. And it's just baffling to me that as consumers, many of us don't have a lot of money to spend on things. So I'm just like, when is this going to break? Surely it can't go on forever that reaching our limited supply of dollars is going to be worth as much as these companies are making.

[01:47:18] But AI is going to fix all that for us, Stacey. And that's why we have to win the AI race. AI is really interesting because if you have agents going out looking for things that you actually need as a consumer, then there's no reason to look for ads and there's no reason to buy ads if you're a company. You're way better off striking a deal with an agent supplier.

[01:47:42] I was just thinking about this because Amazon Alexa AI, the personal shopper, they announced that Alexa would start looking out for books from authors you've liked historically, things that they know you're looking for, basically. And like for Jeff's book, if I just said, hey, Amazon, just let me know when Jeff's book comes out. I like his stuff. His publisher doesn't have to buy an ad.

[01:48:10] They could just make sure that they update the Alexa API with the fact that Jeff's book's out. And that's just a really interesting kind of switch. That's a good way to do it, right? Well, it's a question of whether or not companies, a lot of ad companies are now rejiggering to sell to agents. Right. Well, that's the agent. So in Amazon's own PR briefing on this, that's exactly what happened. So they said, so we said that we were going to,

[01:48:40] we didn't want you to pay more than this, a penny more than the highest bid you beat. So, so there's an auction for an ad placement and you bid a dollar and somebody else else bet, bet 50 cents and you beat the 50 cents. But Amazon said, but good news, we're not going to charge you a dollar. We're going to charge you 51 cents. That makes sense in principle until they say, but then we realized this is in their own press release

[01:49:08] that we were undervaluing our ads. We could, we didn't charge enough. And they could have set the price for those ads on their own. They were free to do so. They should have just said what the price was instead of lying. Yeah, the issue is they lied. It's a deceptive ad practice. That's the issue. So they said, with relevant ads increasingly winning at prices below market value, we began to test a concept called soft reserve prices,

[01:49:37] a real-time minimum value that seeks to better reflect what the actual placement is worth. That's called setting a price, Amazon. Which again, they could have done. That's anti-capitalism right there. We also introduced what we called a hard reserve, which is the minimum a bid must have passed to enter an auction. That helps cover our costs. Okay, that's fair. But what they did is they used machine learning

[01:50:04] to figure out what the price was that an advertiser was willing to pay. Our auction looks at a combination of which ad is most relevant to the customer and the price, they'd say this out loud, an advertiser is willing to pay. Amazon is really the mafia. You know, I, again, being a hypocrite person, I have Amazon Prime, right? And I watched Spider Noir and now they have commercials where I've had it for years and I never had any commercials. And I was like, well, you get commercials or you're going to pay extra.

[01:50:34] And when I say- By the way, I immediately paid the extra. They got me. I did not. It's like you too. I'm not going to complain about people being criminals and downloading things somewhere else, but I'm just saying when I watch it and I suffer through the commercials, I refuse to even touch my phone or walk near my phone. I'm about to start taking the Stacey approach. Now, in fact, when I watch a show, I'm probably leave my phone in the other room because they love to do the, oh, what did you search on and leave in your cart? And put it up on the screen to remind you that you left it in your cart.

[01:51:02] And I'm not scanning no QR codes. I'm not clicking no buttons, nothing. I'm going to suffer through this minute and 30 seconds of trash that you're trying to sell me. And I'm never buying anything that you tried to sell me because I was getting it for free. And now you're charging me and trying to upcharge me. I refuse. And pro tip, if you leave stuff in your cart and don't buy it immediately on your little impulse buy, they come back in three or four days and was like, would you like this 5%, 10%, 15% coupon? The item in your cart has just dropped by 10%.

[01:51:31] You're dang gum right it did because I didn't need it. And I'm getting the savings because I'm on to your tricks phone. And I know, oh, I'm sorry. I think you should do dramatic reading of this Amazon's response to the FTC's lawsuit. I think it would be much more interesting. In Amazon's defense, they consider themselves a data-driven company. And what they're really saying, what they want to do is have that pricing set by data. Which again, they could have done. And they should just say it out front.

[01:51:59] This is how much it's going to cost. Yeah. I mean, Google, there was amazing about Google. When AdSense started, it was wonderfully counterintuitive. You didn't get the higher position because you paid more. You got the higher position because your ad performed. Yeah. Right? Because Google said that's relevant and that's more useful to the users. And so, da, da, da, da. Right? And so, they set new rules for how these auctions operated. And you can set the rules. You can set the rules.

[01:52:22] They do say, and this might be persuasive to the court, that the cost of ads, the average winning bid for sponsored products to ads fell 50% over the five-year period. Because advertisers are paying less for their ads. And they say, and this is the critical language. There was no harm to consumers as a result.

[01:52:52] The advertisers are paying less, not more. But in this case, the customer, the FTC is protecting. The FTC app doesn't require you to have harm. They don't? Oh, but the court might. Okay. Let's see. Don't you have to? Don't you have to? If you're going to? For antitrust, you've got to have harm. But this is not antitrust. This is deceptive practice. Does UDAP require harms? Let's see. Watch your policy walk at work.

[01:53:21] This is straight out of this week in Google. And you know what's funny? I'm so sorry. When you left, Paris came in and she does the same darn thing. There must be the researcher chair or something. The reporter chair. We just blather, Leon, me. Yeah, we don't look up facts. We just go, well, it seems to me. Well, obviously, I'm a rantor, so I don't know what's going on. I'm still worried about doing the fax machines in time. All right. Moving right along.

[01:53:51] It does require some form of injury or risk of harm. To who, though? To the consumer or to the advertiser? Well, in this case, the advertiser is the consumer. Yes, right. So I'll tell you what got me ranting this week. What got you ranting, Leo? I might have a counter rant because Owen may agree with this.

[01:54:13] Bernie Sanders, along with Greg Kassar of Texas, calling for a permanent ban on the development and deployment of artificial super intelligence and saying there should be jail time. Individual developers could face up to 20 years in prison for trying to circumvent the ban. It's OK, Leo, because there's not such a thing. Never will be such a thing as super intelligence. It's BS. So it's never going to exist.

[01:54:42] It's the Ban Artificial Super Intelligence Act. Baza. When he started reading that, Jeffrey, was that you? It's the Ban Pink Ponies Act. It may have been me. Oh, well, I love this idea, of course. I knew you would. Ban it all. I'm all for banning. You know what? And again, I'm wearing a sports photo. Somebody put in my Google earlier. It says I'm a sports person now. I still love technology because I use it all the time.

