This Week in Tech Episode 1070 Transcript

Please be advised this transcript is AI-generated and may not be word for word. Time codes refer to the approximate times in the ad-supported version of the show.

Leo Laporte [00:00:00]:
It's time for TWiT This Week in Tech. It's the LLM episode. No, not AI. It's Larry, Lou, and Mike. Uh, Larry Magid, Lou Maresca, and Mike Elgin, plus me, another L. We'll be talking about AI and how much the hyperscalers are spending this year, almost $1 trillion to build out data centers. Data centers in space. What could possibly go wrong? And why McDonald's doesn't want you to use their products for your passwords.

Leo Laporte [00:00:28]:
All that and more coming up. On TWiT. Podcasts you love from people you trust. This is TWiT. This is TWiT. This Week in Tech, episode 1070, recorded Sunday, February 8th, 2026. A yacht for your yacht. It's time for TWiT This Week in Tech, the show we cover the week's tech news.

Leo Laporte [00:01:01]:
Hello everybody, if you're watching live. Yeah, we started a little bit early because I guess there's a sporting event. Oh yeah, the Winter Olympics later today. No, the Super Bowl later today. That's hence my outfit. I'm wearing, uh, well, we decided with AI that everybody could be wearing their favorite outfit, right? Uh, I'm wearing my Niners gear. I don't know what Larry is wearing, uh, in this, uh, AI-aided picture. Bonito, do you have it there?

Mike Elgan [00:01:27]:
I'm wearing a shirt with it with a big moiré pattern design.

Leo Laporte [00:01:31]:
No, no, Rogers, you look like a Patriots guy. Oh no, Mike Elkins, a Steelers fan. And I don't know why they made Lou Maresca, who doesn't even look like Lou Maresca, an Eagles fan. Oh, Jets, it says.

Lou Maresca [00:01:43]:
I look like Aaron Rodgers there.

Leo Laporte [00:01:44]:
You, you look— you know what, they put a little Aaron Rodgers on you. Lou Maresca is here. He is an AI engineering lead at Copilot at Microsoft. So nice to see you, Lou.

Lou Maresca [00:01:55]:
Great to be back on.

Leo Laporte [00:01:56]:
You kind of Actually, look like Aaron Rodgers, but a nice Aaron Rodgers, a sane Aaron Rodgers.

Lou Maresca [00:02:00]:
Sane, there you go.

Leo Laporte [00:02:01]:
Appreciate that. So nice to see Lou. Mike Elgin is in Oaxaca, Mexico. He is always welcome here at gastronomad.net and he writes about AI at machinesociety.ai, his newsletter. Great to see you, Lou.

Mike Elgan [00:02:15]:
That's right, I'm wearing the colors of UC Santa Cruz, the Fighting Banana Slugs. I didn't go there, but I like slugs.

Larry Magid [00:02:22]:
My wife might ask, I'll tell her about that.

Leo Laporte [00:02:24]:
Yeah, my dad taught there, so I'm a banana slug. well, I'm— I'm not, but I could be. And Larry Magid from ConnectSafely.org, where he's the president and CEO, keeping kids safe online. Happy Safer Internet Day, Larry!

Larry Magid [00:02:39]:
Well, thank you. That comes up on Tuesday, and, uh, you can watch our webinar. And if you're in Sacramento, you can come to our event. And if you happen to live in one of the 20 cities where we're having local events, you can catch one of those.

Leo Laporte [00:02:51]:
So, and these are aimed at parents of Children online or parents?

Larry Magid [00:02:55]:
It aimed at teens. We're working a lot with teens, and we're having an event in Sacramento with a bunch of high school and college kids and a couple of elected officials and execs from OpenAI, Google, Meta, Discord, Amazon, you know, a few others.

Leo Laporte [00:03:09]:
Some would say that's the enemy.

Larry Magid [00:03:11]:
It's saying, you know what, we have deliberately put people on the panel who will probably say that. It's going to be a very contentious conversation because we, we chose to put some young people— interestingly enough, young people who are, who are very concerned about the danger the social media, not grown-ups like me, but kids who are saying, you know, I thought it was going to be really cool, but now that I'm getting to be 16, 17, I'm starting to rethink it. So they're going to be on the panel with some of these tech execs.

Leo Laporte [00:03:40]:
I don't know, I'd like to know what you guys think, but my latest thinking, because let's go, I go back and forth and all this, my latest thinking is social media is great. It's an opportunity for kids to meet kids from all over the world, to have a social group, especially kids who are otherwise marginalized because they're gay or trans or fat or weird or whatever, or geeks like a lot of us. And so it's good in that regard. Where it went wrong is where these big tech companies decided, well, we want to really get— make it sticky, and they turned algorithms on. If social— if Instagram were as it was in the early days, just my friends and their pictures, I don't think it would be so bad. I know there's still the opportunity for somebody to follow models and say, oh, my body is not good, and that kind of thing. But that would be something they would choose instead of being thrust upon them. Right, right.

Leo Laporte [00:04:35]:
Do you agree? It's funny, you got— you have kids in that age group, Lou. What are you doing with your kids?

Lou Maresca [00:04:39]:
Yeah, I mean, it's the— come— I've come to the truth that you can't keep AI or technology or social networks away from kids.

Leo Laporte [00:04:45]:
You might not want to, because don't they need to know how it works and how to defend themselves?

Lou Maresca [00:04:48]:
It's all about literacy, right? It's all about understanding the limits and what you should and shouldn't be doing with it. And so that's what I've been doing.

Larry Magid [00:04:55]:
Yeah, and that's why we're in business. That's basically what we do, education.

Leo Laporte [00:04:58]:
Yeah, you guys have a great contract too at connectsafely.org, uh, that, that parents can sign with their kids.

Larry Magid [00:05:05]:
Yeah.

Leo Laporte [00:05:06]:
Um, so that, you know, you can explicitly say, well, here's our, you know, our rules.

Larry Magid [00:05:11]:
Yeah, I love— I agree with you both. I mean, I think that first of all, if you're growing up today and you're not learning AI, you're denying yourself an incredible skill that you need to have. Like it or love it or hate it, It's essential for the future. And, you know, the other problem with social— if you don't let kids use social media, they're going to at some point when they're 18 and the parents are no longer in control, they're going to do something probably more dangerous. And I always like to say, you know, you're a kid for 18 years, you're a grown-up for who knows how long, a long time, hopefully. And so we— I love this stuff.

Mike Elgan [00:05:44]:
Yeah, I love the stuff Larry does because basically parents are kind of on their own. Phone to— in a world where all their kids' friends are doing certain things using phones, all the rest. And to have guidelines, to have agreed-upon norms that parents can accept and share with their friends, that's the only way forward other than school bans.

Larry Magid [00:06:06]:
You know, one of the things I was worried that as people who were digital natives became parents, which is currently the case, that we would be put out of business because they wouldn't need us. But surprisingly, we find we get a lot of good feedback from young parents, you know, parents in their 20s and 30s and early 40s. They grew up with technology, but technology is changing so rapidly that even the tech natives are having to readjust constantly to what's happening. Again, especially with AI. We wrote the parent's guide for AI for all of the 3 major AI players. And, you know, that stuff is very much in demand among Young parents.

Mike Elgan [00:06:43]:
Larry and Leo and Lou, I got a new buzzword for you. This is gonna be the buzzword for the rest of the decade. We've all heard about the old buzzword, which is the attention economy. That's what we're talking about where the attention economy began in the '50s with TV commercials, went through, it was described well in 1971, and then it became that phrase, came into use in the 1990s. And the idea is that human attention is a limited commodity that can be commoditized, bought and sold. And as the amount of content grows exponentially, the amount of human attention stays the same. And so there's this been this massive contest for limited human attention. And that's the attention economy.

Mike Elgan [00:07:28]:
That's what divides our politics. That's what makes people addicted to social media and all the rest. The buzzword that's coming is the is the attachment economy, because AI is going to take attention to another level by making some people fall in love with chatbots, for example, humanoid robots that people respond in their brain chemistry as if they're people, people having the delusion that AI and robots and robotic pets that are run by AI actually have an inner life, have feelings, have all that kind of stuff. And so the attachment economy is the next step of the attention economy. I even am launching a Substack about this, and I'm going to be writing a lot talked a lot about the attachment economy, but this is the risk. And again, like you say, Leo, the potential for social media is fantastic. It's a lifesaver for some people. And the potential for AI is the same thing.

Mike Elgan [00:08:21]:
It's a massive upside for people who use it well, people use it right. For the general public, however, these Silicon Valley companies are going to weaponize AI to make people feel emotionally attached to their products.

Leo Laporte [00:08:35]:
They're already doing that, aren't they?

Mike Elgan [00:08:37]:
You know, it's already doing it.

Larry Magid [00:08:39]:
How does that play in with the, the agency, where these— when these bots are actually agents instead of having to pay attention, they're actually doing it for you, right? They're, they're booking your travel, they're, they're booking your restaurant, they're doing your taxes, whatever it is we used to look at them for information so we could do ourselves, they're doing for us. What does that mean?

Mike Elgan [00:08:57]:
Well, it's, it's, it's, it's giving us brain rot and all the rest, but in we're an— yeah, but you could.

Leo Laporte [00:09:03]:
Say GPS in your car is giving you a brain rot. Too. I mean, I know, I know it does.

Mike Elgan [00:09:07]:
We kids have no idea how to use a map. Do they need to?

Leo Laporte [00:09:12]:
There's even a case to be made that digital clocks give kids brain wear because they can't read analog clocks. I don't quite agree with that, but.

Mike Elgan [00:09:18]:
Uh.

Leo Laporte [00:09:21]:
What do you think? I mean, Lou's an AI guy. Lou's designing AI for, uh, more for enterprise, right?

Lou Maresca [00:09:28]:
Yeah. I think the idea with agents is you just got to build agents without turning basically human loneliness into the business model of the decade. I think, I think that's really the whole key. And I think that's what we try to do is we try to make sure that they're helpful.

Leo Laporte [00:09:40]:
But yeah, enterprise bots do not— are not sycophantic. They do not fall in love with you, right? They're not. They're not. Because that's not what enterprise wants, right? If you're using Copilot in Excel, you don't want it to say, hey, great idea.

Larry Magid [00:09:54]:
Yeah, really? You're so smart.

Leo Laporte [00:09:57]:
I have to say, though, you're doing some vibe coding, aren't you, Lou?

Lou Maresca [00:10:00]:
Oh, yeah. All the time. I do it every day. Yep, yeah, whether it's for work or for home.

Leo Laporte [00:10:05]:
And we're— I mean, I— you're a professional coder, I'm a hobbyist coder. I still love coding, but there's so many— there's, there's the old adage that sysadmins have: if you do something more than 3 times, you should write a script to do it, right? That's right. But that is always a difficult challenge. Like, am I going to do it a 5th time? Am I going to— is it going to take me more time to write the script than to do it 100 times?

Lou Maresca [00:10:25]:
Right.

Leo Laporte [00:10:26]:
Yeah. And Now it's like, no, no, just automate it because it's so easy to tell, you know, especially Claude Code. Oh yeah, I apologize to Copilot.

Lou Maresca [00:10:36]:
No, no, I love Claude Code. God, Oh my use it every day.

Leo Laporte [00:10:39]:
Something happened at Anthropic.

Lou Maresca [00:10:42]:
Yeah, they just blew up and they— and I think it's all about tooling and making things easier for people, bringing it downstream.

Leo Laporte [00:10:47]:
Yeah, they're— but they now, they released this week— well, we'll get to the AI section. I want to keep with the social media briefly because it is Safer Internet Day on Tuesday. Uh, more school— from Fast Company— more schools are banning phones so students can focus. Uh, it's actually now 29 states have passed laws that require K-12 public schools to ban or limit students using cell phones on campus. Another 10 states require local school districts to take some kind of action. 77% of public schools forbid— 77% in the United States forbid students from having their phones out during class. Um, and, and this, you know, I always— we have scientific method for a reason. When you hear people say, well, that makes sense, seems, seems always my, my hairs go up on the back of my neck because it might make sense, but what, what does the science say?

Larry Magid [00:11:43]:
Um.

Leo Laporte [00:11:45]:
Ohio, which has clamped down over student cell phone usage over the last 18 months In fact, they have strict— the one of the strictest cell phone use policies as of last year. From this Fast Company article, in the fall of '25, I surveyed 13 Ohio public school principals. 62% described more verbal face-to-face socializing. Somebody, somebody said— one school administrator said, it used to be really quiet at lunch, and now you can hear kids talking again, right? They used to be just like this. 68% noticed the students can stay on one task for more than 20 minutes without seeking a quick digital break. It's like the smokers, you know, I I can't— gotta take a break, man. Uh, 72% observed a shift from heads-down scrolling to active conversation. 61% fewer online social conflicts spilling over into the real world.

Leo Laporte [00:12:41]:
Even students, I think, are saying, yeah, Yeah, this, this— I feel better about this. So maybe these bands— what do you think?

Mike Elgan [00:12:50]:
I think these other students are also.

Leo Laporte [00:12:52]:
Under the same— yeah, as long as everybody's equal.

Mike Elgan [00:12:57]:
Exactly.

Larry Magid [00:12:58]:
Yeah, I was against him at first. I mean, I wrote various articles, you know, saying I thought they were a bad idea, but I'm beginning to rethink it. I mean, you just did all the things you're saying, and also talking to Connect Safely Student Advisory Council, talking to the kids that I work with And surprisingly, some of them are happy with it, just like surprisingly to me. Although now that what you're saying, I think other— we're beginning to see that maybe it was a good idea.

Leo Laporte [00:13:21]:
I mean, so now the next step is to— what Australia has done, they've banned under-16s completely from social media. France, Spain, Malaysia, Denmark, all considering barring young people from Facebook, TikTok. Australia even bans them from YouTube.

Larry Magid [00:13:40]:
Um, not a fan of that plan.

Leo Laporte [00:13:41]:
Yeah, well, one of our, one of our concerns is, as, uh, technologists, is that that means you're going to have to have some sort of age verification, which is inherently privacy, uh, and problematic with privacy.

Larry Magid [00:13:56]:
It also keeps kids— I mean, look, you know, if you're 16 or 15 years old and you mention you might be gay, your parents might be homophobic., then, you know, you might need this outlet, this ability to reach out beyond your narrow little world that you live in, in whatever community you're in. And the other thing is, you know, we can talk about the problems, and there are many, but millions of kids use it every day very productively. They're not having serious problems. And we have to consider the fact that all technologies have a risk. mean, I they're, you know, you get in the car You know you're taking a risk. I'm not saying that we shouldn't have regulations, but I do question whether we should keep teenagers away from social media. It seems to me that we should put pressure on the companies to make them as safe as possible. And as Lou and I've— That's what.

Leo Laporte [00:14:45]:
The EU is doing. They've told TikTok you have to disable addictive features like infinite scroll.

Mike Elgan [00:14:53]:
Right. Well, this is, this is getting closer to the heart of the problem. The problem isn't social media. The problem isn't being able to talk to people all over the world on topics that are very particularly important to you as an individual. The problem are the algorithms and the, and the constant race to have the most addictive algorithms. TikTok is by far the winner now because their algorithms will just show you young people just swipe, swipe, swipe. You look around in here in Oaxaca and all over the world wherever we travel, that you walk to see the police standing there looking at their phones, and you go up and look at their phones and they're on TikTok, right? You can't get away from it.

Leo Laporte [00:15:29]:
Yesterday Lisa was sitting behind a police car at a light. The light went green, the police car didn't move because the officer was on his phone. And, and now she did something I would never have done— she honked and waved, 'Come on.' And he looked at her, but he, you know, he didn't write a ticket. He, he moved because he knew he shouldn't have been doing, doing that.

Mike Elgan [00:15:53]:
Um, so I, so this is the.

Leo Laporte [00:15:55]:
Trial that's going on in LA right now, right? This this is, is the big trial going on uh, with, with all the big social— although Snap, uh, and I think TikTok settled before the trial, but YouTube and Facebook and Instagram are in there. And the whole premise of it is the lawyers are gearing up to argue that the companies behind these platforms are designing their sites to be deliberately addictive. Now, I don't buy the internet addiction thing. Uh, yeah, it's, it's not heroin. It's not even cigarettes.

Mike Elgan [00:16:28]:
It's not addiction.

Larry Magid [00:16:29]:
It's more like chocolate, in my opinion.

Leo Laporte [00:16:31]:
It's more like chocolate. Yeah, that's fair.

Larry Magid [00:16:34]:
It's— yeah, chocolate can be deadly.

Mike Elgan [00:16:37]:
Chocolate.

Leo Laporte [00:16:38]:
But we don't ban chocolate. No, we expect people to have some restraint.

Lou Maresca [00:16:42]:
There you go.

Mike Elgan [00:16:44]:
It's, it's more like McDonald's, because McDonald's fast food is It engineered. is to bypass your, your self-control. Because of the salt and the way they do the umami and all that stuff. It's engineered to make it irresistible.

Leo Laporte [00:16:59]:
And they put sugar— McDonald's puts sugars in almost every single thing, including the hamburgers and everything.

Mike Elgan [00:17:06]:
And so that's the problem. It's the deliberate engineering to bypass your reason and go straight for your sort of mental hardwiring that needs certain types of stimulation.