[01:55:08] But I hate when someone buys a team and they want to build a new stadium and taxpayers have to fund the stadium. And they say, all these jobs. Well, guess what? A stadium brings a hell of a lot more jobs than a data center does. And those fans still never get anything. If you want to use my taxpayer dollars to build a stadium, at least everybody in that area should get one free ticket to a game sometime in their lifetime. So they can actually go into the place. That's a good point.

[01:55:31] But you want to go get a data center, you're going to tell me that my pa Mabel that just had 140 acres taken from them are going to front of $40 billion. Let me be realistic. A $40 million tax-free exam for 10-year build on the backs of 20 people working in this data center and bringing in jobs. Well, this isn't a ban on data centers. We can argue about the ban on data centers. This is a ban that would.

[01:55:57] So the ban artificial superintelligence act would permanently ban the development and deployment of whatever this is, superintelligent AI. I don't know what that is. I don't know if anybody does. No one knows what it is. That's a made-up term. And more to the point, temporarily pause advanced AI development until a federal regulator has established safety rules. It would also direct the U.S. to pursue international agreements to prevent superintelligence from being developed anywhere in the world.

[01:56:27] I think Bernie Sanders must have seen a movie. Well, this is the problem with the Tesquerell AI boys and their ridiculous dreams. Well, there have been, and I understand this is Sanders responding to this drumbeat over the last two months of stories about AI escaping containment.

[01:56:48] There was another story this week about an AI from OpenAI hacking a German wiki to create their own AI message board. And I think that- They didn't hack it. Didn't they just share messages on it? Maybe, yeah, they didn't hack it. But that's where the language is going to be in media. You just, we really just need to clearly state that these companies are liable for harms, which is something we actually have never literally done. That's a good idea.

[01:57:18] So- Why can't I let Stacey run everything? Because when you say everything, not only does it sound appealing, but it's very concise and simple. But that's what I say about the fax machine thing. Bernie Sanders is out here saying stuff that makes everybody on this panel go, that's not what it is. But we get what you're saying. So I have a question along this line. If a human being had done what the OpenAI model did to Hugging Face, was there anything illegal about that? Yes, it's a violation of the computer fraud. Exactly. Absolutely.

[01:57:46] And this is not a case of an outside user using a platform that OpenAI can't control. This is a case of OpenAI's own model under its own control with negligence did this and that. But I, as you know, and we talk about this on intelligent machines, I believe AI is a significant major advance in technology. It is very, very important. Maybe the most significant advance ever since the invention of the personal computer. And I'm all for it.

[01:58:16] And I think it could be developed reasonably without harming the environment, without harming individuals. But I think it's very important that we continue to work on it. This is just nutty. Let me ask you one serious question, Leo. What in the utopia, if AI could do the grandest thing for humans, what would it do that makes it so viable and useful? It doesn't have to be curing cancer.

[01:58:42] It could be taking people like you or sole proprietors and helping you make it so that you can have a website in 15 minutes that's beautiful, practical, and effective. Could help you develop your business. Could help you write a marketing plan. It could be bigger than that. Well, it is. I'm just saying it could be little things. Think about GPS that allows you to track your kid. So it's, yeah, it's foundational in the sense that like the internet is foundational. GPS is foundational.

[01:59:10] It is going to be a technology infrastructure that allows things to happen at greater scale. Again, faster, faster, cheaper, easier. Again, in your wildest dreams, if you could snap your fingers and wish for one thing that it would make or fix or change, what would it be? Like, new drugs. New drugs. New drugs that could work. Yeah. Easy. Personalized medicine based not on like your actual, you know, what you do every day.

[01:59:37] We already know how to make cancer drugs that are specific to the cancer of an individual, but it's very, very, very costly. If you could sequence that and do that, that would be incredible. If it could be done at scale, it could cure cancer in effect. I mean, there's a lot of might that hasn't happened yet. We should point out. I'm not, and that's what I'm trying to get to. It's so there's a proponent of people that like AI is going to do all the things you're saying. And then there's other people that are like, dude, have you checked the economy in a job report?

[02:00:06] People are losing jobs and saying that AI is going to help me build a website or create my own job. There are no jobs. What am I building a website for? There's no one there to buy the job. So I, you're better off saying what you just said is for the people trying to sell. Like we can cure cancers or individual things or help like those things that it could possibly do just sounds so far fetched for the average person. That's why they don't want an extra data center in their backyard because like, what do you mean? People are losing their jobs everywhere. How is this going to help me get a job when you're giving them a tax break?

[02:00:35] I don't understand. And I'm losing water. It's just a weird dichotomy we have right now with the sales pitch of how AI is going to help change the world. And so, yeah, and what's happened, here's our issue. AI is built by and deployed by people. And right now the people who are building it and deploying it are motivated by basically returns to their private investors, not even the public markets.

[02:01:02] So if you think about it, and there's no legal repercussions for them at all, right? They can do whatever they want. So everything that people are mad about, totally valid. I mean, these people are, you're letting like the scummiest people in the world, like come into your house and, you know, poop in your neighborhood, basically turn your neighborhood into a cesspool. And legitimately they hate that.

[02:01:29] So what we need to do is realize that, A, AI is a technology tool. It's not like this magical thing. And it's built by people, deployed by people. So you have to hold those people accountable to it and build the types of guardrails and incentivize them to build the guardrails needed. And I hate when we start talking about it like this, it's this magical thing.

[02:01:55] Because that's how we used to talk about technology and even things like the internet. And really all it is, is a sea change. It's a technological infrastructure sea change. And we know how to deal with that. And also if we talk about it as if it's human to the anthropomorphization is the risk. Leo, if you want to see more nuttiness this week, go to line 107 if you wouldn't mind. Oh, Jeff, you're up to your old tricks. Why don't you do that? I am indeed. Oh, yeah.

[02:02:25] This is just nuttery. This doesn't really have any to do with technology at all. Oh, don't give them anything. Please, no. Just don't give them. I mean, because Peter Thiel already thinks it's a antichrist too. Yes. This is just Vance talking to Thiel. Yeah. The vice president said that AI is satanic. It's not just anthropomorphizing. But this is exactly the tension I've been talking about. Yeah.

[02:02:52] Between Trump's populist instincts and Vance's populist instincts, which got him elected. And their donors. And the fact that the other people that got him elected are Elon Musk and Peter Thiel. And the people who are promoting AI. And again, this third constituency of the intelligence community, the military industrial complex, all of which want AI. And I don't. This is to me.

[02:03:19] This is actually pretty cynical. My guess is that the Trump administration and it's the people who are running the elections said, look, J.D., you're going to have this is what the vice president always says. You're going to have to go out and you're going to have to say something. It's going to you may not agree with it. You may agree. But it doesn't matter. This is to appeal to our base.

[02:03:45] Now, unfortunately, this story got in the Wall Street Journal, which is exactly what Peter Thiel's reading. So I don't know if it was effective. The Trump administration officials are wrestling with the spiritual consequences of AI as the president embraces it as an economic engine. It's two. It's two. It's two mints in one. So it doesn't help either when our media complex does the same thing. The next line is Stephen Levy saying, who cares if AI is conscious?

[02:04:14] It's basically alive. This is such BS. Stephen Levy knows better. What the heck is he doing? And that's why I keep complaining. You would think that those people do know better, but they don't. And the way they reference it and the way they talk about it, you're not. You want people to embrace the technology and the people that are pitching it are doing such a terrible job at doing it and explaining why it's necessary. Someone in the chat said, China's going to eat our lunch.

[02:04:43] I have a MacBook Studio. I don't think it was made in the USA. The iPad, the camera, everything. They already ate our lunch. There's literally $100,000 in this room, and I don't think there's one thing that was made here. So, I mean, like, what are we talking about when it's like, we're going to lose? You know, I can't even get a BYD vehicle over here, an electronic vehicle that's worth a dang because Ford and Elon are blocking real change in technology because they've got us in a chokehold. So it's such a weird dichotomy, the game that we play with all this technology, and it's evil.

[02:05:13] It's this is that. So that's why I complain about it. That's why I grunt about it because we're all messed up right now. But we're human. That's what we do. Meanwhile, NVIDIA, who makes the hardware for all of this and is doing quite well. They're making money. They're making a billion dollars a day profit. A billion dollars a day profit. So they spent a little less than two weeks profit to buy Hugging Face. This has been going on for a couple of weeks, but it actually finally happened. Hugging Face is started off.

[02:05:43] By the way, I didn't understand this. So the name Hugging Face comes from the emoji where you've got a face and there's two hands. That's the Hugging Face emoji. So that's where it comes from. JJ, let's all do it so we can get it. So Bernardo can get a card. Yeah, everybody do the Hugging Face. Come on. You can do it. You're trying to go against our overlords? You're trying to get me in trouble? No, no. You're welcome. It's just for the thumbnail. You know that. Don't you, Owen?

[02:06:11] If you're going to get downloads on YouTube, you've got to go. I know the game. You know the game. Anyway, Hugging Face, which started, I didn't know this, as a chat service. And that's where the Hugging Face came from. I never understood where. I thought it was just two random words. It's one of those great stories of a pivot. And they ended up being the kind of the. Remember Downloads.com and Two Cows, the places where you'd go to get shareware?

[02:06:39] They're basically that for AI models. They have more than 3 million downloadable AI models. Yeah, they're two cows. And they're Download.com. And they just sold for $12.9 billion to NVIDIA. Congratulations. After turning down an investment from NVIDIA, was it six months ago, at $7 billion? Wow. The irony is, it's not like those models don't exist elsewhere.

[02:07:06] I guess it's sort of like Microsoft buying GitHub for a little bit less money, I guess. But there's nothing. I mean, there are other places you can go to get models. So I don't really. But what I do like is the notion that maybe NVIDIA is going to get behind open source or more properly open weight. Yeah, I think I've been way behind it. And this is where I need your schooling, Leo.

[02:07:35] And I was trying to get it this morning on chat. Is the value of the harness. I called it a saddle this week. The harness. It's a bridle. It's a bit. It's a saddle. It's spurs and stirrups. Versus the underlying model. And is the underlying model just the steam engine? And is the value going to be in the harnesses and how those operate? And is that kind of where Hugging Face can play? Is that we're going to help you? It doesn't matter what model you use. It's all interchangeable underneath.

[02:08:05] But there's a layer above. The geek's best answer. It depends. Yeah. Maybe. Sort of. Well, the harness, which could be like Claude Code or Codex or Antigravity or Cursor. Or the agent I use, Hermes or OpenClaw. That's the software that surrounds it that calls on the AI model and feeds it context, gives it information, gives it memory and all that stuff.

[02:08:33] And the AI model then gets all of that as a simple string of text characters, tokens that go in. And then tokens come out of the AI model and the harness gets that, displays it on the screen and, you know, kind of puts it out in the world. Actions on it. But the model still does a lot more than just simply. It's more than a pachinko machine. That's what Benito likes to call it, a pachinko machine or a wind chime.

[02:08:59] It's more than something you put information in and it goes, and it comes out. The people who build the models are also doing a lot. But you're using them practically interchangeably. I feel like they're mostly interchangeable. They've gotten more and more similar. You know, people remember when OpenAI killed ChatGBT4.0? People were so sad because they had fallen in love. This was their virtual boyfriend and girlfriend.

[02:09:29] And they said, you're taking away our loved ones. It was because, and OpenAI kind of fixed it. OpenAI had made that model, independent of the harness, made that model be very personable. And that's in the weights. So you take the model, you train it, and then you adjust the weights so that some stuff happens more and some stuff happens less. And all of that's the post-training.

[02:09:59] And they had post-trained it to be very friendly and personable and lovable. And they realized, oh, this is not good. People are getting too over-involved with their models. So they actually took a lot of the personality out. And that upset people quite a bit. So it is everything. You know, there's more than just the training. There's a lot of post-training and reinforcement learning. And it's also the context you put in. You know, every model, when you start it up, you have a file, a text file. You can read it.

[02:10:29] In most cases, it's called soul.md that tells it who it is. You know, you are a smart, you know, social scientist or you're a great writer or whatever. And that kind of also shapes the outputs. Because it goes in and the outputs are modified and it goes out. But it shapes what happens. So it's a complicated, it's a very, it's almost a chaotic system. That's part of the problem with it. It's probabilistic, not deterministic. It's hard.

[02:10:58] Different, you put the same inputs in at one point and different ones will come out. Unlike a computer program, which pretty much will always produce the same result, given the same input. It's supposed to anyway. How can I talk NVIDIA into buying America? They could buy the Lake America. They could buy the Gulf of America. I don't think America is available. They're making a billion a day. We're 40 trillion in debt. They could just buy us.