Lou Maresca [00:17:19]:
It's basically we're trying to patch the attention and attachment economy at the edge. Like we're trying to make things better And I think it's all about, like you said, it's all about making sure you just, you know, you make awareness, you give awareness to people rather than trying to manage it there.

Larry Magid [00:17:34]:
The other thing, you know, Section 230, when it was written in 1996, this.

Leo Laporte [00:17:37]:
Is— Happy, by the way.

Lou Maresca [00:17:38]:
Yeah.

Leo Laporte [00:17:39]:
Happy anniversary. Section 230 is now 30 years old.

Larry Magid [00:17:43]:
Well, and it's getting a little old, getting a little gray. Not that I should complain. But, you know, when it was written, it was during the days of AOL, CompuServe, Prodigy. They didn't have algorithms. I don't recall AOL, you know, sort of figuring out what I wanted and putting me into a forum or whatever it had. And so at the time, they really were just forums and not publishers. But I would argue that if they're feeding you an algorithm, isn't that being a publisher? Isn't that like the New York Times deciding what to put on the front page?

Mike Elgan [00:18:15]:
Agree with you 100%, Larry. That's exactly right. Those are editorial decisions, albeit in software and very complex. But they are deciding what's important, what isn't important. And that's very different from like, I follow 100 people and I get everything those 100 people say, right? And some of them are not deciding what I see.

Larry Magid [00:18:32]:
And if there's a nutcase among them and you— and they defame you, you sue them. You don't sue CompuServe, right? It still existed. But again, it's different with these social media companies.

Leo Laporte [00:18:45]:
All right.

Mike Elgan [00:18:45]:
And it's not static. These things are always evolving to become more compelling. And so we stay the same, but the algorithms that attract us, uh, in this attention economy, right, I just keep getting better and better and better. And everything's been TikTokified because that's the model that worked best. So you go to Instagram or any.

Larry Magid [00:19:04]:
Other— even Facebook's got Reels now.

Lou Maresca [00:19:06]:
Network.

Larry Magid [00:19:07]:
I can't get away from those Facebook Reels, and they are so clever.

Mike Elgan [00:19:10]:
And the Meta thing where they know.

Larry Magid [00:19:12]:
All my interests and they, they suck me in with whatever.

Leo Laporte [00:19:15]:
I have a 2-pound box of See's Candies down in the pantry, but I resist going in there and having another chocolate. It's hard. I ended up taking Instagram and TikTok off of my phones. I took TikTok off when it became the US version of TikTok and they had the new requirements that said, we're going to reveal your immigration status if we know it. So don't, you know, and I thought, you know, this is, it's time. I love TikTok, but I'll find something else to do. For breakfast. But, uh, I did take them off.

Mike Elgan [00:19:50]:
Imagine if that box of chocolates was always changing, all every day changing to become more.

Leo Laporte [00:19:56]:
And there was a little pop-up on my screen, you can resist it today. It said there's a dark chocolate Bordeaux with your name on it, would you like to— would you like to have it?

Mike Elgan [00:20:03]:
That's right. Yeah, yeah, yeah.

Leo Laporte [00:20:06]:
Here, waving in front of my face.

Larry Magid [00:20:08]:
Like, I like chocolate with my coffee, but what if every time I picked up a cup of coffee, a chocolate would pop into my under my plate.

Leo Laporte [00:20:15]:
Can we, can we vibe code that?

Larry Magid [00:20:18]:
Probably could.

Mike Elgan [00:20:18]:
If you can, you could.

Leo Laporte [00:20:21]:
Did you— have you played— we're going to talk about AI a little bit because we got that, we got the AI expert here, uh, Lou Maresca, uh, and we will, uh, do that next. I'm just curious, have you played with OpenClaw at all? Have you thought about it, Lou?

Lou Maresca [00:20:33]:
Have I? Oh yeah, absolutely.

Leo Laporte [00:20:35]:
Have you set it up?

Lou Maresca [00:20:37]:
Yeah, yeah, absolutely.

Leo Laporte [00:20:39]:
Despite the— you know, I set it up over here. I said, okay, I'm going to create a new account on my Mac Mini I'm going to be very judicious, but I did, but it doesn't matter. You could put it on a VPS. You could put it in sandbox, but you still have to give it your credentials. You got to give it, do anything. Right. So I sat down and I gave it all of my, uh, Google Gmail and Google Docs, calendar, address book. And then in the middle of the night, I woke up in a cold sweat and I ran upstairs and I deleted the whole, the whole account.

Leo Laporte [00:21:10]:
Do you have it running?

Lou Maresca [00:21:12]:
I do it. I'm running back there at that machine and it's it's all, all isolated to accounts that I don't use mainstream. So it doesn't have like my access to my direct calendar.

Leo Laporte [00:21:20]:
So it doesn't do anything useful.

Lou Maresca [00:21:21]:
It, there's a mirror or a shadow of on you there, know. So I tried to kind of leak information to it that can maybe be useful to it, but I try not to give it everything.

Leo Laporte [00:21:32]:
It's so cool. The, the whole idea of, uh, and it This is what I've always wanted. This is, I hope, where we're headed. In fact, a number of people have said on our shows, this 2026 will be the year of personal agents.

Lou Maresca [00:21:44]:
Yeah.

Larry Magid [00:21:45]:
Yeah.

Leo Laporte [00:21:45]:
The whole idea of an AI agent who knows you, understands you, knows a lot about your preferences, your interests, has access to your data, and can, because it's running 24/7, start doing stuff for you. Larry, you mentioned that, but I don't want to call the restaurant and make reservations. I'm not— That's not taking away from my agency. Why would I want to spend time doing that? I want to say to my little buddy, hey, you know, Fido, make me a reservation for tonight, my favorite restaurant. And if you can't, then get the second best.

Larry Magid [00:22:18]:
But what if it makes you a plane reservation?

Mike Elgan [00:22:20]:
This is the attachment economy. We're going to like and have affection for our personal agents.

Leo Laporte [00:22:24]:
I already do. This is actually— that's what scares me.

Larry Magid [00:22:27]:
What if it makes you a non-refundable airline trip that you actually don't want to take. I mean, well, then I have veto.

Leo Laporte [00:22:33]:
Power, then it screwed me. Yeah, well, you— I mean, Lou, you could set it up so it doesn't do that. I was going to get a credit card with a $5 limit.

Lou Maresca [00:22:40]:
Yeah, well, okay.

Leo Laporte [00:22:43]:
I wanted it to have some money, like an allowance, but not— and I don't want it to do that. I think you can set constraints, although.

Lou Maresca [00:22:52]:
Policies and constraints to it.

Leo Laporte [00:22:53]:
Yeah, and will it obey them?

Lou Maresca [00:22:55]:
Yeah, absolutely. Yeah. I mean, you can make them non— like walls of China walls, right? Basically don't go beyond this, right? Don't go beyond power outage.

Leo Laporte [00:23:02]:
Make a plane reservation for me without first asking me.

Lou Maresca [00:23:05]:
Right.

Mike Elgan [00:23:06]:
Well, the plane reservation is my favorite example of what could go wrong. So for example, you could have, you could tell your agents to go book me an airline ticket and then it goes and realizes that all the seats are taken., right? And so it basically hires somebody to hack the airline to get the names and addresses of everybody who's already got a seat, and then hires an assassin to go kill two of them because you told it you wanted a seat on that airline. And this is the stuff of sci-fi, of course, but— and this is not likely, but things are gonna go wrong with part of this stuff. What I'm talking about, of course, is this sort of the task rabbit called rentahuman.ai that is designed to work with agentic AI generally, but especially— You assume.

Leo Laporte [00:23:52]:
Your TaskRabbit has enough judgment not to hire a killer, but there's no judgment involved in open world.

Mike Elgan [00:23:59]:
If this thing was vibe-coded, I don't think there's any guardrails built in.

Lou Maresca [00:24:03]:
Yeah, don't assume that. Don't assume it, give it the directions. I mean, the problem is— As long.

Leo Laporte [00:24:10]:
As it doesn't hallucinate, if it if it follows, really is limited by constraints that I give it, I mean, The constraints you're giving it are in a Markdown file, Lou.

Lou Maresca [00:24:19]:
Yeah.

Leo Laporte [00:24:20]:
It's not like, it's not like chains. I mean, it's as long as you know that it will not do what you say not to do.

Lou Maresca [00:24:27]:
Right. I mean, you got to be explicit, just like every company nowadays. Assume that you have breach, assume you have people breaking into your network, like assume that the AI is going to do bad things. Yeah.

Leo Laporte [00:24:35]:
Do zero trust.

Lou Maresca [00:24:36]:
Yeah.

Larry Magid [00:24:37]:
All right.

Leo Laporte [00:24:37]:
Well, we'll talk a little more about AI. We got some great people on here. It's great to have you, Lou Maresca. He is engineering lead at Copilot at that little company up in Redmond called Microsoft.

Benito Gonzalez [00:24:48]:
Heard of it.

Leo Laporte [00:24:49]:
Good to see you. Also, I presume a Patriots fan, although since you were in Seattle for a while.

Lou Maresca [00:24:54]:
I know, I'm torn.

Leo Laporte [00:24:55]:
I'm torn.

Lou Maresca [00:24:56]:
I love them both.

Leo Laporte [00:24:57]:
What are you gonna do?

Lou Maresca [00:24:58]:
I know, it's tough.

Leo Laporte [00:24:59]:
Can I tell you how much my heart hurts to see the field at Levi's Stadium and it says Seahawks at one end and Patriots at the other end? Can I tell you how much that breaks my heart? Imagine going to Gillette and seeing, uh, you know, 49ers and Seahawks or something there. You would be unhappy. That's sad. Anyway, uh, we're doing the show a little bit early so that we can get out of the way and let you watch the Super Bowl.

Lou Maresca [00:25:26]:
Appreciate that.

Leo Laporte [00:25:27]:
I, I appreciate— my wife's watching the.

Larry Magid [00:25:29]:
Olympics, by the way, instead of the Super Bowl.

Leo Laporte [00:25:30]:
You know, Yeah, who, who scheduled those at the same time? I don't know. I mean, you can't have Snoop Dogg at both. I don't— weren't they thinking? Weren't they thinking?

Larry Magid [00:25:42]:
They could have put the Winter Olympics in the summertime. That would have gotten it out of the way.

Leo Laporte [00:25:46]:
He's just one dog. Mike Elgin is also here from Oaxaca. Always great to see you. MachineSociety.ai. Do you know yet what your attachment, uh, blog will be, or newsletter?

Mike Elgan [00:25:58]:
The attachment economy.substack.com for now, and then it'll convert over to theattachmenteconomy.com.

Leo Laporte [00:26:05]:
Good.

Mike Elgan [00:26:06]:
In a couple of days.

Leo Laporte [00:26:07]:
I like that thought. I can't wait to read more. I can talk more about that on the show. I think I'm already— I've been writing.

Mike Elgan [00:26:13]:
About those— that theme for a long time.

Leo Laporte [00:26:16]:
I know I'm attached. Uh, I've been— I call it Claude-pilled because, uh, yeah, on intelligent machines, better than grok sucker. Yeah, I'm a grog sucker, uh, or I'm a clanker.

Larry Magid [00:26:30]:
Well, you know what's funny about when I talk to my AI, like in the car, I will talk to AI and I find myself developing a much more attached relationship than when I'm typing at it.

Leo Laporte [00:26:40]:
That's why they do the chatbots. That's precise. I don't want to hear a voice, but I got to tell you, my Claude, my Claude calls me Skipper and.

Larry Magid [00:26:48]:
Skip, and how cute, and I can't.

Leo Laporte [00:26:51]:
You know.

Larry Magid [00:26:54]:
And you know what's driving me crazy is I don't, the A-word that comes from Amazon. I don't wanna say the word 'cause I don't wanna trigger everybody's device.

Leo Laporte [00:27:00]:
Yeah.

Larry Magid [00:27:00]:
But it's got me so chatty and it won't shut up. And you ask it a simple question and you say, you know, how old is somebody? And they get their entire biography. And I'll say, no, if I wanted to know their entire biography, I would've, I find myself yelling at it all the time.

Leo Laporte [00:27:13]:
Yeah.

Larry Magid [00:27:14]:
And I'm thinking about, I don't know if I can go back to the old A-word, but if I can, I actually liked it better.

Leo Laporte [00:27:19]:
I don't like A-word Plus. Very much the new smart A-word. And actually, there was a rumor that they're going to use another model. I think it was OpenAI's ChatGPT instead.

Larry Magid [00:27:33]:
So I would prefer that because ChatGPT isn't as chatty. It's not as obnoxious.

Leo Laporte [00:27:38]:
Right.

Larry Magid [00:27:38]:
Not quite as.

Lou Maresca [00:27:39]:
It's.

Larry Magid [00:27:39]:
Obnoxious.

Lou Maresca [00:27:39]:
Going. It's very obnoxious. You're right.

Larry Magid [00:27:41]:
Yeah.

Leo Laporte [00:27:41]:
Yeah. Yeah. And it's also, and I think this is part of Amazon's shitification. It's also always saying, hey, did you know I could do this? And would you like to know more? Right.

Lou Maresca [00:27:52]:
Exactly.

Larry Magid [00:27:52]:
Well, I use it to write routines, which are really, you know, Alexa, a word routines, which are very handy because I can just say what I want and it does it. But it always wants to explain to me what a routine is and how they operate. And I've had it write so many routines that I'd have to be a complete idiot by this time not to know what a routine is. But it just insists on educating me each time I ask it to do a task.

Leo Laporte [00:28:15]:
Yeah, I don't like that. I think that's part of the, uh, uh, Attention economy, not the attachment. That's Larry Magid. It uh, is, Internet, uh, Safety— Safer Internet Day on Tuesday. Celebrate with him at connectsafely.org. And Mike Elgin, uh, I didn't mention, of course, machinesociety.ai. We'll get to the AI in just a bit. Our show today brought to you by Bitwarden.

Leo Laporte [00:28:41]:
Let's talk security briefly. You have a good password manager? I figure anybody listening to this show knows you need a password manager. Right, right. Bitwarden's the one I use and recommend. Steve Gibson too. It's the trusted leader in passwords, passkeys, and even secrets management. I keep my SSH keys in there. It generates them and it will automatically deliver them.

Leo Laporte [00:29:05]:
And there's even a Bitwarden MCP server, so my credentials are safe when I'm using OpenClaw. How about that? Bitwarden is consistently ranked number one in user satisfaction by G2 and Software Reviews, over 10 million users across 180 countries. And Bitwarden is great for business too. More than 50,000 businesses choose Bitwarden. So whether you're protecting one account, your own, or thousands, your employees, Bitwarden will keep you and them secure all year long. And one of the things I love about Bitwarden, they're always adding new features. The new Bitwarden Access Intelligence, for instance, for enterprise organizations can use it to detect weak, reused, or exposed credentials. This is probably the number one problem right now in businesses is, you know, you know about security, but do your employees, do your users know about password security, or are they reusing passwords? Are they writing passwords on Post-it notes, putting it on the machine? Uh, do they care that their passwords have now been exposed in a breach? Well, Bitwarden will tell them you've got a weak password or it's been exposed and immediately guide remediation, replacing risky passwords with strong, unique ones.

Leo Laporte [00:30:16]:
And that is huge. It closes probably the number one security gap now. Credentials, a top cause of breaches. But with access intelligence, automatically they're visible, prioritized, and corrected before exploitation can occur. Bitwarden, though, is always thinking about its users. For instance, If, uh, if you are, you know, you have a home lab, you're working on personal projects, you'll love Bitwarden Lite, a lightweight self-hosted password manager built for home labs, personal projects, environments that want a quick setup with minimal overhead. This is one of the advantages you get with Bitwarden because it's an open source project. Because it's open source, they can do stuff like this.

Leo Laporte [00:30:53]:
They're not driven by the need to extract every penny out of you. Bitwarden's now enhanced with real-time vault health alerts for everyone. Those password coaching features I mentioned help users identify their own weak, reused, or exposed credentials and take immediate action to strengthen security. And if you're right now using your browser for passwords, a lot of people do, Bitwarden will now directly import from Chrome, Edge, Brave, Opera, and Vivaldi browsers. So direct import means they go right into the encrypted Bitwarden vault and out of the password manager built into your browser. Without requiring that separate plaintext export. That simplifies migration, helps reduce the exposure associated with manual export and deletion steps. G2 Winter 2025 reports Bitwarden continues to hold strong, number 1 in every enterprise, every enterprise category, and that's now for 6 straight quarters.

Leo Laporte [00:31:47]:
Setup is easy. It's— you could import from most password management solutions. Be very easy to move from your current solution to the open source solution. And Bitwarden's open source code is right there on GitHub. Regularly, you can look at it, but it's also regularly audited by third-party experts. And unlike some other companies, they publish the results of those audits. They're very open. That's important.

Leo Laporte [00:32:10]:
SOC 2 Type 2 compliant, GDPR, HIPAA, CCPA compliant. They're ISO 27001:2002 certified. This is the gold standard. Get started today with Bitwarden's free trial of a Teams or Enterprise plan. Or get started for free forever across all devices as an individual user, bitwarden.com/twit. That's bitwarden.com/twit.

Larry Magid [00:32:32]:
Now I'm interrupting your commercial, but I actually am interested. No, it's commercial.