[02:11:27] He could just buy us. By the way, that's why the administration doesn't want to piss off these guys. Because, yes. Yes. Right? Because it's so insane. Because again, like all the stuff we're talking about, NVIDIA is just making money. They're the ones making the money. Forget the fact that when you're building the data centers, half the time, by the time they get built in the two years, the chips are going to be different in that point anyway. And other companies are going to be wanting to refresh their chips and update them.

[02:11:55] That's the biggest thing when somebody says that China is coming to eat our lunch and stuff like that, like we talked about earlier. We're forcing China to find their own manufacturing, their own chips that they want to create. And if they get on NVIDIA's level, that's when you're really like, oh boy. Because we don't have the materials to build the things. We don't have the infrastructure to build the things. But NVIDIA could just buy us and fix everything. You've been talking, Jeff, a lot about anthropomorphizing AI. And a number of people have said that this is sort of the problem.

[02:12:24] One of the consequences is the Bernie Sanders bill. The other one would be the president pulling fable from the marketplace. Treating it more like it's a human or it's acting like a human than it's a computer program. I just was saying to Darren Oki in the chat, it leads to the mysticism we see. This is a strange power. And I was talking to my colleague at Stony Brook. We need new vocabulary for this. It doesn't think it calculates.

[02:12:54] Well, that's one reason I use the word agent more than artificial intelligence. Because artificial intelligence is by itself anthropomorphizing it. Like it's intelligence. Right? It was a marketing term John McCarthy came up with in 1956, 55, hoping to convince people to the government. It's very hard. I mean, Dworkish Patel in discussing the hugging face. He highly anthropomorphized. He highly anthropomorphized.

[02:13:22] He talked about AI civilizations rising and falling. Oh, yes. And he gave Gary Marcus just material to work with. But I sympathize to the effect that to the limit that it's hard to discuss what these things do without using human verbs. Yes, because it looks like they're thinking. And in fact, I'm not convinced that we do anything that special. Kind of. Speak for yourself.

[02:13:47] We're kind of tokens in, tokens out based on our context that we've learned over the years of our lives and from our parents and our schooling. And we also have a little bit of, I don't know, is it intuition or something stuck in there at birth? Listen, I mean, we're not that sophisticated. How history gets things wrong. In the Gutenberg parenthesis, I wrote about this.

[02:14:11] And the argument in that book, which is a fascinating book, is that the theory of mind is BS. Yes. That we don't really have this diagram of structure. That instead, it's more like we have, and this is my phrasing, a whole bunch of videotapes in our head. Well, what did I do last time I did this? Oh, I did that stupid thing. Okay, I'll do it again. Right? Because that's what I did before. And we have ruts that we're in rather than reasoning. In 1978.

[02:14:39] I'm so glad you brought that up because I was going to break up our theory of mind. So good job, Jeff. Stacey, I think you'd like that book a lot, How History Gets Things Wrong. I'll check it out. It's classic. In 1978, I took the, it should have been a comic book. It would have been a better comic book, but you know. Oh, no. You can't have everything. It was a lot harder to read. I had to take my time on this one. A lot of words. No pictures. Yeah, words, words, rats, experiments, all kinds of things.

[02:15:08] In 1978, I took something called the EST training, which was at the time, you know, that was this human potential movement. Earhart seminar training. Excellent. Earhart seminar training. In the examiner, I was mocking it, but keep going. Well, everybody mocked it. Yeah. They don't let you pee. It's two weekends. It's very intense. It's a lot of stuff. And towards the end, something happens and you're supposed to, and this is the phrase they use, get it.

[02:15:37] And then, so you're at that point, there's a trainer leading this, there's 200, 300 people in the room and you're all supposed to be getting it. And I'm looking, I'm going, I don't really get it. Oh my God. This is the landmark guy. Yes. This is a cult. Well, it was cult-like. It was cult-like. It's a cult. No, no, no. It wasn't a cult in the sense that they didn't keep coming at me for money and I had to go to a special place. Oh, but there were particularly loyal followers. Oh, there were some people who got into it that way.

[02:16:07] They were cultish. I found it very valuable. You know, I was 21. It was something very valuable. But. I was about to ask how old you were at the time. I was 21, but, and it changed my life in many ways. And how I, how I, how I thought about things in many ways. But recently I got it. I realized how many years later, 50 years later, that what they were really saying is we, and they, this, it was actually explicit. I just wasn't paying attention. We're just robots.

[02:16:37] We are just act, just as you just said, Jeff, we're just playing back the tapes in our head of previous actions. It's why we keep doing the same stupid things over and over again. We keep making mistakes and go, why did I do that again? Because we're essentially automatons in many, many ways. And the point of the S training was to understand that so that, you know, you can kind of break out of the loop.

[02:17:05] Therapy is probably, good therapy is probably a better way to tackle it. It's like therapy. It's very similar, right? The idea was, well, what if we could do this in two intensive weekends? But you're right. I mean, it's very similar to therapy. And it's, it's, it's not inconsistent with therapy as well. I mean, I've done a lot of therapy too. But, and that's, that's, and that's the same thing we do with technology because people think they're being rational, but they're just reacting to things. And they're like, no, this is totally rational.

[02:17:35] And you're like, eh, is it though? Why do they keep putting money in a slot machine? Yeah. And the same thing with AI, we keep thinking that this is based on data. This is data driven. This has no agenda, but it, it absolutely does. Yeah. And it's just feeding back to us, our own. Yeah. It's us. Or the creator's biases and what we want to see. I think I'm a higher being than what you guys are saying. I have freedom of choice and free will. My favorite other story to tell people is, uh, I had a friend in high school.

[02:18:04] His dad was a drunk. He had two sons. One son grew up to be a drunk. The other one never drank any alcohol. The one says, oh, I have to drink. Our dad was an alcoholic. And the other one says, I'm never going to drink. Cause our dad was an alcoholic. Those two kids made choices in life. But there's also genes involved. Well, my response to that would be, yes, they made choices. They may have been different choices, but they made choices in an automaton response to the stimulus.

[02:18:31] There were different choices, but it wasn't because they used free will. Well, it was because they were triggered by that initial experience. And that's why we have therapy because this is a free will or is it like a gene thing? Is it genetics? Am I genetically predisposed to be? I'm not. I believe in free will. I think we have free will. Yeah, we have free will. Well, I guess the idea is, did they make a considered decision or did his videotape of his dad's drunken behavior make him work one way or the other, as opposed to saying my dad was a drunk.

[02:19:00] It caused me the following. Right. I don't drink, but my dad was not a drunk. So. Or I take opioids and I get constipated. Other people take opioids and it changes their life. Right. Because there's a different receptor in there's genetic differences in us. If I die, if I'm on chatbots and this is a chatbot conversation, you are in trouble. That's all right. You're watching. I don't know how we got very philosophical here.