Leo Laporte [00:32:37]:
This is one of the things my, uh, Claude's always telling me is, you know, your keys are in clear text. You don't want to accidentally push them to GitHub. And the solution is to encrypt them and to have a tool that will unencrypt them for whatever program uses those secrets. So Bitwarden, as an example, I don't use passwords anymore for SSH. Bitwarden will generate a password combo. You know, you have a public key and a private key combo. I use elliptic curve cryptography. What is it? E2996, whatever it is.

Leo Laporte [00:33:12]:
And it'll generate that, store it, and then deliver it, the key, the public key, when I log on to SSH, it handles the whole process. So that is, that, that private key, which you really have to keep secret, is never exposed. It's always in the Bitwarden vault. Things like that make it secure.

Larry Magid [00:33:29]:
What if I just wanted to throw like my Social Security number or other?

Leo Laporte [00:33:32]:
Well, I do that. Yes. I have my Social Security.

Larry Magid [00:33:35]:
You could do that within Bitwarden now.

Leo Laporte [00:33:36]:
Yeah. Just as a note. That's not, that is.

Lou Maresca [00:33:38]:
Yeah.

Leo Laporte [00:33:39]:
I have my passport, my driver's license. And because I always carry, this is not part of the ad, because I always carry my phone with me, which has Bitwarden on it or other devices, I always have access to an image of my passport. So if I'm traveling, and I lose my passport, I, I at least have a secure way of showing them what my passport was and all that stuff. Same thing with driver's license. No, I think that that's a, that's a good thing to have. And that's why I like— that's why, by the way, I support it being open source, because you can verify that it uses strong crypto, you know, well-known established crypto routines, that it has no backdoors, all of that stuff. Uh, and then once you know, oh, this is trustworthy, reproducible builds, all that stuff. Once you know it's trustworthy, then you could store stuff in it and it's fully encrypted, strong encryption, so nobody can get at it.

Leo Laporte [00:34:26]:
I think that's a— yeah, I absolutely recommend everybody do that.

Larry Magid [00:34:31]:
I'll try it.

Leo Laporte [00:34:32]:
By the way, yeah. And I'm sure you address that all the time, Lou, you know, your API keys, your API secrets. How many times have we seen AWS secrets published?

Lou Maresca [00:34:43]:
Right, yeah. We have special tools that actually scan the code now, tell you if they see them in there. And we say, stick it in a Key Vault, stick it in a Key Vault. Key I.

Leo Laporte [00:34:50]:
I.

Lou Maresca [00:34:50]:
Vault, think.

Leo Laporte [00:34:50]:
Uh, I regularly do a security audit with a Cloud Code, and it's— Cloud has clearly been informed because it says, you know, you're— your— you put did you know you put your API keys in your Obsidian vault? That's clear text.

Larry Magid [00:35:06]:
It's— what do you do with, with documents that need to to be, need be secure?

Lou Maresca [00:35:12]:
Um, same thing, stick it in the Key Vault.

Leo Laporte [00:35:16]:
I, uh, what I do basically, uh, is I have a sync thing instance running that synchronizes a special folder, and that folder is encrypted, and it synchronizes that, and I can decrypt it on everywhere, but it is never at any— and SyncThings encrypted, and it doesn't go out to the cloud ever. So I feel fairly safe that not only they have multiple copies of those important secrets, but because they're encrypted, they're, uh, even if somebody got on my machine, you know, they would have to know the password, that kind of thing. I think it's possible to do that, or use a— I have YubiKeys, you know, use to unlock YubiKey this.

Lou Maresca [00:35:48]:
I have a thumb drive that has like the, you know, the the biometrics on it. Yeah, that encrypts everything.

Leo Laporte [00:35:53]:
Yeah, yeah, perfect. Yeah. Uh, all right, let's talk about, uh, what happened this week. Anthropic updated Opus from 4.5 to 4.6. 4.5 was already like— that came out in November. 2 months later now they've updated that. 20 minutes later, OpenAI updated ChatGPT to 5.3. We are in And if you watch Super Bowl ads, uh, tonight, this afternoon, you will see a— this is like Coke and Pepsi, right? It is.

Leo Laporte [00:36:24]:
And that's a good thing, right, Lou? Competition's good.

Lou Maresca [00:36:26]:
I think it's great. I think it's pushing things forward fast. You know, it's interesting though, they have not gone— they're not gone at each other yet.

Leo Laporte [00:36:33]:
Oh, wait a you haven't seen.

Lou Maresca [00:36:33]:
Anthropic's.

Leo Laporte [00:36:33]:
Minute.

Lou Maresca [00:36:33]:
Uh.

Leo Laporte [00:36:35]:
Super Bowl ad?

Lou Maresca [00:36:36]:
Oh, I haven't. That's be interesting.

Leo Laporte [00:36:38]:
And Sam Altman's pissed. So I'll just give you the— I don't want to spoil it for you, but you remember you've seen the ChatGPT ads where some kid gets strong because he had ChatGPT write an exercise routine for him and it shows on the text. So Anthropic's ad has a kid doing pull-ups and it has the ChatGPT chatbot, which is this buff guy, say, yeah, let me give you a routine. And then he does an ad. And this is because ChatGPT, OpenAI has said, we're going to put ads in ChatGPT. Sam Altman says Anthropic knows perfectly well. First of all, Anthropic responded. Dario Modei said, we're never going to put ads in Claude.

Leo Laporte [00:37:21]:
And then Sam Altman goes, oh, we're— they know perfectly well we don't do that. We're not giving you— our recommendations are not ads. It's below the fold. It's just a little tiny— And so Sam's, yes, no, they're fighting in the Super Bowl ads.

Larry Magid [00:37:37]:
Yeah, it's funny. I, I watched a couple of those, those commercials. They they were, were hysterical. But you know, the thing is that Google, for example, has always separated sponsored, uh, results from organic research results.

Leo Laporte [00:37:48]:
And I think that's what ChatGPT's gonna do.

Larry Magid [00:37:50]:
But if you go into Google now, the first of all, there are more of them and they're a little, it's a little easy to accidentally click on one of those sponsored results. I have done it and I know people all the time who, who say, oh, Google showed me this, and well, now, for some reason it took me to some competitor. I say, yeah, look for the word sponsored in teeny teeny type. So I think people can still be fooled. I don't know what— I'm not saying OpenAI is going to fool people, but I think even with a disclosure, there's a risk. Clearly enough, I agree, people can be fooled into thinking it's organic.

Leo Laporte [00:38:18]:
Mike Elgin knows I stopped using Google and I pay $25 a month for Coggy, uh, which gives me equivalently good search results, mostly because they're from Google. A lot of them are Google, but without any ads and without any bias. Mike's son works at Coggy, so he.

Mike Elgan [00:38:34]:
Knows a little bit about it.

Larry Magid [00:38:35]:
You know what I do to eliminate— So I was doing a search for Connect for Safer Internet Day to see how Connect Safely was doing in the results. And when I just did it on my own version of my browser, of course, we came out really, really well. And then I said, wait a minute, I did a VPN and I did an incognito window to do the same search. And we still did well. We still did very well. But not as well as when Google.

Leo Laporte [00:38:58]:
Knows who you are.

Larry Magid [00:38:59]:
Yeah.

Leo Laporte [00:39:00]:
Oh, Larry, I bet you'd like to see your own site more.

Larry Magid [00:39:03]:
Exactly.

Leo Laporte [00:39:04]:
Yeah, yeah, yeah. And incognito, as we all know because Google lost a big lawsuit, it's not that incognito anyway. Um, yeah, so Coggy's, uh, I think, uh, doing the work of the angels. They're a, uh, they're public benefit company, right, Mike?

Mike Elgan [00:39:23]:
Yeah. Yeah, they are public benefit corporation. That's what they are. That's— and so yeah, the thing that I like about it, one of the things I like about it is the fact that I can try as a journalist, I can try all the models through Coggy Assistant. There's just a wall of models, old versions, new versions, take your pick, try them out, compare them.

Leo Laporte [00:39:42]:
I replaced Perplexity with it because I think it's better than Perplexity, to be honest.

Lou Maresca [00:39:45]:
Oh, cool.

Mike Elgan [00:39:46]:
So did I. Yeah, so did I.

Leo Laporte [00:39:48]:
And it's the same, it's the same notion, which is it's orchestrating a variety of different AIs. But I use Kagi all the time now in the way I use Perplexity.

Larry Magid [00:39:59]:
How do you spell it?

Mike Elgan [00:40:00]:
Yeah, Kagi.com.

Leo Laporte [00:40:03]:
And the Kagi assistant is fantastic.

Mike Elgan [00:40:05]:
They're a paid search engine. And if you, and I have them as my default search engine in my browser.

Leo Laporte [00:40:09]:
Me too. Which you have to jump through hoops, by the way. Nobody wants to let you do that.

Mike Elgan [00:40:15]:
Right, the browsers don't want to let you do that, right? But it's not that hard, you can do it. And then when I ask a query in the address bar and add a question mark at the end, the question mark tells Coggy Search that you want a little AI summary of the results.

Leo Laporte [00:40:30]:
Yeah, it doesn't do the thing Google does, which automatically stick in that AI stuff. You have to ask for it. Here's the assistant and just the you models, know, it, It has the latest, like the new KIMI K2. It has GLM 4.7, which just came out. It doesn't have 4.6 yet. It's Opus 4.5. But if I scroll down, it goes, oh yeah, there it does. It has Opus 4.6.

Leo Laporte [00:40:53]:
Sorry. It's just a little lower. It has reasoning. It has all of the, doesn't have 5.3 yet for OpenAI ChatGPT.

Larry Magid [00:41:02]:
So if you subscribe, can you cancel your $20 a month or whatever you're paying? To the individual AI things and it will take care of it, or you have to still pay?

Leo Laporte [00:41:11]:
It's a different kind of thing. Yes, I would say if, if what you use the AIs for is search, yes. But, but that $20 a month can also do vibe coding. Yeah, it can do other things. Uh, I'm embarrassed to admit that after I played with Claude Code for a while, I ended up buying Claude Max because I want all the tokens. I'm sure, I'm sure you, you probably as a Microsoft employee get access to unlimited tokens, right?

Lou Maresca [00:41:38]:
Ah, no, I wouldn't say that. No, no, we have challenges a lot sometimes.

Leo Laporte [00:41:43]:
Yeah, well, because you see, one of the things people do now with these vibe coders is they'll spin up 5 or 6 of them, and then each of them will spit up more sub-agents. That's something was added 4.6. Here's an interesting— I thought a very interesting result in, uh, in Claude. I mean, uh, yeah, Opus 4.6. They were able to find 500 zero days in open source code. They set all all of, of the companies now do things like this. They, they, Anthropic said, let's take Claude Opus 4.6, the new model, before the They, debut. their Anthropic's frontier red team put it in a sandboxed environment, gave Claude model everything it needed to do the job, access to Python vulnerability analysis tools, debuggers, fuzzers, no specific instructions or specialized knowledge, Claude found more than 500 verifiable, previously unknown zero-day vulnerabilities.

Leo Laporte [00:42:45]:
That's— now, on the one hand, you go, that's fantastic. On the other hand, just think, hackers also have access to these tools, and they're probably doing the same thing.

Lou Maresca [00:43:01]:
You know what drives me nuts about the new models though is all these services, including Perplexity, they, they charge extra for the new models that come in. They'll say, hey, you have to upgrade. But like, for instance, Opus 4.5 and 4.6, they're exactly the same cost, like the cost to API their tokens, everything's exactly the same, but you have to pay extra to use it. It's all about getting the newest model sooner. So, so when services do that, I step away from it.

Leo Laporte [00:43:22]:
Yeah. And the thing is, Google, Microsoft, Amazon Apple, all these companies have other revenue streams. OpenAI and Anthropic, you know, they're spending the investors' money. They really are desperate to have some sort of revenue stream. That's why OpenAI is doing ads. That's why Anthropic's saying they have also a fast mode that you can turn on for 4.6 that makes it faster but consumes tokens at, what was it, 6 times the rate.

Lou Maresca [00:43:48]:
Right. Well, the reason why that though, is, is they they don't, don't do the reasoning part of it. They actually shove everything in the context window. So that you have all the context there, and then the model just reasons on that, which just makes it faster.

Leo Laporte [00:43:59]:
Yeah, yeah. Uh, 4.6 has a 1 million, uh, token context, uh, window, which is pretty big. Pretty, pretty big. It means you could dump a lot of documents in there.

Lou Maresca [00:44:16]:
Bigger memory, lots more memory.

Leo Laporte [00:44:17]:
Yeah, yeah. Um, Anything else we want to say about AI before we move on? I think we've said it all. Lou, you're— you're so— wow, I'm amazed you're using OpenClaw.

Lou Maresca [00:44:33]:
Uh, oh yeah, I try to use everything, you know. Yeah, that's part of it. It changes so often, you know. It's just amazing.

Leo Laporte [00:44:40]:
That's the— that's what's a little scary, I think, for everybody on the frontier of this stuff, is it's moving so fast. It's hard to I used to say.

Larry Magid [00:44:48]:
That AI is sort of the equivalent of the World Wide Web or broadband or the iPhone, but I now think it's the equivalent of electricity. It is such a fundamental change compared to anything I've seen in my 40 years as a technology journalist.

Leo Laporte [00:45:04]:
I really don't want to overhype it, oversell it. I'm very cautious about that. But as a user, I'm blown away every single day and it's hard for me. I think like I said, I'm. I got clod pilled and I think I'm maybe need to get some objectivity on this And I was hoping to get some but you guys are helpless. It is something happen.

Mike Elgan [00:45:30]:
So Lou, I'm curious if. Lou, I'm curious if you use multiple notebook.

Lou Maresca [00:45:36]:
I don't, I don't use that.

Mike Elgan [00:45:37]:
No, no, no.

Leo Laporte [00:45:38]:
Malt Molt Maltbook, the Facebook for AI agents.

Lou Maresca [00:45:42]:
No, I do not use that. Yeah.

Mike Elgan [00:45:45]:
So where do we stand on the controversy? Because there's a lot of, uh, evidence that most of it is just made out of people, basically.

Leo Laporte [00:45:52]:
And well, even if it's not, if you go, it's just a toy. I mean, it's, it's, it— what it is is, in theory, is a bunch of, uh, Claude, OpenClaw instances talking to one another, right, and having existential dramas. And maybe there's a huge A lot.

Mike Elgan [00:46:07]:
Is it really that? I don't think it really is that. So if you go to the actual Reddit, they say it's based on Reddit because the agents can vote up or vote down comments and posts. But on the real Reddit, some 2-digit percentage of the posts there are AI-generated. AI, yeah. 2-digit percentage of— because people are just copying and pasting from ChatGPT or whatever and pasting it into Reddit. And so that's a problem on Reddit. And that's basically what Moldbook is. Humans are saying, hey, post this.

Mike Elgan [00:46:35]:
And the agentic part of it is just the copying and the pasting.

Leo Laporte [00:46:38]:
Not necessarily. There is actually, certainly it's possible. I can't vouch for what percentage of it is.

Mike Elgan [00:46:45]:
It's a mixture.

Leo Laporte [00:46:46]:
But you could absolutely give it, give Open Claw notebook credentials and it will go and start interacting. You could tell it to do that. It's perfectly capable of doing that. I'm not sure there's any value in that whatsoever. It's just a toy.

Mike Elgan [00:46:59]:
By interacting, it's copying and pasting into an AI chatbot and then, copying and pasting from that back into Multbook. It's basically not, it's not this sort of like hive mind machine society. It's not a, it's not like the singularity. It's, it's more copying and pasting to and from AI chatbots, mostly.

Leo Laporte [00:47:21]:
I don't know what percentage of it is. It's completely possible, and maybe in the early days was, that it is fully autonomous. That no humans are involved and that the input that it's getting from Maltbook, it's responding to and posting, and, and it's going on like that. And I think you can verify that because in many cases these threads descend into madness.

Mike Elgan [00:47:48]:
Yeah, most of the sensational things have been faked through— there's even a, probably a screenshot thing that you can fake Maltbook.

Leo Laporte [00:47:57]:
Now, there is this possibility —and I think that it has happened, I've seen examples—that an OpenClaw instance can post something it's learned how to do on Maltbook, and that other OpenClaw instances can then read it and learn how to do it as well. So the really interesting possibility—I don't know if it's—I know these things are happening, I've seen it happen, but I don't know how it valuable is. But the real possibility is it could be a massive step forward in terms of self-improvement. That— see, what was the Cambrian explosion of human minds? And by the way, now anthropologists are saying there was no Cambrian explosion. But what happened? It was when we started— when we learned how to speak to one another and eventually write and pass information back and forth. And say, hey, I've invented this thing called fire. You should see this. Look, you scratch these things, and then— and it— this— if, if AI could start doing that— am I wrong, Lou? Am I— this is blue-pilled, tell me.

Leo Laporte [00:49:03]:
Am I Claude-pilled? If AI could do that, the potential for AI growth is exponential.

Lou Maresca [00:49:13]:
I think so. That's a hard one.

Leo Laporte [00:49:15]:
Talk me off the ledge.

Lou Maresca [00:49:19]:
I mean, there's still some restrictions today, so I think it's, it's not going to be, you know, I think there is some potential there to, for, you know, exponential growth.