[02:19:27] This is good because we have deep thinkers on the show today, including Owen JJ Stone, who does believe in free will. I do. I do. I do believe in free will. Oh, look, he's got Saul Hank flaky. You can't even get that anymore. And the Calabrian flavored. Uncle Leo, when my cousin puts out a product, I said to myself, let me put these in the freezer. They might be worth money one day. I think they will be. The original. It's like the Archie comic books that my grandmother threw in. Not me.

[02:19:56] They have one thing in common with those Archie comic books. No one's making those anymore. Nope. I told you they were there. I'd never tell you a lie. I didn't finish the story. Steve Gibson's wife called the people who made it and said, would you make me some more? No. And they said, no, we can't make some more. It's Saul Hank's. And she said, well, if I get Saul Hank to say so, can you make me some more? They said, well, okay. So she calls me. I call Saul Hank.

[02:20:23] Saul Hank says, calls the lady, said, yeah, you can make some for her. So they made 20 more bottles. And I said, you're kidding me. No. That's it. Don't, don't, don't you be calling me Owen JJ Stone. You want to preserve the rarity of those. That's right. I think I would have finished that story, Leo. I think I would have let that story just die a death. Now they're coming after you. Now, now you've opened up a door. I should explain for those who don't know what we're talking about.

[02:20:52] Saul Hank is my son. Well, I'm very happy to say both on the CBS this morning, on Saturday and in the Wall Street Journal at the very top, they say Henry Laporte known as Saul Hank. So at least the family name gets mentioned once. I like it that he didn't use my name at all. Right. There's no, nobody knows he's related to me. Anyway, he's very successful. Got a restaurant in New York. Was at the U.S. Open tennis tournament. Sold half a million dollars worth of sandwiches in five days.

[02:21:20] They were doing 2,000 sandwiches a day. So he was very happy. Maybe I'll run up to New York and get him to sign my OG canisters. I actually, he has now sold the name and the brand to Wonder, Mark Lohr's company. And, but he's going to continue on and they're going to keep the original store and everything. So I'm going out in a couple of weeks. Going to finally go out because I've never had one of his sandwiches. But you're not on that.

[02:21:50] Are you on the GLPs and that kind of stuff that you kind of can't eat it? Enjoy it? I can't eat it. I may not enjoy it. Oh, I see. Okay. No, I'll enjoy it. I just won't eat all of it. All of it. But it's a lot anyway. It's a lot. Yeah. It's very rich. I bet it is. I think I might have one bite and go, well, that's delicious. That's probably all I can have. I'll eat the rest. I love a good French dip. Oh, it's the best French dish to be ever had. Paris and I stood in line and went there. It's incredible.

[02:22:20] We didn't use our clout to get in early or anything. We stood in line like civilians. I will put a link. I'm using my clout. I'm showing up to Hank's my cousin. Everybody, who's this guy? Oh, no. He told me. I said, because Wesley Faulkner was going. I texted him. I said, how do I get Wesley in? He said, don't bother me. He said, just tell them at the front. They've been told. They've been warned. If somebody says, oh, I shouldn't say this out loud. Don't say it out loud. It's a secret word. And you don't know it, people. Tell me off air until I get there.

[02:22:50] You know, I need these things signed. Oops. Reel it in, Uncle Leo. I came very close. Now everybody's going to want sandwiches and salt. Give it away. The secret word. It's almost like you haven't done 1100 shows to not know how to act. I have no. I'm telling you, we're just automatons. We're machines, my friend. You haven't learned anything. I have learned nothing. This is why he doesn't do NDAs. You're the worst agent model available. You're fired. I am. Fire me. I am.

[02:23:21] By the way, that's what Astra said too. I said, Astra, what happened? He shouldn't have done that. He shouldn't have done that. I'm not kidding. The final thing is, don't do it that way anymore. It says, require useful progress after 10 minutes. Give coding tests a focus session. Reconsider the settings. I would not repeat this unrestricted high reasoning setup. Okay, fine.

[02:23:50] Too long of a leash. Stupid model. Serves you right. Reel it in. Reel it in. The assignment was substantial enough to require care, but entirely reasonable. Managing that complexity was the agent's responsibility. Oh, thank you. At the end, after saying all the things I shouldn't have done, it was the agent's fault. You know what? That's casting shade on your co-worker. Knock it off. Yeah. This is Twit, This Week in Tech.

[02:24:19] We're glad you're here. Thank you for watching. Have a happy Labor Day, everybody. We'll break with a short plot twist for your glass. Erdbeere can more than dessert. Now new. Granini Selection Erdbeere. Fruitful, pure, as strawberry matcha latte or in Spritz. And more fruit taste? Discover the whole Granini-Fillfalt. McZini, Habini, Donzini, Friendini. Cyber Cab launched. Woo! Woo!

[02:24:48] Did you see all the celebrations? All the excitement on September 4th? It was very quiet. Oh, God. Why wasn't he on a stage? I saw a video about people in the wheelchair. It's very quiet. And it's already under investigation. As it should be. This is Elon's Tesla that has no steering wheel and no pedals. The government says, unlike Zooks. Remember Zooks? The one in Las Vegas with no steering wheel, no pedals. Did you end up running in one?

[02:25:18] I didn't get to. I couldn't. They were always booked up. They're very popular. They went to the National Highway Transportation Safety Administration. They said, can we do this? They went through all the hoops. They did all the things. And they got permission. That's not Elon's way. Nuh-uh. So he just started picking up people in Austin two days ago. You know, I'm not going to put my life in his hands in that vehicle. But the problem is, it's not my choice. That thing can run me over. Right.

[02:25:47] Stacey, he rolls up in a wheelchair, then stands and puts the wheelchair in the trunk. Was this made with AI? What is going on? Where is that? I got to find that. Is this commercial great? Did the crutches? I think the guy has the crutches. I've never seen it before. I just Googled it. And I'm like, wait a minute. I'm trying to listen to Leo and look. And I'm like, wait a second. Did you put it in the Discord? Put it in the Discord if you would. Hold on. They gave me a link. I'm watching it on Facebook.

[02:26:17] I'm really curious to see this. Larry in our Discord went to the U.S. Open and had a Salt Hank. But he's already had one before. It wasn't his first time. Someone find me this commercial on YouTube. Oh, I keep finding it on. Oh, I could find it. It's called a Google search Tesla robo taxi commercial. Just do Cyber Cab wheelchair. I bet it pops right up. Oh, yeah. There you go. Cyber Cab wheelchair.