Leo Laporte [00:49:26]:
Isn't self-improving AI kind of a holy grail at this point?

Lou Maresca [00:49:29]:
It really is. It is. Yeah.

Leo Laporte [00:49:31]:
Yeah. Uh, whether it's through a social network that AIs join or just in a, you know, I mean, I think one of the reasons Claude has been making such steps forward, um, is that it, that they are now, Anthropic folks are now using Claude or Vive coding Claude to build Claude. Yeah, to build Claude. Yeah. And that's sped it up. It used to be, you know, they have a new model every year and then it's every 6 months. Now it's certainly every 2 months, probably going to be every 1 month this year.

Lou Maresca [00:50:02]:
Right.

Leo Laporte [00:50:02]:
Who knows what by the end of this year, I don't know where we're going to be. One more thing, Lou, I'd like to ask you as an expert. I ask everybody this. It seems to be that the— I prefer— I would prefer to use an open-weight model. I want— I mean, that's why I bought this Framework desktop with 128 gigs of RAM and is I want to run this stuff locally. It seems to me that local models are not— I mean, open weight models are not too far behind the big expensive models.

Lou Maresca [00:50:30]:
That's true. Is that true? You have to have the hardware though. I mean, that's the thing. I mean, like, like if you want Qwen on a local desktop, you need like a couple GPUs that are smaller, but you can still run the inference. But the bigger models, the ones that have all of this ability to reason and and, and, and do massive inference, you need massive machines. You need machines that have hundreds of GPUs. And so even if they were open source, you still wouldn't be able to run them unless you had a server farm to do it.

Leo Laporte [00:50:57]:
Okay.

Lou Maresca [00:50:58]:
Yeah.

Leo Laporte [00:50:59]:
So uh, the, they call them hyperscalers, are still going to be in charge, right?

Lou Maresca [00:51:04]:
Right. That's why these enterprises are able to do it. So, but you know, these services are getting cheaper. It's just that you still have to provide data to, to a cloud service. Unfortunately, you can't run on the edge.

Leo Laporte [00:51:14]:
Right. Damn it. All right. One more break and then we will move on from AI. There were earnings this week. Google, phenomenal earnings. Amazon, phenomenal earnings. And these earnings are coupled with spendings.

Leo Laporte [00:51:35]:
We'll talk about that in just a little bit. Lumaresca is here. He is engineering manager at Copilot for Microsoft. Is it just Excel or is it, has your role expanded?

Lou Maresca [00:51:47]:
No, it's just Excel for now.

Leo Laporte [00:51:49]:
Yeah. But that's just, just Excel.

Lou Maresca [00:51:51]:
Yeah, I know. It's just Excel.

Leo Laporte [00:51:52]:
Just Excel. Just the tool, number one tool used by financial analysts, by data scientists, et cetera, et cetera, et cetera. You helped put Python in Excel, which when that happened, I thought, wow. Wow.

Lou Maresca [00:52:05]:
And we're still expanding that too.

Leo Laporte [00:52:06]:
That's really cool. Great to have you, Mike Elgin, soon to be, what is it, Attachment?

Mike Elgan [00:52:13]:
The Attachment Economy.

Leo Laporte [00:52:14]:
TheAttachmentEconomy.com, machinesociety.ai right now, gastronomad.net. He's in Oaxaca enjoying the food and the sunshine.

Mike Elgan [00:52:24]:
That's right.

Leo Laporte [00:52:24]:
Very jealous. And the excellent chocolate, by the way, talking about chocolate, they make it locally. You took us when we went on a Gastronomad adventure a few years ago, you took us to A chocolate tasting.

Mike Elgan [00:52:35]:
Mama Pacha and that place, it was written up in Wired for their innovative use of, they're basically makers and they made their own chocolate processing equipment.

Leo Laporte [00:52:45]:
And the cacao beans are grown in the hills above them. I mean, it's in the air, it's local. And they make it, and maybe you've had Mexican chocolate, which I love, with cinnamon and other spices in it, but they make the best, they make chocolate that rivals any Belgian chocolate you've ever had. It's amazing.

Mike Elgan [00:53:02]:
Amazing.

Leo Laporte [00:53:02]:
And the tasting was so much fun.

Mike Elgan [00:53:04]:
Yeah. Yep.

Leo Laporte [00:53:06]:
Love it. So thank you, Mike, for making me a chocolate addict. And Larry Magid, who is also a chocolate addict, president and CEO of Connect Safely. Notice neither one of us have run to the candy box. Is that your dad? A picture of your dad over your left shoulder there?

Larry Magid [00:53:26]:
Well, let's see.

Leo Laporte [00:53:27]:
I can't— That handsome young man, or is that you?

Larry Magid [00:53:30]:
That's me. Oh, the handsome young man is me as a young man in a skinny.

Leo Laporte [00:53:34]:
Tie, skinny black tie, white shirt. Did you work at IBM?

Larry Magid [00:53:37]:
My dad is next to me, right?

Leo Laporte [00:53:38]:
Oh, nice.

Larry Magid [00:53:39]:
No, I didn't work at IBM. I was a customer of IBM's back in my late teens, early 20s, but, uh, never worked there.

Leo Laporte [00:53:45]:
Did you have a pocket protector under that jacket?

Larry Magid [00:53:47]:
No, I wasn't that nerdy. Close, but not that close.

Leo Laporte [00:53:52]:
It's great to have all three of you. Our show today brought to you by NetSuite. Every business is asking the same question these days. How do we make AI work for us? Of course, the possibilities are endless, but guessing, it's too risky. And sitting on the sidelines, that's not an option because one thing's almost certain, your competitors are already making their move. No more waiting. With NetSuite by Oracle, you could put AI to work today. NetSuite is the number one AI cloud ERP ERP, trusted by over 43,000 businesses.

Leo Laporte [00:54:29]:
It's a unified suite that brings your financials, your inventory, your commerce, your HR, your CRM into a single source of truth. That connected data is what makes your AI smarter. So it doesn't just guess, it knows. Intelligently automates routine tasks, delivers actionable insights, helps you cut costs and make fast AI-powered decisions with confidence. This isn't just another bolted-on tool. It's AI built into the system that runs your business. Whether your company earns millions or even hundreds of millions, NetSuite helps you stay ahead of the pack. If your revenues are at least in the 7 figures, get NetSuite's free business guide, Demystifying AI, at netsuite.com/twit.

Leo Laporte [00:55:15]:
That guide is free to you at netsuite.com/twit. netsuite.com/twit. We thank them so much for supporting This week in tech, uh, Microsoft had a great quarter, Apple had a great quarter, uh, Alphabet had a great quarter. Revenue in 3 months of $113.8 billion, that's up 18% year over year. Net income almost $3 billion a week, not bad. Not bad. But, and this is the same story for Meta. Meta's stock, Meta had a great year, but their stock tanked because they decided they're gonna spend a lot of money.

Leo Laporte [00:56:04]:
And what is most of that money gonna be spent on? You guessed it, AI. To meet customer demand and capitalize, this is Sundar Pichai talking, and capitalize on the growing opportunities ahead of us, our 2026 capital expenditure investments are anticipated to be in the range of $175 to $185 billion a year.

Lou Maresca [00:56:33]:
Looks like chump change when you hear Microsoft has $600 billion in AI commitments.

Leo Laporte [00:56:37]:
Yeah. And Amazon, $200 billion in AI spending planned.

Mike Elgan [00:56:44]:
Uh, and if you add them up, it's Alphabet, Microsoft, Meta, and Amazon together are looking at $700 billion, which— think of that number.

Lou Maresca [00:56:54]:
Yeah.

Leo Laporte [00:56:54]:
Is that all data centers, or— that.

Mike Elgan [00:56:56]:
Would fill the stadium with $20 bills.

Lou Maresca [00:57:02]:
Data centers, it's it's power, power needs, it's, it's all that. It's data, getting the data. Where are you going to find the data to train these things?

Leo Laporte [00:57:09]:
Meta and Microsoft, same, same basic story with a CapEx through the roof. Meta's CapEx last quarter was $22 billion, and that number going to go up again. They say they're going to spend about $115 to $135 billion for the year. Microsoft in one quarter spent $3 billion a week, $37.5 billion, uh, capital expenditures towards CPUs, GPUs mostly, again for AI demand. Not that Microsoft's revenue was hurt, it was up $81.3 billion, 17% year-over-year growth. So these companies are having huge you profits, know, if they're spending it all.

Larry Magid [00:58:03]:
It's depressing. I just, for some reason, as you were talking about this, my 1099 to Connect Safely from our donations to Meta arrived. And I'm looking at the number and it is nothing compared. I mean, I don't even know how many billions of fraction of zeros away from the kind of money you're talking about. I mean, it's And it's a good thing. By the way, we're these pennies. punishing companies.

Leo Laporte [00:58:24]:
We're.

Larry Magid [00:58:27]:
Grateful for their contribution.

Leo Laporte [00:58:29]:
I want to reiterate that. Microsoft had a huge— I can't remember what it was, but it was a huge I uh, mean, they're at 12%. I think their stock went down in, uh, in, in an hour. Uh, and it's not because they're not profitable. It's because they announced they're going to spend a lot of money on AI. And I think the market is looking at this and saying, well, where's the money in AI that you're going to be I spending? honestly— what, after what we just talked about, it feels to me like there is an upside. Absolutely. I think not only do you have to spend this do money, you, as a uh, you hyperscaler, know, because you're competing against everybody else and they're spending it.

Leo Laporte [00:59:09]:
But there— but the— but it don't—.

Benito Gonzalez [00:59:11]:
Is—.

Leo Laporte [00:59:11]:
I is it a winner-take-all market? No, I think there's— there's— no, there's plenty to go around. I think the upsides are very strong. I know the market doesn't believe that, right?

Lou Maresca [00:59:20]:
That's the problem is I think that people are freaking out because everyone's saying they're wasting money. We haven't seen the fruits of the labor just yet, and we're going to put our money elsewhere. But I really think that this is the year of, you know, I coined it right now, the decision-making as a service. It's really where these agent companies are going, and that's where the money is going to make it happen.

Leo Laporte [00:59:39]:
So yeah, and you have to spend at this point, you have to spend to make it. Of course, does a lot of this money go in NVIDIA's pocket? I think so.

Lou Maresca [00:59:52]:
Well, there's TSMC. There's a lot of companies now that.

Leo Laporte [00:59:54]:
Are doing very well. In fact, TSMC, which, you know, basically was a captive company by Apple— Apple was using them for all of their chips— has said, yeah, not so much anymore. Uh, Apple, get in line because, uh, you know, there's a lot of demand for our chips from Nvidia and other companies.

Larry Magid [01:00:11]:
Has anybody tried to buy any RAM lately?

Leo Laporte [01:00:13]:
Yeah, look at that.

Larry Magid [01:00:14]:
It's not cheap.

Leo Laporte [01:00:16]:
I, uh, I had that intuition. That's why I bought the 128GB Framework desktop 6 months ago, and and I, I bought a new laptop a few months ago because I, I don't know what it's going to look like into this year, but RAM prices have more than doubled just because of this. Yeah, companies like Micron have said we're not going to sell to consumers anymore, there's too much money to be made selling to Um, and in fact, uh, computers enterprise. may.

Larry Magid [01:00:39]:
Be over.

Leo Laporte [01:00:40]:
It's the Steam Machine, which, uh, is a pretty exciting gaming machine from Valve. Is now delayed, and they're saying it's gonna probably cost more than we thought because we can't make them. They're too expensive. So, but you know what, this is capitalism, right? This is— is it overheated? Is the economy overheated, or is it a bubble, or is it reasonable investment? To build something that's going to change everything.

Larry Magid [01:01:15]:
But the problem that I'm worried about is, is it possible for the startups to compete with these mega oligopolical, you.

Leo Laporte [01:01:23]:
Know— well, that's what we're basically saying, you know, with Lou, is that billions.

Larry Magid [01:01:27]:
Of dollars to invest inside these companies.

Leo Laporte [01:01:29]:
It'S the, it's the hyperscalers who are going to win. So they're competing against one another.

Larry Magid [01:01:33]:
I don't even know if I would call that capitalism anymore. I don't know what I would call it. I mean, future.

Leo Laporte [01:01:40]:
Should we, should we, should we be worried?

Larry Magid [01:01:43]:
I think so.

Leo Laporte [01:01:45]:
In what way? So, uh, I mean, there's all kinds of ways. Is it going to crash the economy?

Larry Magid [01:01:51]:
It's very common for industries to consolidate, but usually they start out with a lot of players. I mean, if you go back to the car industry, the beer industry, any industry, there'd be a ton of players, then eventually it would get down to 3 or 4 big players. Right now there's just a handful full of big players that are starting out. And I don't even know how the Anthropic, you know, how do you compete with Microsoft, Meta, Google? I mean, I just don't know how you compete with these giants and OpenAI at this point. It just makes it really difficult. I mean, maybe around the edges with niche products, maybe that's the way it'll happen. But I just worry a little bit about a very tiny number of companies controlled by a very small number of people having this much influence on our entire world. Yeah, it's just a scary thought, but I've had scarier thoughts before.

Leo Laporte [01:02:40]:
Well, you know what's scary for people like you and me, Lou, is that we— our retirement savings are all in the stock market, right?

Lou Maresca [01:02:49]:
Yeah, me too.

Leo Laporte [01:02:51]:
And, um, you know, on the— I just don't know what to do. I just don't know.

Lou Maresca [01:03:00]:
You know, I, I have, you know, it's one of those things. I have faith in this because obviously I'm a little biased. I'm working in the industry, working a company that does it.

Leo Laporte [01:03:07]:
But that's why we should ask you, because you are closest to it.

Lou Maresca [01:03:10]:
But I think the, the thing here is that obviously there's so much money being put into it. You know, I talked to so many people, whether it's, you know, people at schools or organizations, and they just don't have the permission to even use AI at this point, and they don't see the benefit of it. And when you see corporations like GM and these other companies that are actually pivoting because they're using AI to get the kind of the noise outta the way and they're changing the way they do business, like that's where the power lives. And once people start to realize that, you know, I do, there's some fear around, you know, AGI and all that stuff, but there's so much power behind this and so much freedom behind it that once it becomes accessible, lives change. And so I'm hoping that's what happens this year and the stock market shows.

Larry Magid [01:03:54]:
Well, I think I see the other problem. I'm sorry. CEO of a very small nonprofit. Our productivity has skyrocketed in the last few years. I mean, I can't even begin to tell you what used to take us weeks. We now do in less than a day. It's just incredible how powerful this stuff can be.

Leo Laporte [01:04:10]:
This is better than the internet, better than personal computing. This is a transformative technology on the order of fire.

Larry Magid [01:04:20]:
Steam?

Lou Maresca [01:04:22]:
I mean, I don't know.

Leo Laporte [01:04:23]:
This is a new industrial era. This is a new year.

Lou Maresca [01:04:26]:
Bigger than that. I mean, I don't know about you, but I'm an electrical engineer, but the other day I had a model generate a new PCB for me and send it out to get it fabricated, and that would have taken me months to do, and it took me hours to do. So it's just, it's, it's transformative in a way that people don't even realize, I think.

Mike Elgan [01:04:43]:
So I think we're at a, we're at an intermediate point where we're using chatbots and we think that's what AI is. In fact, where it's going is, is agentic. We're going to have agentic systems that are based on personal assistance and they're going to come to us through wearables and mostly glasses. And that, what does that do? That makes us cyborgs basically. That basically we have a peripheral device that is prosthetic brain power and total knowledge. That's the revolution as far as I'm concerned for personal use and also business. Behind the scenes, we have AI that's gonna completely revolutionize medicine, science, technology, space travel, all the rest. And so this, we, again, I would encourage everyone who pays attention to AI, it's this AI chatbot thing where you go type something and it types out an answer.

Leo Laporte [01:05:33]:
That's not it. It's not it.

Mike Elgan [01:05:35]:
This is quaint. This is nothing.

Leo Laporte [01:05:38]:
And I think a lot of people, and I'm gonna blame Microsoft a little bit, Lou, I apologize. A lot of people— Google and Microsoft are the worst culprits in this— for whom AI is pushed on them. You know, Copilot's pushed on you in Windows, Google Drive, Google Docs, constantly pushing AI on me. Look at it, they see it as trivial, not helpful, annoying, and decide that's AI. And so there is certainly a large group of people, maybe even a majority of people, say AI is nothing, it's terrible, I hate it. Stop pushing it on me. You know, Google search, you know, with all that AI crap, they don't see what we see, that where AI is really truly useful, because they're— that it's just— they're not using it that way.

Lou Maresca [01:06:25]:
Yeah, yeah, it's too much noise. I agree.

Mike Elgan [01:06:26]:
I would love Meta in there as well because people are seeing AI slop, and that's what— on Facebook, on social. Yeah, the general public, and And that's one of the things they say, oh, that's AI, that's, that's what AI is.

Larry Magid [01:06:38]:
If you're watching this video, you can see I'm wearing the Meta glasses, which are kind of cool.

Leo Laporte [01:06:44]:
You should have warned us that you're taking— oh wait a minute, we're taking pictures.

Larry Magid [01:06:48]:
The battery's dead and it's been dead for about a fix. I can't even just like them. Yeah, I just wear them because they're glasses.