[02:26:46] So this was put out by Tesla. So this is put out by them. They filmed it at their local factory, I guess. Okay. Here it is. It's on their X account. There we go. So he rolls up. Okay. No problem with that. Yeah. Or goes up in crutches. Whatever. Yep. Yeah. Puts everything in the trunk. It serves all types of people. Puts the luggage. Puts the wheelchair in there. You stand there. How is he standing? Puts the wheelchair in there. Walks around in the front. Gets in. Nope.

[02:27:15] We don't see them walk around. We just see the car. There's no driver there, though, to put the wheelchair in the trunk. How do you open the trunk? Yes. All the... What? What? This is great. You're like, hey, what if I'm in a wheelchair? And they show you this and you're like, wait, so what if I'm in a wheelchair? I got to tell you, this is definitely generated by AI, right? This is an AI commercial. This is AI for sure. The color. Because first of all, he rolls up in a wheelchair. Then he walks up with luggage. Then he walks up with crutches.

[02:27:45] That's the logic of AI. If I had the crutches, I would just put them in the front seat with me because it's so open and there's no brakes or steering wheel or anything like that, I can just take the crutches with me. But why does he... Unlike the Zooks, it has a front seat and a back seat. If there's no front or... If there's no steering wheel or pedals, why make it look like an automobile? Oh, because it is an automobile. They just took the pedals and the brakes out.

[02:28:08] NHTSA is investigating because literally hours after the automaker shipped, they are updating their standards to make it easier and faster for driverless cars to deploy, but you still got to go to them. And... If anything should be regulated in technology, this is it. I think this should be regulated. Yes.

[02:28:32] Tesla says it doesn't need an exemption because the vehicles comply with all existing safety standards. Okay. But it doesn't have a brake. I think that's a safety standard. I might be wrong. And they don't even use LiDAR. They're just using cameras, I believe. They don't even have the full range of the technology. That's what bothers me. And I'm... It just... Again, this dude, because Uncle Fester's his buddy, just feels like,

[02:29:00] I can just go do whatever I want to do and push it through. No, you can't. This makes no sense. And what if I get stuck in this car or something happens? Like, I'm not really too worried about it, like, running over someone, even though that is a possibility. But just the fact, like, if something goes wrong and malfunctions, I have no way of physically turning a wheel to help myself. Like, that would freak me out. So they're doing what they call an audit query to examine the data and process Tesla used

[02:29:26] to make the determination that they don't need to get an exemption. They kind of want you to do that. They kind of want you to apply and explain. You know what? I'm not going to trust. It needs consumer reports. Yeah. CR needs to get in on this. Yeah. I know people complain about Waymo and, like, when things would go wrong or something like that, you know, it would drove you out of town or they got stuck in traffic in front of each other. At least there was still a wheel and a brake.

[02:29:56] Like, there's something there that I could get to. You know what I mean? Like, my one friend's like, oh, you sit in the back of a plane and you don't have a steering wheel or... Yeah, but there's also a co-pilot and then a backup pilot. I like it. We don't need safety standards. We don't even have drivers. What do we need safety standards for? Yeah. Well... Woo! Okay. And then Ray Kurzweil... Actually, before we do that, let me just do one little plug for Club Twit because I

[02:30:22] want to thank all of our fabulous Club Twit members who make this show possible. You know who you are. Ten bucks a month and you got, I think, some benefit out of that. I hope you do. You get ad-free versions of all the shows. You wouldn't even hear this begging commercial. You also get chapter markers in all the shows so you can jump around if you want. You get access to the wonderful Club Twit Discord, which is a great place to hang out with smart people. Not just talking about the shows, but talking about everything geeks care about.

[02:30:52] No, we also do some special podcasts, but I don't think we're going to do Salt Hank this week in Salt. Although, Hank really, from day one, he said, I'm taking over Twit. When you're done, I'm taking it over. So it may be, this is... I don't know if we're going to put salt on our pumpkin pie, but it may be. I guess a pumpkin. Snacks. Yep. Snacks. Absolutely. This is the kind of thing that's from the club.

[02:31:20] You also get lots of special programming. Stacy's Book Club, as a perfect example. Our photography group, our AI user group, which we do twice a week, twice a month, rather, now. And we cannot do the Apple events, the keynote events in public anymore because Apple doesn't like it. And so we do this as a private event just for our Club Twit members. So on Wednesday, 9 a.m., we're going to get... Actually, I think it's 10 a.m. We're going to get together.

[02:31:47] Maybe I'll get there earlier with Micah and I, and we'll watch together as Apple announces its new phones, give you that commentary you expect. But because we didn't want Apple to be disadvantaged compared to all the other companies, we do all the keynotes that way now in the club only. But most of... That's the only thing we do that's club only. Almost everything else ends up on YouTube where we stream it live as we're doing it, like Micah's Crafting Corner, Chris Markwert's Photo Time. Micah's doing a media club, too. We're going to do...

[02:32:14] On September 18th, we're going to do the top animated episodes from Rick and Morty, The Simpsons. We're going to do the Marge and the Monorail episode. That's going to be a lot of... I can't wait. And that crazy show, we're going to do it September 11th with Jeff Atwood off by one. We have some very interesting viral video to show you. That's all I'll say about that and the AI user group. Join the club. Ten bucks a month, lots of benefits, but most importantly, the warm and fuzzy feeling of knowing

[02:32:43] that you're supporting the programming here. If you like what you see, if you love our programming, if you listen to it every week, please join the club. We really, really appreciate it. Computronium's creator, Baker Swile. Would you explain that to me? We interviewed Ray. I think I like Ray. I've interviewed him many times. He has been for years saying the singularity is getting closer and closer, by which he means

[02:33:13] machines get smarter than we are and more importantly, design themselves so that they can accelerate at a high speed. And at some point, yes, we will merge with the machines. He's highly respected. He created the synthesizer for Stevie Wonder. He actually, because he was friends with Stevie, created the Kurzweil speech synthesizer. The first read.

[02:33:41] I mean, he's a brilliant, brilliant guy. And among other things, he is a Google fellow. Now he's joining a startup developing a brain computer system that, according to Fortune, involves snorting nanoparticles. Now I don't subscribe to Fortune, so I can't get any farther than that headline, but I'm not sure I want to. Jeff, did you read it? Nope. I don't subscribe either. And I'm not snorting anything, or I'm not letting Elon put a chip in here.