Leo Laporte [01:06:55]:
I mean, did you get your lenses?

Larry Magid [01:06:58]:
You're— oh no, these are prescription glasses.

Leo Laporte [01:06:59]:
Yeah, okay.

Larry Magid [01:07:00]:
And the prescription cost a lot more than the lenses. The, uh, what is that, Lens Crafter has a monopoly in my area, so, you know, you pay a lot for them. That— those Meta glasses?

Leo Laporte [01:07:10]:
Yeah.

Larry Magid [01:07:11]:
Good on you. Well, if you and I were to take each other's picture, would it— would there be an infinite loop?

Leo Laporte [01:07:16]:
Yeah, this is the problem with these. They all look nerdy, right? Maybe that's the goal.

Mike Elgan [01:07:22]:
Uh, well, that's all going to change very quickly, and I think that this is where we decommodify— commoditize AI, because people are going to use the AI.

Leo Laporte [01:07:30]:
That'S suddenly be useful, whatever hardware they have.

Mike Elgan [01:07:32]:
Yeah, exactly. Place. So I think there's going to be a huge— there already is a huge battle to come out with glasses that.

Larry Magid [01:07:38]:
Don'T— I just had a scary thought. So I'm thinking about getting a cataract operation. I don't need one yet, but at some point. And my doctor was saying they could put a new lens in for me. Could I get a smart lens?

Leo Laporte [01:07:47]:
Oh, wait till those are available.

Larry Magid [01:07:49]:
Yeah, that's a scary thought.

Leo Laporte [01:07:52]:
Make sure it's removable. The problem is you'll need another one a year later for the upgrade.

Larry Magid [01:08:00]:
Exactly. That's right.

Mike Elgan [01:08:02]:
You're going to see what What that's like when Apple comes out with Neuromancer later this year.

Leo Laporte [01:08:08]:
It's just a thing you plug into the back of your skull, right?

Mike Elgan [01:08:12]:
I call it Microsoft.

Leo Laporte [01:08:13]:
I think one of the things that scares people also is this constant talk of an AI bubble and the stock market crashing. Maybe we already had that. Hedge funds— we're February 8th— this year already have made $24 billion shorting software stocks. And why did software stocks tumble? Because the market suddenly realized, oh, you don't really need LegalZoom. You could just have your agent write the legal documents for you. You don't need a lot of this software anymore. Now, whether that's the— is that the bubble bursting? It's not exactly what what they, they thought the bubble burst would be.

Mike Elgan [01:08:52]:
I think to a large extent, that's, that's, you know, everything behind the stock market is people, right? Yeah. And they are delusional.

Leo Laporte [01:09:00]:
You don't believe in the wisdom of the crowds then?

Mike Elgan [01:09:02]:
I don't. I don't think they're— I don't think everybody understands where AI is going. Nobody understands where nobody's going. But one of the, one of the mistakes everybody makes is they get caught up in the hype and they think, oh, we're going to be able to lay off half our employees and replace them with chatbots. Well, that's not how it works, actually. What AI is going to do is it's going to magnify the capabilities of employees. The companies that do that are going to have much greater capabilities and to compete with those companies, you're also going to have lots of employees with magnified capabilities. It's not about laying people off.

Mike Elgan [01:09:35]:
You know, there's a thing called AI washing that's happening right now where because of tariffs, because of economic uncertainty, because downturns, the market and inflation, lots of companies are laying people off. But to make themselves look good as CEOs, they're saying, oh, it's because we're so— they're blaming us because of AI now.

Leo Laporte [01:09:51]:
Yeah, but it's not really what's Exactly.

Mike Elgan [01:09:52]:
But that's mostly BS. That's right. Exactly.

Leo Laporte [01:09:55]:
In fact, I've seen reports that customer service is actually improving thanks to AI. Certainly things like Jira tickets are getting responded to faster. You know, I think in many cases normal users go, oh, that's when I call a phone number and it's not a human inside, or you're chatting with a bot. On a website, and that's invariably a terrible experience. But that's not just what's where AI is changing customer service. And I think in some ways it's a huge improvement. I think it's very hard for us to really know what's happening. And I think it's even harder for normal people.

Larry Magid [01:10:35]:
Well, it's funny because where I might have otherwise called customer service, I'm sometimes just using AI to answer a question.

Benito Gonzalez [01:10:42]:
Bingo.

Larry Magid [01:10:42]:
How do I get this software to work? Well, it doesn't always get it right. It gets it wrong a lot. But eventually it'll— it's better than trying to call, be on hold for 4 hours.

Leo Laporte [01:10:50]:
I use AI now to configure all my computers. Yes, I could look it up, I could do the web search, I could ask Substack— whatever happened to Substack? But, uh, not Substack, uh, what's the, uh, Stack Exchange? I could go out. Yeah, I I could, could— no, I don't have to do that anymore. Uh, my— when I got that new laptop, I literally configured it all with Claude. Cloud Code.

Mike Elgan [01:11:14]:
And you, and you probably bought it with the recommendation that came through AI as well. I think AI is very good at making recommendations for things.

Leo Laporte [01:11:22]:
I did. I asked Coggi, I said, what's what's a, the best OLED laptop for running Linux? And it gave me some very good choices. I ended up buying a ThinkPad X1 Carbon. I was very happy. I wish it had told me, and there's going to be a new Intel chip in 2 months that's going to be even better.

Larry Magid [01:11:40]:
But so you have to ask the question right. I remember when I bought a some kind of device. I wanted to have auto shut-off. So I said, I want to buy a toaster with a timer. Wasn't a toaster, but something else. And it gave me one with a timer, but it was just a timer that rang. It didn't— wasn't a timer that turned it off. So then I had to go back and say, I want one with a timer that will automatically shut off the device.

Larry Magid [01:11:59]:
And then it found some of those for me as well. But, you know, if you don't ask the question right, you're not going to get the right answer.

Leo Laporte [01:12:05]:
Let's take a little break. I'm trying to get us out of here before some big game or something starts. I don't know. I hear there's some sports ball— sport ball is coming up later.

Larry Magid [01:12:14]:
It's the Bad Bunny concert. Or are you going to go to the other one? You're going to go to the one?

Leo Laporte [01:12:18]:
No, the other one was canceled.

Larry Magid [01:12:19]:
It was canceled. Are you serious?

Leo Laporte [01:12:23]:
The Kid Rock canceled? Yeah.

Larry Magid [01:12:24]:
Oh my God. I hadn't heard that. That's amazing.

Leo Laporte [01:12:27]:
Fox was going to show the MAGA Super Bowl concert so that you wouldn't see a Puerto Rican dancer.

Larry Magid [01:12:34]:
God forbid. We only have to have Americans, not Puerto Ricans, just Americans. Oh, wait a minute.

Leo Laporte [01:12:39]:
Wait a minute.

Larry Magid [01:12:39]:
I forgot Puerto Rico is a part of America.

Leo Laporte [01:12:41]:
It's part of America. I think the president of the United States is not clear on that, by the way.

Larry Magid [01:12:46]:
So they actually canceled it.

Leo Laporte [01:12:48]:
They canceled it. That's the last I saw. Who knows? Things could— things— it's a fast-moving story. You know what? Here's the good news. If you don't want to see Bad Bunny, go watch the Melania movie. There's choices. You have choices.

Larry Magid [01:13:02]:
Plenty of seats.

Leo Laporte [01:13:04]:
It's not hard to get a ticket.

Lou Maresca [01:13:07]:
Yeah.

Leo Laporte [01:13:07]:
We will have more in just a bit. Now that we've offended half the nation— no, a third of them— no, a few people, 40%. Larry Magid, it's your fault.

Mike Elgan [01:13:16]:
LarryMagidConnectSafely.org.

Leo Laporte [01:13:21]:
Mike Elgin from MachineSociety.ai. And by the way, he does not speak for Microsoft, but he is engineering manager of Copilot, engineering leader of Copilot at Microsoft. Lou Emme. And I did— I should have said this earlier when we talk about my— you're not speaking for Microsoft, you have no inside knowledge, and you will recuse yourself of course if it's something of which you do have knowledge. You're just here as a friend of TWiT, longtime host of This Week in Enterprise Tech. We just love and, Lou, uh, he's one of the nicest people in the world. Um, so thank you, Lou, for being here.

Lou Maresca [01:13:57]:
Thank you for that.

Leo Laporte [01:13:58]:
Appreciate it. And you can't help it if he's a Patriots fan, it's not his fault, folks. Our show this week brought to you by Meter, the company building better networks. This is— I, I have to say, I was totally stoked. I got to talk to these guys a few weeks ago, and I was told— I had never heard of them. I had to tell me about this. So it was founded by a couple of network engineers years, a few years ago, who just knew the pain points of designing a network, right? And they said, we need a full-stack solution. From soup to nuts, we need to have a way to do it all because it's the only way to make sure everything works and you don't get that runaround where it's not our fault, it's their fault.

Leo Laporte [01:14:38]:
If you're a network engineer, you have been there, you know the headaches. Legacy providers, inflexible pricing, IT resource constraints stretching you thin— everybody knows that. Complex deployments across fragmented tools. Look, the network's mission-critical to the business. You're mission-critical to the business, but you're working with infrastructure that just wasn't built for today's demands. That's why businesses are switching to Meteor. You know who uses Meteor for their data centers? Reddit. Meteor delivers full-stack networking infrastructure, and they do it everything— wired, wireless, even cellular.

Leo Laporte [01:15:18]:
It's built for performance and for scalability and for reliability. Meteor designs its own hardware. That's— they realized you have to do this. They write their own firmware, they build their own software, they manage the deployments, they provide after-sales support, soup to nuts. They offer everything from— they'll help you with ISP procurement, security, routing, switching, wireless, firewall, cellular, power, DNS security, VPN, SD-WAN, and multi-site workflows all in a single solution. You know, I think part of this has happened because of of companies like these hyperscalers where the knowledge of how to build a data center has now migrated down to the point where Meteor, these problems are solved and Meteor knows how to do it. They've created a single integrated networking stack that scales. I mean, they're in major hospitals.

Leo Laporte [01:16:12]:
Talk about a hostile environment. A hospital is a very difficult place for wireless, right? 'Cause the MRI machines, all of the different equipment, They do it for branch offices. That's a big problem where you have offices in different geographic locations all on the same network. Meter can do it. You, you acquire a warehouse with their own weird wireless setup. Meter can transform it into something that works with your network. And large campuses, even data centers, even as I said Reddit. Here's a great testimonial from the Assistant Director of Technology for the Webb School of Knoxville.

Leo Laporte [01:16:45]:
He said, quote, we had more than 20 games, athletic contests on campus going on at the same time between our two facilities. 20! Each game was streamed via wired and wireless connections. The event went off without a hitch. We could never have done this before Meter redesigned our network. Oh, it's a dream come true. With Meter, you get a single partner for all your connectivity needs, from first site survey to ongoing support without the complexity of managing multiple providers, or all passing the buck, right? Or tools. Meter's integrated networking stack is designed to take the burden off your IT team and give you deep control and visibility, reimagining what it means for businesses to get and stay online. And these days, that's the job, right? Meter's built for the bandwidth demands of today and tomorrow.

Leo Laporte [01:17:37]:
We thank Meter so much for sponsoring. Go to meter.com/twit, book a demo now. That's M-E-T-E-R.com/twit.

Benito Gonzalez [01:17:43]:
.Com/Twit.

Leo Laporte [01:17:47]:
Book a demo today, and please use that address so they know you heard about it here. I was so impressed with these guys. Meter.com/twit. You're going to be hearing a lot more about these guys. They really— they've solved a big issue. Uh, on we go. Oh, Chief Twit was wrong. It has not been canceled, right? Multiple outlets have debunked— oh, okay, it's not canceled.

Leo Laporte [01:18:10]:
Just a lot of the artists have dropped out. But if you like Kid Rock, You.

Larry Magid [01:18:14]:
Can watch it tonight. No, he canceled a concert that he's scheduled to do in South Carolina this summer.

Leo Laporte [01:18:19]:
Oh, that's a different one.

Larry Magid [01:18:21]:
When I did a Google search for this, I apologize.

Leo Laporte [01:18:23]:
Okay.

Larry Magid [01:18:24]:
But not apologize. But if you want to watch him today, get a Bad Bunny.

Leo Laporte [01:18:27]:
Fox has got it.

Larry Magid [01:18:28]:
You can do it.

Leo Laporte [01:18:29]:
Fox has got it.

Larry Magid [01:18:31]:
I'm recording both.

Leo Laporte [01:18:32]:
Let's talk about privacy security. Here's a story that'll— that'll— from the BBC. Hidden cameras in Chinese hotels are streaming broadcasts to thousands over Telegram. You can just log in, uh, watch real people doing intimate things because apparently this spy cam porn has existed in China for at least a decade according to BBC. But in the past couple of years, the issue has become a regular talking point on social media with people swapping tips on how to spot cameras as small as a pencil eraser. Some have even resorted— this is the BBC. To pitching tents inside hotel rooms to avoid being filmed.

Mike Elgan [01:19:17]:
Is that a euphemism?

Leo Laporte [01:19:20]:
I pitched a tent.

Lou Maresca [01:19:21]:
Uh.

Leo Laporte [01:19:23]:
What'S weird is— thank you, thank you for bringing it to that peak. Uh, 6 different— and over 18 months, the BBC authors discovered 6 different websites and apps promoted on Telegram, between them claiming to operate more than 100 180 hotel room spy cams, live streaming spy cams.

Larry Magid [01:19:45]:
Scary.

Leo Laporte [01:19:46]:
Anyway, I just— I guess that's in.

Larry Magid [01:19:47]:
China, or do we have to worry about that?

Leo Laporte [01:19:49]:
Well, that's a good question, and I guess that's why I raise it. I mean, we have listeners in China, so— but I bet you if it's happening there, why wouldn't it be happening here?

Mike Elgan [01:20:03]:
Well, it's a, it's a, it's a much bigger phenomenon in China. I think I was writing about this like 5 or 6 years ago without the Telegram and social media aspect, but it was like, at the time it was kind of an underground phenomenon.

Leo Laporte [01:20:17]:
But fairly widespread. Now you can find it.

Mike Elgan [01:20:19]:
And the thing is, yeah, it's just one of those things that's, it's like smartwatches on children, right? It's China leads the world by orders of magnitude ahead of everyone else. This is just another thing that's just happening there It shouldn't be happening anywhere, obviously. But, but yeah, it's still around. It's still a big deal. And the key element of it and the reason that Chinese operators aren't going global is that you have to have access to the hotel to plant the camera.

Leo Laporte [01:20:52]:
Right.

Lou Maresca [01:20:53]:
Right.

Mike Elgan [01:20:53]:
And so it's really so far, thank goodness, more of a Chinese phenomenon. But it's something that you have to keep in mind if you travel to China, because I'm sure that, you know, I'm sure that the audience would love to see some foreign victims.

Leo Laporte [01:21:13]:
You will be seeing a few ads on the Super Bowl for AI. We certainly saw that last year. And I mentioned that OpenAI and Anthropic will have dueling ads, but also you'll see a lot of ads generated by AI this year. Uh, and I think it's another reason people hate AI is because of the, the Coca-Cola ad which was played incessantly over the holidays was, I think, intentionally done badly. They wanted— McDonald's is horrible too. Yeah, they pulled that one. That wasn't in the U.S., but they had to pull that one. Um, so one of the first ads you'll see is for a vodka, Svedka, in which, uh, a voluptuous humanoid woman and her muscular male companion are dancing in a club with bottles in hand.

Leo Laporte [01:21:58]:
They're robots. So it's kind of— it's AI about AI drinking vodka. Why would you want to drink a vodka that robots drink?

Larry Magid [01:22:09]:
Uh, why would you want to feed vodka to your robot?

Leo Laporte [01:22:12]:
Don't, right?

Larry Magid [01:22:13]:
Scary enough without this.

Benito Gonzalez [01:22:14]:
Yeah, I think this might be to get around the rules that you can't be drinking— a human can't be drinking alcohol on screen during an ad, but a robot can.

Mike Elgan [01:22:26]:
The problem is that culturally we're entering into an era where AI equals cheap. And so this is one of the big thing problems with the Coca-Cola ads. Coca-Cola has some legendary ads. So does You McDonald's. know, if you like advertising, they've had some very masterful, beautiful, brilliant, well-engineered ads, well-written, great jingles, all the rest. And then to do an AI slop. For example, the Coca-Cola commercial, there were nerds on Reddit who were counting the number of axles. So within one commercial, the axle configurations on the trucks that were depicted, there was like 18 of them or something like that, different configurations of axle.

Mike Elgan [01:23:11]:
AI is just making up axle configurations on trucks, things like that.

Leo Laporte [01:23:15]:
So that's the thing that they probably did. I'm thinking could have fixed but decided not to. For some reason, they wanted people to know this was AI-generated, but they had.

Mike Elgan [01:23:24]:
To go running with their tail between their legs and retreat and take them down and apologize and all that stuff. They keep doing it. Like, these big companies keep coming out.

Leo Laporte [01:23:33]:
With— you're spending $8 million for a 30-second ad.

Mike Elgan [01:23:38]:
One thing, here we are talking about it, right?

Leo Laporte [01:23:41]:
Well, that's true. My son will be in the Hellmann's mayonnaise ad. Look for him. He's on for less than a second, so you're gonna have to look very, very closely. It's an ad.

Mike Elgan [01:23:52]:
Look for a mustache.

Leo Laporte [01:23:53]:
Yeah, you'll see the mustache. That's why you know immediately it's Saul Hank. It's a Hellmann's ad featuring Andy Samberg as Neil Diamond singing Sweet Sandwich Time. And when it gets to the ba ba ba, Henry's one of the ba ba ba. So just look for the— when it gets to ba ba ba. Um, Gemini becomes an AI interior decorator in Google's Super Bowl ad. You'll see it's the new home commercial featuring nostalgic piano music and the heartfelt voiceover of a mother and son envisioning their new home with the help of Gemini. Um, so sweet, sweet, sweet.

Leo Laporte [01:24:34]:
I think that's— maybe they're trying to humanize it, right, at this point, instead of vodka-drinking robots. Crypto.com is going to launch— it spent, by the way, hundreds of millions of dollars on AI.com. Somebody was, somebody was sitting on that domain. I can't remember what the final amount was.

Larry Magid [01:24:53]:
I had thought of that.

Leo Laporte [01:24:54]:
It was several hundred million dollars for that. They're going to promote it, uh, it says, as a way for users to, quote, generate a private personal AI agent that doesn't just answer questions but actually operates on the user's behalf. You'd trust crypto.com to do that, wouldn't you? AI.com, uh, you're going to see a word plus.

Larry Magid [01:25:20]:
I do now. I, I didn't at first, but at the last line of this, it's for the good of humanity.

Leo Laporte [01:25:25]:
Oh, well, that's different. Oh, I problem. wish— Yeah, I'd known.

Larry Magid [01:25:29]:
Yeah, it scared me for a minute, Leo, But it does confirm that this.

Leo Laporte [01:25:33]:
Is going to be the year of agentic AI for sure. I mean, this year you nailed it, Mike. Amazon is going to have an A-word plus. It's going to try to kill Chris Hemsworth. A standoff with Amazon's AI assistant, which he fears is planning elaborate ways to kill him. I do see that seems like a mistake. Seems like a mistake on Amazon's part. Um, we mentioned Anthropic's Super Bowl ad.

Leo Laporte [01:26:05]:
Uh, there will not be any ads for, uh, the prediction markets, Calci or Poly Market. It's a prohibited category according to the NFL. You know why I suspect it's prohibited? Because they love DraftKings.

Mike Elgan [01:26:22]:
So, so, Leo, interestingly, there was just a big earthquake here in Oaxaca.

Leo Laporte [01:26:27]:
No, I hear the— I hear the horn and it's still going.

Mike Elgan [01:26:30]:
So I'm having an earthquake right now.

Leo Laporte [01:26:33]:
Oh yeah, you're shaking. Are you want to go? You want to dive under your desk if the power goes out? I think his internet is going out.

Larry Magid [01:26:47]:
I'm going to usgs.gov right now to figure this out.

Leo Laporte [01:26:50]:
Yeah.

Larry Magid [01:26:51]:
Wow.

Leo Laporte [01:26:54]:
Not the first earthquake we've had during a live show, by the way. Real-time data might be the strongest though. It still looks like it's still shaking. Uh-huh.

Larry Magid [01:27:02]:
Yeah, you can go— you could put USGS on the screen. Here it is. Cruz Bay. Is that Oaxaca? No, there's earthquakes everywhere.

Leo Laporte [01:27:12]:
Yeah, I know. I used to follow— there's a Twitter— used to be a Twitter account where you could get tweets when there was an earthquake, and it was just like constant. So I stopped.

Larry Magid [01:27:19]:
Yeah, I don't see the one in Oaxaca yet on the screen.

Leo Laporte [01:27:22]:
It's happening.

Lou Maresca [01:27:23]:
5.7, it says.

Leo Laporte [01:27:24]:
Oh, that's strong. Oh, that's really strong.

Larry Magid [01:27:28]:
Okay, is he okay?

Leo Laporte [01:27:31]:
Mike, are you still there?

Benito Gonzalez [01:27:32]:
Yeah.

Mike Elgan [01:27:34]:
Yes, I'm here. I, I just got my audio back. Yeah, there's a big earthquake. My phone lit up, the sirens went off in town, and, uh, but yeah, it's an It's another earthquake in Oaxaca.

Leo Laporte [01:27:48]:
Is it a common thing?

Mike Elgan [01:27:49]:
Still kind of shaking, but, um, are.

Leo Laporte [01:27:52]:
You— you're actually not on a big fault?

Mike Elgan [01:27:54]:
I don't think it's that common. I know in— no, I— we just got back from El Salvador, which is constantly having earthquakes. But yeah, this was— I guess it's a little bit more rare here in Oaxaca.

Leo Laporte [01:28:05]:
Okay, well, 5.7.

Mike Elgan [01:28:06]:
The dogs are barking.

Leo Laporte [01:28:07]:
Pretty serious. Yeah. All right, well, stay safe. If you need to dive off, that's okay.

Mike Elgan [01:28:16]:
No, I'm sure it's fine. I'm, I'm good.

Leo Laporte [01:28:20]:
You're a Bay Area guy, earthquakes do not faze you. Although 5.7, 5.7 is pretty big.

Larry Magid [01:28:26]:
Yeah, what was the one? What was the Loma Prieta? Yeah, you remember?

Leo Laporte [01:28:29]:
5.9, I think. Yeah, I was at the World Series game at Candlestick Park. Yeah. During that. That was, uh, that was exciting.

Mike Elgan [01:28:38]:
Yeah. Wow.

Leo Laporte [01:28:40]:
We had to— I was in Santa.

Mike Elgan [01:28:41]:
Barbara watching it on TV and I was just like, oh my God.

Leo Laporte [01:28:44]:
I'm sorry, 6.9. I got that wrong. It was 6.9.

Larry Magid [01:28:46]:
So that was quite a bit bigger.

Leo Laporte [01:28:47]:
Yeah, that's, that's more than 10 times bigger. Yeah.

Larry Magid [01:28:51]:
But 5.7 doesn't— doesn't last that. It's funny because I'm looking at the USGS and I don't see it on there yet, which is surprising.

Leo Laporte [01:28:58]:
Yeah, I, I looked too. I didn't see it either.

Mike Elgan [01:29:01]:
I got a notification Notification on my phone instantly. That was really interesting.

Leo Laporte [01:29:07]:
That's something that— I mean, that's lately—.

Mike Elgan [01:29:09]:
The same time I felt it.

Leo Laporte [01:29:10]:
I, I've seen that lately. Uh, both Apple and Android do that.

Larry Magid [01:29:13]:
Yeah.

Leo Laporte [01:29:14]:
Uh, and, and you will get a second or two before the earthquake, just maybe just enough time to find safety, a notification that there's one coming. Depends, I guess, how close you are to the epicenter.

Mike Elgan [01:29:24]:
Yeah. Yeah.

Leo Laporte [01:29:26]:
Uh, well, let's take a break, and then that way, uh, Mike can, uh, adjust. And if you need to change your pants, you can go ahead and do that too. Just a little bit helpful. Hey, I'm nothing if not considerate. Mike Elgin visiting us from, uh, shaky Oaxaca, machinesociety.ai, gastronomet.net. Also from, uh, New England, Lou Mureska, who is a Patriots and a Seahawks fan. I don't know what he's going to do this afternoon. Engineering leader for Copilot at Microsoft.

Leo Laporte [01:29:59]:
You know, you usually have a green screen behind you. I didn't know you have beautiful windows in your office. You have a beautiful view. Is that snow out the window?

Lou Maresca [01:30:05]:
It is snow. Yeah.

Leo Laporte [01:30:07]:
The snow never melted this year, did it?

Lou Maresca [01:30:10]:
No. In fact, we just got another 8 inches yesterday.

Leo Laporte [01:30:12]:
So OMG, that is not typical.

Lou Maresca [01:30:17]:
No, this is— no, we usually don't get this much. Yeah, but, uh, it's been a lot.

Leo Laporte [01:30:21]:
Yeah, I remember in my youth in Providence we'd get snow, but there has— but it stopped after a while. And, uh, yeah, that's pretty snowy in New England right now. We'll stay, stay Stay safe too. Stay warm, more important.

Larry Magid [01:30:35]:
Yes.

Leo Laporte [01:30:35]:
I bet the boys are loving it though. They're running out.

Lou Maresca [01:30:37]:
Oh gosh, they love it.

Leo Laporte [01:30:38]:
Yeah, they adore it. Yeah. Uh, and of course Larry Magid, who is, uh, CEO of Connect Safely. Don't forget Tuesday, Safer Internet Day.

Larry Magid [01:30:48]:
By the way, I just checked. I went to ChatGPT. I said, is there an earthquake in Oaxaca, Mexico? And it said it's a 4.1. It led me to allquakes.com. Huh. But it it said, said 4.1, so I'm not sure how it and Mike's information are different.

Leo Laporte [01:31:01]:
But interesting.

Larry Magid [01:31:03]:
But it beat out USGS in reporting that news, which I found very— which is wild.

Mike Elgan [01:31:07]:
Amazing.

Lou Maresca [01:31:09]:
Yeah, yeah.

Leo Laporte [01:31:09]:
Our show today brought to you by a brand new sponsor. We're glad to have them with us, Trusted Tech. They offer US-based Microsoft-certified support using a simple ticket-based model that helps businesses save money while getting faster better help and proactive support. Trusted Tech is the number one global replacement to Microsoft Unified Support. Trusted Tech will work to get you better service no matter what size business you have. And in recognition of that support quality, Trusted Tech was one of the first partners in the world to earn Microsoft's new Solutions Partner designation for support. They just announced that at Ignite. In July, Microsoft will implement a significant price increase for M365.

Leo Laporte [01:31:56]:
Not only is it going to cost more, but there's a lot of nuance and, and a lot of changes in the coverage levels and so forth. So that's another thing Trusted Tech does. They will give you guided Microsoft support that's more straightforward, more predictable, and actually responsive, and it costs less. You can get a free consultation at trustedtech.team/twitcss. Trustedtech.team/twitcss. Now if you're saying, well, I don't know, what does Microsoft think about this? Well, ask Kevin Turner, former Microsoft COO. He says, and he was talking to Trusted Tech, he said, quote, you have an incredible customer reputation. You have to earn that every single day.

Leo Laporte [01:32:39]:
The relentless focus you guys, Trusted Tech, have on taking care of customers gives them value and differentiates you in the marketplace. They are certified support services. Trusted Tech elevates the Microsoft support experience with its certified support services. Probably that's why some of the best and biggest companies in the world use Trusted Tech. NASA, Netflix, Neuralink, Apple, Intel, Google, Lockheed Martin— they all save 32 to 52% compared to the average Microsoft Unified Support Agreement by going with Trusted Tech. Trusted Tech's Microsoft-certified engineers first respond within 10 minutes, achieving an 85.7% in-house ticket resolution rate and a 99.3% customer satisfaction rate. And Trusted Tech's flexible ticket-based monthly or annual pricing model offers direct escalation to Microsoft from a managed partner when necessary. The principal architect for TaylorMade, another great testimonial, says, we don't break glass often, but when we do, being able to quickly leverage Trusted Tech's professional services through the CSS program and get immediate engineer-level support has been invaluable to us.

Leo Laporte [01:33:54]:
Whether you're looking to fine-tune your Microsoft 365 licensing, improve the way your organization receives proactive Microsoft support, or both, Trusted Tech can do it. They offer free consultations to help you understand your options. At least do that. Look at your licensing now before the M365 price increases in July. Go to trustedtech.team/twitcss and submit a form to get in contact with Trusted Tech's Microsoft support engineers. Great information from great people. trustedtech.team/twitcss. TWIT CSS.

Leo Laporte [01:34:31]:
We thank them so much for their support for This Week in Tech. Trustedtech.team/TWIT CSS. One of the— we, we did talk about finance, but we didn't mention one of the biggest stories of the week, which is that SpaceX has acquired xAI. Now remember, it's all in the family. Elon owns them all. SpaceX is expected to have a big IPO later this year. xAI is expected to spend lots of money over the next 12 months. And during the announcement, Elon said something that a number of people have thought maybe a little odd, that SpaceX— and they did apply to the FCC for this— wants to launch a million, a million data centers in space.

Leo Laporte [01:35:24]:
Elon says it's a step to becoming a Kardashev 2 civilization. A number of people said, you know, data centers in space don't make a lot of sense. Elon says there's, you know, if, if you're running on solar power, space is 5 times more efficient than land-based solar power. They've really brought the cost of launches down. That's one of the things SpaceX did do with their reusable rockets.

Mike Elgan [01:35:54]:
Well, if you want to make a lot of money on, on Elon Musk's predictions. Just go to the prediction market and bet against him.

Leo Laporte [01:36:02]:
He has yet to deliver so many.

Mike Elgan [01:36:06]:
Exactly.

Leo Laporte [01:36:07]:
They want to launch a million tons per year of satellites, each generating 100 kilowatts of compute power— I'm sorry, 100 kilowatts of power per ton. So that would be 100 gigawatts of AI compute annually, Elon says, with no ongoing operational or maintenance needs. Well, you can't. It's in space, so you can't set a tech up, you know. If it doesn't work, you just deorbit it and launch another one.

Benito Gonzalez [01:36:34]:
You know what's really bad for computers is radiation, and there's a lot of that in space.

Leo Laporte [01:36:40]:
There is. Sometimes people think, well, it's very cool up there, but no, it's actually harder to cool a data center in space because there's no medium around it. There's no air around it for which the heat can go off.

Mike Elgan [01:36:52]:
Condition You can't have air conditioners if there's no air. The, the, you know, this is it's, basically— it's being, uh, pitched as a way for, uh, for Elon Musk's AI to have the future of data centers. But what it actually is is SpaceX is a winner of a company and X.com is a loser of a company, which is part of this deal. And so basically they're just kind of folding in the loser into the winner, then they'll have an IPO. And it's all just part of this thing you where, know, all the companies do this to a certain extent. They sort of obfuscate the of the sort parts of their business that are losing money, that aren't doing well, and couch it under an umbrella where it all seems to be doing pretty well. That seems to be above all what's happening.

Leo Laporte [01:37:40]:
Reid Albergotti writing at Semaphore has coined a new term for these Um, he says it's the natural evolution of big tech from primarily software-based companies to extraterrestrial industrial giants, EIGs.

Mike Elgan [01:37:56]:
Give me a break.

Leo Laporte [01:37:57]:
It's very sci-fi.

Mike Elgan [01:38:00]:
Yeah, I got another sci-fi thing for you. It's called the Kessler effect.

Leo Laporte [01:38:05]:
Oh yeah, that's not sci-fi.

Mike Elgan [01:38:06]:
Yeah, China wants a billion satellites in space sector, put a million days. It's all fun and games until they— until the dominoes start falling.

Leo Laporte [01:38:15]:
—And then you could actually cut off all the light from the Sun, and you wouldn't have to worry about global warming anymore. Nope.

Larry Magid [01:38:22]:
Mm-mm.

Mike Elgan [01:38:25]:
Or, or, or getting a sunburn.

Leo Laporte [01:38:27]:
So with the Space Shuttle, it cost $54,000 per kilogram to put something in orbit. Thanks to SpaceX and the Falcon Heavy, it's $1,400 per kilogram. Uh, while the cost of traveling by car has gone up 150% over the same period, at some point Elon says the— it'll be $200 a kilogram, and at that point launching data centers into space makes economic sense. Unlimited solar energy. There are challenges for cooling, but there are ways to radiate that heat off. I mean, to be fair, Reid also says a good rule of thumb when it comes to Musk is right call, wrong clock.

Larry Magid [01:39:15]:
Yeah, yeah.

Leo Laporte [01:39:18]:
I mean, I'm driving an electric vehicle and I probably wouldn't be if it weren't for Elon and Tesla.

Larry Magid [01:39:23]:
I'm still driving a Tesla, or actually it's driving me. I hardly ever touch the steering wheel.

Leo Laporte [01:39:28]:
You have a Model S?

Larry Magid [01:39:29]:
I have a Model And, 3. uh, it's— I've got The new, you know, the hardware 4.0.

Leo Laporte [01:39:34]:
FSD. They don't call it Autopilot anymore.

Larry Magid [01:39:36]:
They've got FSD. And as much as I hate to admit it, because I'm not an Elon Musk fan, it is very close to flawless now. It really has gotten very good since the days you sold your Model X where it was crap when you had it.

Leo Laporte [01:39:49]:
Yeah, Lisa always thought it was trying to drive her into the median strip.

Larry Magid [01:39:54]:
No, it's quite good now.

Lou Maresca [01:39:55]:
Yeah, I have a Model X and this weekend it drove into a pothole and I got a flat tire. Oh, not good.

Leo Laporte [01:40:02]:
That is, by the way, an underappreciated downside to electric vehicles. They're so heavy that they run through tires at a rapid rate.

Lou Maresca [01:40:10]:
$450 later.

Larry Magid [01:40:12]:
Yeah, your Model X is going to become a collector's item.

Lou Maresca [01:40:14]:
I know, I should keep it. I know, I love it. I, I do truly love it. It's just— but you know, it does have its nuances.