[02:34:11] I'm not snorting anything in here. Well, better snorting them than in surgery, right? I'm not so sure. Yes, if those are your choices. Sure. Although, you know, injecting it might be a better option. Than snorting it. And what are they going to do? They're going to go up into your... I mean, are they going to go in... But like, for brain computer, like actual brain computer interfaces, putting it in the right spot is actually really important. So...

[02:34:40] But they would know where to go, wouldn't they? Like, are they magnetic? Are you going to like drive them? No, they would know because they're smart. They're smart nanoparticles. Okay. That is just a bunch of science fiction words together. I mean, they're real things. Nanoparticles exist. But like, what type? What are they made of? Do they have chips in them? Are they just really tiny bits of metal? Plastic? We have all of the options.

[02:35:07] I think Ray, who's getting on, deserves a little bit of peace and quiet. And if they want to use his name to promote their nanoparticles, God bless them. You want to see my phone? Yeah. Oh, yeah. The pinky? And I realized my last phone. Look, guys, what does this remind you of? Oh, yours was a lighter color, Rose. Now you've gotten deeper and darker. Do you like it?

[02:35:37] That's the new Pixel 11. Yeah, I do. I think it's actually a little lighter. And you know how I hate a heavy phone. You do. But it is... Because you've got kind of floppy hands. Is that the larger size? We forgot all the floppy hands. Is that the larger size, Pixel? No, this is the smaller one because I hate big phones. Because she has floppy hands. She can't. Well, I can't tell hand size, depth, perception, you know. Yeah. It looks... It's hard to tell, isn't it? It looks like a big phone. It does.

[02:36:07] Well, it's the same size. It's a little... It's a smidge larger than the 8. I'm trying to think. Do I have anything that looks like a... Put it next to a banana. Do you have a banana there? I used to have... A hero standing size. Yeah. I used to have a giant pencil that I would put things next to it. I used to have a card or a credit card or a coin or something. I used to have a six-foot long pencil. Here's a photo. This is a... What's that? 4x6. It's an old photo.

[02:36:36] Everybody knows that. Oh. It's a photo. Back in the day... Oh, it's the size of a 5x4 photo. Yeah. Yeah. Okay. It looks pretty. You know what I have? That is hibiscus. What else do you have? I have a ruler. I have a little pica. Oh, see, that's even better. My dad used to carry a ruler. He's a geologist, paleontologist. When he was out in the field, he would always carry a geologist pick,

[02:37:04] which is a hammer with a pointy end at one end. He would... And a hand lens and a ruler. So it's about... It's five and... It's a little... It's almost six inches long. And then almost three inches. Which is, by the way, completely meaningless. I don't even know what six inches is. I have no idea. Dude, I have a ruler. Your hand is six inches? Is it? I'm making up. Didn't they used to say stuff like that?

[02:37:34] That's six inches, but it's not really. My wallet has a little centimeter scale in it. By hand to the tip of my index finger is six inches. Yeah. All right. I got the reference. I appreciate your help. Stacey, what you're working on these days at that old Consumer Reports there? What am I working on that's exciting? I am still working on my legislation.

[02:38:00] It was introduced last legislative session in three states. Right to repair? What is it? Oh, it's my Connected Consumer Products End of Life Disclosure Act. Oh, that's right. That's huge. If anyone has a better name for it, feel free to send me an email. And we're going to be reintroducing it in California, hopefully New York, maybe Massachusetts and Virginia and possibly Oregon. So these are nice. This is what I'm doing at the moment. Yay.

[02:38:30] Yay, me. Congratulations. And then in October, I have a big report and a big research project coming out. So yay. Do you want to talk about that? Or is that a secret? I think they're like embargoed or secret or something. Okay. Good. I'm really bad at that. We love Consumer Reports. You're better than Leo. At least you caught yourself. The secret password is... Yeah. No, we love Consumer Reports because, of course, Paris also works at Consumer Reports, Paris Martineau.

[02:38:57] And we have Nicolas De Leon frequently on the shows. He also... So it's a great group of people. How long have you been working with him now, Stacey? Three years. Oh. And you're happy? You don't want to come back to podcasting ever? I'm actually contemplating starting out my newsletter again. Oh, good. Oh, good. That'd be great. You did such a nice job. We didn't mention this GigaOM thing, but when OM passed, everybody loved OM Malek.

[02:39:24] We did, of course, talk about it, but I would have loved to have got you as one of the original GigaOM folks on. You wrote a lovely piece. I guess you kept Stacey on IoT around just for that. I kept it around as an archive, but yes, it was a good place to put that. And if you go to Stacey on IoT, it's probably... It's probably the front page still because I don't think you... Yeah, it's still the story because I haven't written anything else for there.

[02:39:52] Do you want to say anything about OM or have you said it all already in this story? I feel like I said... I mean, we lost a really good dude. He was a really sweetheart. A really influential... Yeah. He was not a sweet man. I mean, he was... No. He was a good friend. You know, I never worked for him. He was always great on the show and lovable and cuddly, but I know the people who worked for him knew another side. Well, no, he just...

[02:40:19] He did not accept bullshit of any nature. Yes, yes. I'm not saying that he was like a cruel or mean person, but he was not... He was demanding, right? He was demanding, but he also... He was... He knew... I don't know. He was very good at understanding the human side of tech and all its... For good and ill and how people use technology. And he was brilliant.

[02:40:49] And when you didn't live up to that brilliance, you know, he told you. And so... But he would give you the shirt off his back quite literally. So he was sweet in that sense. But like, he's not like... When I think of a sweet person... He was very patient with me then because he could have easily... He probably was. A lot more trouble. I miss having him on. When he had his heart trouble, we had him on less. Just like you, Owen.

[02:41:18] You know, when people have health problems, the last thing they want to do is spend three hours on a Sunday evening talking about nothing. You know, suddenly priorities become more important. Anyway, loved having Om on the show. And we really appreciate... I really appreciated the lovely piece you wrote at Stacey on IOT if you want to read that. Thank you for being here, Stacey. We appreciate it. Thanks. Y'all want to meet my new dog? Yeah. Oh, yeah. I don't know if... I don't know how this is going to work. Let's see what happens.

[02:41:47] Let's take a ride. Oh! She's so sweet. Is she a border collie? What is she? She's a mini Aussie. Aussie. She's a beautiful little Aussie. Blue-eyed Aussie. She's not gonna... Hopefully she won't be as loud as my other dog. Lisa wants an Aussie. The problem with them is they're very smart, right? And they need attention and they need to work out because they're used to nipping the heels of sheep. Does she nip your heels? She does not. She's seven.