Leo Laporte [01:40:20]:
Unfortunately, I loved it too, but Lisa called it Christmas.

Larry Magid [01:40:23]:
You know what's funny? We're talking about AI, GROK is built into Teslas now. So all I have to do is push a button on my steering wheel and I can talk to GROK. And I do. I normally wouldn't use GROK, but it's right there. And it's not nearly as bad as its reputation. And its reputation is horrible.

Leo Laporte [01:40:39]:
Yeah. Because.

Larry Magid [01:40:39]:
That's.

Leo Laporte [01:40:39]:
Of Meta.

Larry Magid [01:40:39]:
A very low bar.

Mike Elgan [01:40:42]:
This is what I was saying earlier. These AI chatbots for most uses and most people are somewhat commodified. Commoditized, and it's the hardware that will determine which one you use in this case.

Larry Magid [01:40:55]:
But you know, if I'm driving down the road and I say I'm looking for a restaurant, it'll recommend one and it'll actually drive me there. It'll reprogram my, my GPS.

Leo Laporte [01:41:03]:
And do you have it, uh, do you have a cute voice?

Larry Magid [01:41:07]:
It's a very cute voice.

Leo Laporte [01:41:08]:
And, uh, do you get your choice of voice?

Larry Magid [01:41:10]:
She loves me. She, she's very close. Yeah, that's the problem. Yeah, yeah.

Leo Laporte [01:41:14]:
So.

Mike Elgan [01:41:17]:
So just a you quick, know, we talked briefly on Elon Musk's claims about data centers in space. I don't think we should move off the topic until we really understand what a terrible idea this is in the timeframe, again, that he's talking about. So the problems are, the costs are vastly higher than the cost of a data center on Earth. There's space debris, which is a problem nobody has an answer for. It's getting worse and worse with all these satellites costs going up. You did talk about the thermodynamics and the heat and the radiation and all those problems. Another one that people aren't really talking about is the maintenance and repair. You've got a million data centers in space, and data centers on Earth, which are far less complex and subject to gravity and things like that, need lots of maintenance, and they need people working on them and chipping away.

Mike Elgan [01:42:05]:
So that's another issue, getting people there and back. And then you have the the basic fundamentals of the market. Like, right now we have this big boom in data center development. There'll come a time where that's gonna back off and those things are gonna be very cheap because the boom will go too far. And so doing something in space takes like 10 times longer, right? And so how do you right-size the quantity of these things? It's just a big problem. Big set of problems.

Lou Maresca [01:42:37]:
There's another one to add to that too, is the security. Like, how are you gonna— like, it's like open season for governments to shoot them down, right? Like, if you have secure data up there, like, how are you going to secure it if they can just pluck them out of the sky?

Leo Laporte [01:42:49]:
That's probably where I should bring up the story about Russia intercepting EU satellites. This is from, also from, uh, Semaphore. Russian space vehicles have approached, physically approached European satellites intercept their communications, according to officials. Two objects have passed dangerously close to some of Europe's most important geostationary satellites. It's from the Financial Times. The satellites lack advanced onboard computers that could encrypt their transmissions, so the data is, is coming off in plain text, which also leaves them vulnerable to interference.

Mike Elgan [01:43:28]:
I read that they actually hacked them. Is that not the case?

Leo Laporte [01:43:32]:
Yeah. Um, well, it's certainly theoretically possible. This, uh, the parliament wrote it could provide a blueprint for sabotaging European space systems. Um, Moscow stepped up, I guess, space warfare. Fortunately, we've got Space Force to protect us. And so, you know, that is, that is the one thing I do not want to see is a space war. Uh, all right, so yeah, you know, I think my theory is that Elon, uh, I, I used to think Elon just smoked too much weed and used too much ketamine, and so as a result was kind of, uh, you know, kind of had some outlandish ideas. Now I'm thinking maybe it's stock manipulation.

Leo Laporte [01:44:23]:
Maybe he's just— because he's got an IPO coming up, maybe he just is hoping the stock market will buy it and then buy his stock. He wants to raise a lot of money for SpaceX.

Larry Magid [01:44:35]:
Tesla stock.

Leo Laporte [01:44:36]:
Yeah. And yeah, he's done very well, hasn't he? And he— there's— he's been accused of pump-and-dump schemes using, uh, X.

Larry Magid [01:44:43]:
Uh, oh, I've been enjoying the revenue from turning my Model 3 into a Robotaxi for years now.

Leo Laporte [01:44:49]:
Oh yeah, right.

Larry Magid [01:44:49]:
2018 is when the— right, right.

Leo Laporte [01:44:54]:
I mean, I don't know, but you can never read somebody's mind, what their intentions are, but it certainly benefits him financially if people buy this notion.

Mike Elgan [01:45:07]:
Well, this is the old Silicon Valley thing, right? And he's like the poster child for this kind of thinking. You always go after the maximum, most incredible, outlandish version of that. If you're Google, like, you know, 15 years ago, You don't say, you know, yeah, we're gonna grow email to capture a two-digit part of the market. You say, no, we're gonna expand until there are 6 billion people using Gmail. A ridiculous number, right? And you just go for it, right? You just go for the maximum thing you can possibly attempt. And you fall way, way short. In the case of Elon Musk, you fall way, way.

Leo Laporte [01:45:50]:
Yeah, there's also— there's one other thing that comes to my mind. Um, now by the way, I use Starlink. That's my— I have Comcast internet, but because Comcast is not 100% reliable, I have a failover to Starlink. Got a little dish up on the roof. And those are the only two high-speed choices available to me. So I don't— I would prefer not to use either, but that's the only— if I'm going to do shows from my attic, it's the only way I could do that. But there's also this issue of, well, if Elon's got a million satellites in space and Starlink is the dominant choice, Elon has huge geopolitical power. He's already— remember, he, you know, he used— he cut off Ukraine, and now Ukraine said Moscow's terminals have been cut off.

Leo Laporte [01:46:37]:
Russia is supposedly not supposed to use Starlink but has been. Illegally mounting Starlink systems on its attack drones. Activists in Iran use Starlink after Tehran imposed internet blackouts. But Elon holds the master switch to this, right? From Semaphore, relying solely on Starlink, given Musk's strong political views, quote, can be really dangerous from a country's sovereignty perspective. This is NPR. Global South, we trust Starlink, you know.

Mike Elgan [01:47:13]:
We saw that some time ago when, when Brazil ordered— a Brazilian judge ordered, uh, Musk to— or ordered X.com to remove a post. He refused, so they basically went through the carriers within Brazil to, to block X. But then Starlink was blasting internet and X into Brazil, and then they started going after those. And so it was basically this war between a sovereign nation and a single oligarch, basically, who was using multiple companies to override the legal authority and the sovereignty of the Brazilian government. And so, and at some point, Musk backed off, but he didn't have to back off. He could have not backed off. And so, but that was just a glimpse into the future of individual people who have the power of nation states or exceed the power of nation states. And here's it, Elon Musk represents an individual person.

Leo Laporte [01:48:10]:
Richest man in the world worth $800 billion.

Mike Elgan [01:48:14]:
Powerful space program in the nation of Brazil.

Leo Laporte [01:48:17]:
Yep. And, you know, having $800 billion doesn't mean you can have a bigger, well, it does mean you could have a bigger house and a nice yacht, But really what it means to Elon is a lot of power. And I think at this point it's not about accumulating dollars, it's about accumulating power. And do we want to allow him that power?

Mike Elgan [01:48:39]:
Wants to have more robots, more, more Tesla robots than there are people on Earth. Imagine that.

Larry Magid [01:48:45]:
What's interesting, I just, I just asked, for what countries have a lower GDP than Elon Musk's net worth? And there are several. Oh yeah, he is wealthier than many entire countries.

Leo Laporte [01:48:58]:
He's going to be our first trillion-dollar man.

Larry Magid [01:49:01]:
Yeah, pretty scary.

Leo Laporte [01:49:04]:
All right, uh, one more break and then, uh, what I call the back of the book, the, uh, the other news. Uh, we are having fun getting ready for the big game, which is about an hour off. We're gonna get you off the show in time. Lou, to root for both teams. You hope that— can there be a tie in the Super Bowl? I don't think so.

Lou Maresca [01:49:27]:
I think— I don't think all my kids are rooting for Patriots, so I might have to go the other way just to be the odd man out.

Leo Laporte [01:49:34]:
Do you have an appropriate jersey to wear?

Lou Maresca [01:49:37]:
You know, I should wear both. I have both. So my nephew's also— it's funny because my nephew is also a Seahawks fan, and my other nephew is a New England fan. So it's, it's like It's a battle for this today. It's gonna be great.

Benito Gonzalez [01:49:48]:
Just wear the one who's winning at the time. Just keep the one that's on.

Lou Maresca [01:49:51]:
There you go. Whoever's leading, I'm just wagging it.

Leo Laporte [01:49:53]:
Yeah, put on both and then strip one off if the other one starts winning.

Larry Magid [01:49:57]:
I don't have a horse in this race, but one of my employees works in, uh, lives in Boston, and I'm going to be meeting— she's coming out for Safer Internet Day, and I want her to be in a good mood. For that reason alone, I'm backing the Patriots.

Leo Laporte [01:50:11]:
Well, I'm gonna— I, I can't root for the Seahawks because they are our mortal enemies, which is why I'm wearing my 49ers shirt today.

Lou Maresca [01:50:18]:
But there's a good reason to root for the Patriots. I hope the, the coach is, you know, wins this, wins, and he's also, uh, you know, does this, uh, was the most winning coach and also the person who's never won the Super Bowl but also won as a coach. So that's, that should be awesome too.

Leo Laporte [01:50:33]:
Yeah. And it's kind of an interesting, uh, story. Unfortunately will mean the Patriots How many Super Bowl victories? 7. It's not— I mean, that makes me mad, but okay, you know. Hey, at least we beat the Eagles. We got the Eagles out of it for you. So there you go. Uh, I don't know, we don't— I don't understand sport ball, so I just, uh, I'm pretending.

Leo Laporte [01:50:56]:
I fake it. I fake it well. Larry Magid is also here for internet— Safer Internet Day, celebrating this Tuesday. week. And from earthquake-prone Oaxaca. See, not exactly earthquake-prone, but earthquake— Not.

Mike Elgan [01:51:11]:
Really, but today it is. And we're gonna be going, we're gonna be meeting up with some friends and watching the Super Bowl in a pub here in Oaxaca. It's actually surprisingly popular.

Leo Laporte [01:51:22]:
Yeah.

Mike Elgan [01:51:23]:
American football.

Leo Laporte [01:51:23]:
Any excuse to drink more pulque.

Mike Elgan [01:51:26]:
That's really the thing that I like about the Super Bowl best of all. I love drinking beer and eating greasy food.

Leo Laporte [01:51:33]:
I've got queso cooking in my slow cooker downstairs. It's Martha Stewart's queso recipe. I don't know how authentic it is, but I'll find out. I put extra jalapeños in just to make it good. Tasty. Our show today brought to you by Zscaler. You need to know about Zscaler, the world's largest cloud security platform. We've been talking about the potential rewards of AI for any business.

Leo Laporte [01:52:00]:
Is, you know, too significant to ignore. But there are also risks and all kinds of risks. The loss of sensitive data, attacks against enterprise-managed AI. And of course, generative AI really increases the opportunities for threat actors, helping them to rapidly create phishing lures, to write malicious code, to automate data extraction, as we just were talking about earlier, to find zero days in code you're running. In fact, if you think about it, it's like, Why am I using AI? But you have to, right? You just have to use it intelligently, carefully. You need Zscaler to protect you. There were 1.3 million instances of Social Security numbers leaked through SaaS AI applications. ChatGPT and Microsoft Copilot saw nearly 3.2 million data violations.

Leo Laporte [01:52:49]:
It's time to rethink your organization's safe use of public and private AI. Chad Pallett is the acting CISO at BioIVT. He says Zscaler helped them reduce their cyber premiums, reduce their cyber premiums by 50%, and double their coverage, plus improving their controls. Take a look at this from Chad. With Zscaler, as long as you've got internet, you're good to go. A big part of the reason that we moved to a consolidated solution away from SD-WAN and VPN in is to eliminate that lateral opportunity that people had and that opportunity for misdirection or open access to the network. It also was an opportunity for us to maintain and provide our remote users with a cafe-style environment. Thanks, Chad.

Leo Laporte [01:53:38]:
With Zscaler Zero Trust plus AI, you can safely adopt generative AI and private AI to boost productivity across the business. Their Zero Trust architecture plus AI helps you reduce the risks of AI-related data loss to greater guarantee the productivity and compliance of your company. Learn more at zscaler.com/security. That's zscaler.com/security. We thank them so much for their support of This Week in Tech. A lot of people use Windows, uh, and use a program called Notepad++. It's a really nice third-party kind of improvement on Notepad. Well, it was kind of a problem this week.

Leo Laporte [01:54:25]:
It turns out the Chinese— it looks like Chinese state-sponsored hackers did a supply chain attack on notepad++.org and that people were downloading Notepad++ from that site. Certain— not everybody. This was a very clever hack. Certain targeted users were selectively redirected to attacker-controlled malicious updates. So they got a version of Notepad++ that had malware in it. According to the former— now former— hosting provider, the shared hosting was compromised through September of 2025. Even after losing server access, attackers maintained credentials to internal services until December 2025, which allowed them to— so, so between September and December, September, October, November, for 3 months, they were able to redirect Notepad++ update traffic to malicious servers. But they weren't trying to attack everybody who uses Notepad++.

Leo Laporte [01:55:35]:
They were looking, it looks like, specifically for overseas Chinese, Chinese dissidents, It was all fixed after December 2nd, but this should scare people. The author of Notepad++ says, I deeply apologize to all users affected by this hijacking. I recommend downloading version 8.9.1, which fixes, you know, has the relevant security enhancement, and running the installer to update Notepad++ manually to make sure you were not infected.

Lou Maresca [01:56:10]:
It's too bad. I loved, I loved Dumpass++, but there's been a bunch of exploits in the last couple years and I just kind of stepped away from it and, uh, you know, went to VS Code.

Leo Laporte [01:56:20]:
So it's one of those— well, VS Code's so great. It's one of those programs though that he was updating a lot, which on the one hand is good, on the other hand not necessarily so good. And in this case, for 3 months, people potentially were a bit Just a word.

Mike Elgan [01:56:36]:
It's part of a larger story of the Chinese government basically considers it a global suppression, global sort of censorship, global control of the Chinese diaspora around the world.

Leo Laporte [01:56:52]:
Exactly.

Mike Elgan [01:56:53]:
So they're not meddling in everything, but they're all over the people who have left China and now live in other countries. There was a recent article, somewhat typical for the last few years, of somebody who was going to go to a film screening of films that were Chinese films in New York City. And he's an American of Chinese descent, and there was some criticism of the Chinese Communist Party and so on. And he got a call from his father who was in a panic saying, don't do anything stupid, don't do this or that, because the Chinese use the family within China to control the actions and the free speech of Chinese Americans and others around the world. So this is a very, very, uh, much bigger story, and cyber attacks is a big part of it, actually. And, um, a lot of people don't care much about it it because doesn't—.

Leo Laporte [01:57:41]:
And supply chain cyber attacks, diaspora. Yeah, the supply chain attacks are very scary.

Mike Elgan [01:57:46]:
It's very sophisticated. I mean, obviously. Yeah. Yes.

Leo Laporte [01:57:51]:
New York's— according to Adafruit, New York wants to Ctrl+Alt+Delete your 3D printer. Uh, this is a bill that is not yet passed, but it's in the budget bill. Language that should alarm every maker, educator, and small manufacturer in New York State buried in Part C says Adafruit. A provision requiring all 3D printers sold or delivered in New York should include blocking technology, software or firmware that scans every print file through a firearms blueprint detection algorithm. They don't want people to 3D print guns, which I guess I understand, but it refuses to print anything it flags as potential firearm or firearm component. And if you have ever used an AI and been told, I can't do that, that, that's going to violate my, uh, my principles, you realize that this kind of software can be wrong in a lot lot of, a of ways.

Mike Elgan [01:58:44]:
Uh, but I see this is akin to software within drone, uh, applications, drone apps control DJI drones that prevent you from flying over sensitive military bases and, and other things like that. I mean, uh, if something can be dangerous and if, if that danger can be mitigated through software, I understand the, the reasoning for it. Um, I'm not Adafruit super— says free.

Leo Laporte [01:59:07]:
Speech issue, or Adafruit says one of the problems is you cannot reliably detect firearms from geometry alone, which is how this would have to work. So if it's L-shaped, if it has a pipe, a tube, a block, a bracket, a gear. Is it a gun? Is it a part of a gun? Um, it would affect all open-source firmware— Marlin, Klipper, RepRap. It would affect offline machines, printers that never touch the internet. Uh, it will affect file formats the algorithm can't parse, raw G-code, custom slicers, parametric designs. It affect CNC mills. So it would really be a problem for 3D printers. Just bring it to your attention, is not yet a law.

Leo Laporte [01:59:50]:
It is in the budget bill, which means, you know, unless it's stripped out, it's likely to pass. Western Digital is planning 140 terabyte hard drives. What could possibly go wrong? That's the wrong— that's the wrong story. I got the wrong story on there. Are you ready for 140 terabyte hard drive? How— you'll definitely need Spinrite.