[02:42:17] I'm teaching her to dance, though, to entertain her. So she's... What do they call it? She's not a senior dog. She's a middle-aged dog. She's a senior dog. Okay. She's not a kid. She's a sweetie. She's not a kid. You rescued her? We rehomed her. Nice. She's a kid. She's a kid. Then she didn't. Now she does. Yeah. From her perspective, we kidnapped her from her loving home.

[02:42:46] And brought her to our loving home. Put yourself in her shoes. Yeah. What the hell? These people used to feed me. Until AI can make a loving companion like a dog, it'll never be real enough for me. That's really... That's a good point. I don't care what you guys say about choices and videotapes. Nothing to... Create a dog. Create the love of a dog. I double-dog dare you. No, you're right. Artificial intelligence. No, you're right. Make something that makes my heart melt. I'm a hardcore person. I just saw that little puppy dog. I don't care about that dog.

[02:43:15] But I still say, oh, because, you know, it's a dog. So, yeah. Go ahead, AI. Now, we do have a tradition around here. I will tell you who Owen is in a moment. But Owen has... We have a tradition. Unlike anybody else who's ever been on this show, Owen gets to close the show out. And I don't know. You haven't done that in a while, Owen. Are you prepared to do that? I'm always prepared. It's in the contract. Okay. There's only one other person who's talked about Twits Can, and that's me. That's you. You're the only one who's ever...

[02:43:44] Nobody else, even these highly respected longtime colleagues, have never closed the show out. Jeff Jarvis, a real pleasure. A hot type is selling like hot cakes. Everybody should go to Jeff's site. Jeff Jarvis.com. Get an autographed copy. Jeff Jarvis.com. We'll see you back here on Wednesday for Intelligent Machines. I don't. This week, no. Oh, that's why we're doing this now, because you are going away. Where are you going?

[02:44:08] I've got, well, Status, which is a wonderful site in New York, Oliver Darcy, about media and all that is having its first event. And so I'm going. Oh, wonderful. Time's center. And Benito, have we replaced Mr. Jarvis yet? Or are we irreplaceable? Don't you know that? Oh, that's right. Because Micah's got a new AI show. Exactly. Micah Sargent will be found. Micah was almost the immediate host in last week's show because Leo lost his power.

[02:44:38] Oh, I know. I know. But fortunately, the reason I, okay, so this is all connected because I missed your book club, Stacey, because the electrician was out here so that I could put a battery backup in for the system. Oh, good. Turns out, I thought I was going to be down all afternoon, so I was going to miss all the shows. I didn't prepare. He came in. He said, I already did that. Don't you remember last time I was here? I put in the 20 amp circuit and the special plug.

[02:45:08] It's right there. I said. You just didn't flip it? No, I just forgot. No, it was sitting there. I've been using it. I just. I know you didn't read the book and you just needed some sort of excuse, but this one really is not a great one. He came right during the show. So I skipped the show because he was coming to put in the plug. I guess once he said, it's right there, Leo. And that'll be $250.

[02:45:35] I could have probably come on the show, but by that point, it was so humiliating. I just hung my head in shame and went about. Would your power have kept going? Were you plugged into your. No, my power would not have kept going. That battery that I put in. Kept us alive. No, no, no, no, no, no. That's what I'm asking. You thought you only had 45 minutes. Were you hooked up to a bigger battery than you knew? No, no. No, that's another cause for embarrassment.

[02:46:05] The power outage was 16 minutes. But when the power came back, I guess it flipped a circuit. So I didn't get power. And I thought we were off. But we were on. And the battery is even getting juice, but it was just the stuff behind me that wasn't working. So. I'm just an idiot. I really am. Leo does shows about technology, folks. Electricity.

[02:46:34] They all depend on electricity. But this week. I got it. And now, ladies and gentlemen. It's on AI, folks. We thank you so much for being here. We do twit every Sunday afternoon, 2 to 5 p.m. Pacific, 2100 UTC, 5 to 8 Eastern. You can watch us do it live on YouTube, Twitch, X, Facebook, LinkedIn, or Kick, or download shows from twit.tv. We have audio and video at the website. There's a YouTube channel dedicated to the video, of course.

[02:47:01] And the best thing to do, because it's a podcast, and God bless it, you can subscribe. It's just an RSS feed, and you'll get it automatically in your favorite podcast client. Just subscribe. You'll get it automatically the minute we're done. And now, I give you the Reverend Owen J.J. Stone for our closing. Benediction. Our benediction. So it's almost like you read my mind because you were talking about the power going out.

[02:47:27] Now, a lot of people across the country have lost power, and I'm not blaming data centers. I'm not going to talk about the wildfires that go on across the globe. There are landslides in Nepal. People have lost their lives because of earthquakes right now. The climate seems to be changing in a way that it might affect you. So as much technology that you have, make sure you get yourself a life straw so you can drink water. Get one for each one of your family members because you need a gallon of water a day to survive. I got a bug out bag. I don't know if you got one, Uncle Leo, but you really need to have it.

[02:47:56] You need to have a life straw. Jeff Jarvis books so you have some reading to keep your mind straight. Something for lights. When the power goes out, you know, a flashlight will help you out, but you don't want to be burning candles. There's lights that recharge, and when the power goes out, they give you like five to six hours of power. Start looking into things that might go wrong in the household, and check with your insurance companies because they like to tell you that you got flood coverage, but you don't have flood coverage. And I know a lot of people talk about drought right now, but it seems like there's a lot of people's cars floating down the street, especially in my area.

[02:48:25] So check your insurance. Get yourself a bug out bag. Two of the most important things I have in my bag are vacuum sealed cigarettes because in the apocalypse, when the zombies come, I'll tell you what, I can get a lot of work done with cigarettes. You're not in prison anymore, Owen. I don't know how much I can get done with them gold bars I got stashed up, but let me tell you something. Make sure you have that stuff together, and if you have older people in your lives, make sure that you have something you can help them and get them prepared.

[02:48:55] If you don't want to have a bug out bag for yourself, go drop an extra flashlight and a solar radio off to Grandma. You want to know what's in my bug out bag? My NAS and my Vuvuzela. Uncle Leo's old enough that if he doesn't survive, it's okay. But everybody else who weren't in a cult in 1978, we might want to survive. So if you don't have a bug out bag, you can email me. I'll send you a list of all the stuff.

[02:49:24] I got no affiliation with Amazon links or anything like that. I just want you to survive so that you can come back and enjoy another twit that's in the can. This is amazing.