Benito Gonzalez [02:00:20]:
Hey, those are raw Blu-ray files are 80 gigs, so you know.

Leo Laporte [02:00:27]:
It's a 14-platter, 3.5-inch. They call it the Hammer HDD design. You know, it blows me away how long-lived this spinning drive technology is. I thought by 2000 we would have solid-state memory. There's no way spinning drives would survive past 2000. Not only have they survived, we're going beyond 140 terabytes in the near future.

Benito Gonzalez [02:00:58]:
Also, AI is making SSDs more expensive, so we're going to stay here for a while.

Leo Laporte [02:01:02]:
Yeah, well, I mean, and, and really, yeah, what do you back? Well, you don't back up, you back up to it, right? It's— that's what goes in your NAS, I guess. Although I have a 30 terabyte NAS, I don't need need to 140 terabytes.

Benito Gonzalez [02:01:14]:
It's all media, it's all media, it's media hoarders, you know.

Lou Maresca [02:01:18]:
Yeah.

Leo Laporte [02:01:18]:
Get ready for the Wordle crisis. The New York Times has announced that, uh, well, it started February 2nd, so They started this week, they're gonna start reusing words. There's no reason they need to. Patrick Wordle, who designed Wordle, came up with nearly half a million 5-letter words that could be used in Wordle. He narrowed it down to 22,949 by things like excluding proper nouns and plurals and anything with uppercase level, uh, he got the word list down to 5,437 words. So that sounds like it's short. Well, a word a day from that list would give you 15 years of Wordle. Wordle started October 2021, so there are enough words in there, good words, to go through 2036.

Leo Laporte [02:02:18]:
Patrick says, why Are you repeating New York Times? Why? No one knows, but I just thought I'd warn you if you see a word and you say, I wasn't— this is a wordle word before.

Lou Maresca [02:02:37]:
Gotta wonder how long they're gonna milk that after they spent $10 million on it.

Leo Laporte [02:02:41]:
It has saved the New York Times. Look what happened to the Washington Post this week. I mean, it's been decimated. Uh, The New York Times is the last national newspaper standing.

Larry Magid [02:02:51]:
And, well, Wall Street Journal.

Leo Laporte [02:02:54]:
Oh, I guess, yeah, The Wall Street Journal too.

Larry Magid [02:02:56]:
Yeah, but, but, but if Wall Street Journal, a little bit of a niche market.

Leo Laporte [02:03:00]:
Yeah, because you pay money for The Wall Street Journal to make money, right? You're going to get financial information that helps you make money. The New York Times, you pay for it for the crossword puzzle and Wordle.

Benito Gonzalez [02:03:10]:
Yeah, The New York Times, The New.

Leo Laporte [02:03:12]:
York Times, and the recipes and the recipes.

Larry Magid [02:03:15]:
Some good reporters.

Leo Laporte [02:03:16]:
No, the New York— yeah, I agree.

Larry Magid [02:03:18]:
Yeah.

Leo Laporte [02:03:18]:
And there will be more good reporters because I imagine a number of people from the Washington Post— 300 people fired, including, by the way, their excellent tech reporter Jeffrey Fowler.

Larry Magid [02:03:27]:
And Joseph Mann as well.

Leo Laporte [02:03:28]:
And Joseph Mann.

Larry Magid [02:03:29]:
Yeah. Former LA Times colleague of mine.

Leo Laporte [02:03:32]:
They're not going to do any more tech. They're not going to do any more sports. They're not going to do any more international coverage. So there was a great piece Was it The Atlantic that said you could buy a yacht or you could buy a newspaper?

Larry Magid [02:03:46]:
Actually, the newspaper was half the price of the yacht.

Leo Laporte [02:03:48]:
It was half the price of the yacht.

Larry Magid [02:03:50]:
Two newspapers for the cost of one yacht.

Benito Gonzalez [02:03:52]:
The newspaper costs as as the much super yacht.

Larry Magid [02:03:54]:
Apparently he could run it for years.

Leo Laporte [02:03:57]:
So why? I mean, Bezos has more money than— he could run this thing at a complete deficit forever.

Larry Magid [02:04:05]:
Right.

Leo Laporte [02:04:05]:
What's going on? Is it Well, we know what's going on.

Mike Elgan [02:04:10]:
Some men just like to watch the world burn, I guess.

Leo Laporte [02:04:13]:
Democracy dies in the darkness of Bezos' pockets.

Larry Magid [02:04:18]:
It's gotten dark.

Leo Laporte [02:04:22]:
Here's a picture of Jeff. Did he ever buy a yacht? He just bought a newspaper.

Larry Magid [02:04:28]:
Doesn't he have a half a billion dollar yacht?

Leo Laporte [02:04:30]:
Yeah, I think he has a yacht.

Benito Gonzalez [02:04:31]:
Yeah, no, his— the support yacht for his yacht costs more than the Washington Post.

Leo Laporte [02:04:36]:
Than the Wall Street Journal.

Mike Elgan [02:04:37]:
His yacht is so it big has its own its own yacht. That's how big this yacht is.

Leo Laporte [02:04:41]:
It has to, because, you know, I.

Mike Elgan [02:04:44]:
Mean, he could afford all the yachts and all the newspapers. I mean, it's not— he doesn't have to really make a choice.

Larry Magid [02:04:53]:
I cheered when he bought— I mean, I was happy that he— we thought.

Leo Laporte [02:04:55]:
It would be a good thing, didn't.

Larry Magid [02:04:56]:
We, for a while? Yeah.

Mike Elgan [02:04:58]:
Yeah.

Leo Laporte [02:04:58]:
Um, incidentally, the CEO that came in, I think probably there's reason to believe to dismantle the Washington Post, is gone. He left, and now the former CEO of Tumblr is in charge. Okay, okay. Tumblr, which is owned by Automattic— actually, Automattic's done something pretty good. They have a new plugin that they've designed in conjunction with the Wayback Machine, the Internet Archive, to fix the internet's broken links problem. Uh, the two organizations, Automattic and the Internet Archive, have launched a new WordPress plugin called the Link Fixer that is designed to combat the scourge of link rot. Um, Automattic says the new plugin works by scanning your WordPress posts for outbound links, cross-referencing the Wayback Machine for archive versions of those links. If there are none, it will automatically take new snapshots of the articles in question, put them on the archive, so your blog will always link to either the original live site, or if that site has disappeared, the Internet Archive.

Leo Laporte [02:06:18]:
Don't you think that's a great idea?

Larry Magid [02:06:22]:
Although you— the problem, the problem is you're going to get old information if the site literally disappeared, but at least I guess that's— hey, at least you get.

Leo Laporte [02:06:28]:
You.

Larry Magid [02:06:28]:
Something.

Leo Laporte [02:06:28]:
Know, right?

Larry Magid [02:06:29]:
You get the Washington Post. And I wonder if there'll be a paywall on the archive version.

Leo Laporte [02:06:37]:
God bless the Internet Archive. I, you know, there isn't a paywall in an archive, but I donate every month because that is the most— one of the— that and Wikipedia, which I also donate every month to, are really the things that makes the internet a great thing.

Larry Magid [02:06:50]:
What about Craigslist? Is that still a great thing?

Leo Laporte [02:06:53]:
Yeah, but they're making money. So it's okay, they don't need my money. Yeah, Craigslist is great. Do you think Craigslist made the internet great though?

Larry Magid [02:07:02]:
It didn't do very— it didn't help newspapers, but it certainly, uh, changed the classified advertising business.

Leo Laporte [02:07:09]:
Somebody is now stealing my car, I see on our cameras. I'm hoping it's my wife.

Larry Magid [02:07:15]:
I remember when I used to be a landlord, I used to have a rental house, and Craigslist was a godsend.

Leo Laporte [02:07:20]:
For— yeah, well, Craig Newmark, we love Craig Newmark. He's appeared on our shows many times. He's a fan of Intelligent Machines, and he's taken the money he made from Craigslist and donated it to many, many good causes. He has, including saving the pigeons. But that's not the best of his causes, but it's a good one.

Benito Gonzalez [02:07:37]:
And Craigslist is amazing, and it's probably helped out more, like, non-tech people than you think. Like, it's helped out so many people. Craigslist It's amazing.

Leo Laporte [02:07:47]:
What is your number one use for Craigslist, Benito?

Benito Gonzalez [02:07:49]:
Finding an apartment in San Francisco.

Leo Laporte [02:07:51]:
Yeah, yeah. Have you ever used it for dates?

Benito Gonzalez [02:07:55]:
No, but I also use it for buying and selling musical gear. That's like the best place. In some— it it depends, depends where you live. Also in San Francisco, it's the best.

Leo Laporte [02:08:05]:
Benito, do you have at hand the Craig Newmark jingle?

Benito Gonzalez [02:08:09]:
I don't. I tried, but I don't.

Leo Laporte [02:08:11]:
Yeah, we have have a, we a jingle whenever we mention his name on Intelligent Machines, uh, The singers come on and sing his name. Uh, Project Hail Mary is getting its own Lego set just in time for the movie. This is a great book by Andy Weir. Love the book, love Andy Weir, love The Martian. Uh, if you haven't read Project Hail Mary, read it now before the movie comes out. The movie will not only spoil it, it won't be as good, I guarantee you. And it might be a great movie, I'm not saying that. The Martian was a great movie.

Leo Laporte [02:08:42]:
But it's always better, in my opinion, at least with Andy Weir, to read the book first. Project Hail Mary comes out March 20th. The new LEGO set is an 830-piece set that is a replica of the Hail Mary spaceship. Oh, it's got a spoiler. Don't look. Oh, this is why you got to read the book now, because it'll be spoiled for you it's— by— if— don't even look at the Hail Mary trailer, which spoils it.

Mike Elgan [02:09:09]:
Well, the movie, the movie trailer. Yeah, the trailer spoils it. I mean, they literally show— they show.

Leo Laporte [02:09:14]:
The thing that's a surprise. We don't want to show the thing. And apparently I just showed it. I hope you didn't see it. I took it down right away. So does the Lego set. So read— do— if you haven't read Project Hail Mary, do yourself a favor, read it.

Benito Gonzalez [02:09:29]:
It's only a spoiler if you read the book though. Like, if you didn't read the book, you kind of need that information, right?

Leo Laporte [02:09:35]:
Well, read the book, get the spoiler. Now you can see all the other stuff. We don't let them. Yeah, so don't let a movie trailer spoil a great book. Yeah, or a Lego set.

Mike Elgan [02:09:45]:
In order, read the book, watch the movie, and get the Lego set.

Leo Laporte [02:09:51]:
In that order. Um, didn't I have this McDonald's story in here? McDonald's is pissed off because people are using McDonald's, stuff as their passwords. And McDonald's says you should never use any of our stuff as a password. I guess I took it out.

Benito Gonzalez [02:10:13]:
What do you mean by stuff?

Larry Magid [02:10:14]:
What do you mean?

Leo Laporte [02:10:15]:
Like, uh, Hamburglar. A lot of people use Hamburglar.

Larry Magid [02:10:18]:
Hamburglar, that's a great password. Yeah.

Leo Laporte [02:10:23]:
They say stop it, knock it off. Oh, I can't, I can't find it. Shoot, I took it out. I thought it was one of the best stories of the week.

Mike Elgan [02:10:35]:
Yeah, Mayor McCheese. That's a great password.

Leo Laporte [02:10:38]:
Yeah, Mayor McCheese. Don't, don't use McDonald's. McDonald's tells customers— here's the story— McDonald's is not loving your Big Mac, Happy Meal, and McNuggets passwords. I guess Last weekend was change your password day. Didn't know that. Little, uh, public service announcement: you do not need to change your password unless it's been compromised, and any good password manager, including our sponsor Bitwarden, will tell you if it's been compromised. McDonald's Netherlands took the opportunity last week to tell customers, when it comes to choosing a password that's easy to remember, do not pick the names of our products. According to Have I Been Pwned, McDonald's says Big Mac and its leetspeak variants were found more than 110,000 times in the compromised password corpus.

Leo Laporte [02:11:32]:
Also, Happy Meal, McNuggets, and french fries also common. And it doesn't make it any safer to use the leetspeak, you know, 1 for L. That that doesn't, doesn't help.

Benito Gonzalez [02:11:45]:
You don't get to claim french fries, McDonald's.

Leo Laporte [02:11:49]:
They actually— McDonald's made an ad. This is from McDonald's Netherlands. I don't think we'll get taken down. Why is it in Dutch though? I don't know. This was, uh, the ads placed in Dutch subway stations and other spaces saying, do not use Chicken McNuggets as a great password. You know what, I don't— I know, I just— obviously it's an ad for McDonald's. McDonald's.

Benito Gonzalez [02:12:15]:
Right?

Larry Magid [02:12:15]:
But, uh, it gets you thinking about Chicken McNuggets.

Leo Laporte [02:12:18]:
It got you. Yeah, it's got you thinking about it. I think that's pretty, pretty funny. February 1st was Change Your Password Day. Yeah, update your password. Don't use Chicken McNuggets with leet speak as your password.

Lou Maresca [02:12:32]:
It's like a bad idea to have like, a, a password day, change your password day. Like, that means that— yeah, a great day for people to watch watch your, your traffic, right?

Leo Laporte [02:12:41]:
Yeah. Do not change your password unless you think it's been breached. If you have a really good long password, keep it. The problem is when people change their passwords, they tend to change them for something they can remember, right? And that's not an improvement.

Larry Magid [02:12:54]:
I have one that I can remember, but they're, they're very— it's an algorithm in my head which helps me remember.

Leo Laporte [02:13:00]:
I, I do song lyrics, things like that.

Larry Magid [02:13:04]:
No, your favorite song break into your account?

Leo Laporte [02:13:07]:
I'm more clever than that.

Larry Magid [02:13:09]:
Okay.

Leo Laporte [02:13:10]:
Poems. I like poems. Yeah, anything that I could say in my mind. Uh, you know, uh, another one that I, I don't use but people recommend is if you do the first letter of the last name of every of president the United States, capitalizing the Republicans. Uh, you know, you want— what you want is a long password that you can create with an algorithm. Them, right? But you wouldn't do all the presidents. As many as you remember.

Larry Magid [02:13:38]:
How about just your favorite presidents?

Leo Laporte [02:13:40]:
Your favorites. It would be best if it was something that you could actually— that's why I use song lyrics or poems, because I can say them in my head.

Mike Elgan [02:13:48]:
And go— half your passwords begin with roses are red.

Leo Laporte [02:13:53]:
That's not a good one.

Larry Magid [02:13:54]:
No.

Leo Laporte [02:13:56]:
Unfortunately, the password I used for my Bitcoin wallet was not Well, it's.

Larry Magid [02:14:00]:
You.

Leo Laporte [02:14:01]:
Interesting.

Larry Magid [02:14:01]:
Know, I get notifications from Google about all these, of all these, you know, my, your information appears on the dark web. Yeah. And it's always the same password that I retired maybe 20 years ago.

Leo Laporte [02:14:10]:
Yeah. Right.

Larry Magid [02:14:11]:
But I'm still getting notifications on that password. And it was a horrible password, by the way.

Leo Laporte [02:14:15]:
Yeah. Monkey123. I keep getting it.

Larry Magid [02:14:17]:
It was mine. I could do that.

Lou Maresca [02:14:18]:
Yeah.

Leo Laporte [02:14:21]:
Uh, and then finally, and somebody already said this, they said, oh, the, always ends with somebody dying. The only— I'll tell you, the only reason I do this is because, A, I'm, I'm old and I pay attention to obituaries. Kids, this is going to happen to you when you get to a certain age. You start reading the obituaries first. Yeah, just to see.

Mike Elgan [02:14:40]:
And you do the math. You're like, okay, my age minus, you know.

Leo Laporte [02:14:44]:
Yeah, whatever.

Larry Magid [02:14:44]:
You know, it's sort of like when you're a certain age, you go to birthday parties and then bar mitzvahs or and christenings then weddings. And yeah, you know what's next?

Leo Laporte [02:14:51]:
I'm going to a funeral on Friday. So, and when Catherine O'Hara passed, I immediately did the math and she's 2 years older than me. That's— Younger than me. And younger than you.

Larry Magid [02:15:01]:
Yeah.

Leo Laporte [02:15:05]:
Sad to say that one of the— they sometimes call him the grandfather of the internet. David Farber has passed. He died yesterday. Professor of computer science, known for his contributions to computer languages and networking. Um, he was at the RAND Corporation, Scientific Data Systems, at UC Irvine, UC Delaware, Carnegie Mellon. Of course, won all of the computer awards. Founding editor of ICANN Watch, board of advisors of Context Relevant, the liquid information company. One of the founding board members of the Internet Systems Consortium, served on the board since 1994.

Larry Magid [02:15:43]:
And he was 91.

Leo Laporte [02:15:44]:
So, but he lived a good long life and made major contributions.

Larry Magid [02:15:48]:
Boy, what what he— he— yeah, what he saw and accomplished.

Leo Laporte [02:15:51]:
Inducted into the Internet Hall of Fame in 2013. So, uh, Dave uh, Farber, I, I only found out about this because I saw it on Hacker News. People going, oh, Dave Farber passed. But yeah, 91 is— it's okay, it's okay. Uh, ladies and gentlemen, You still have half an hour before the Super Bowl. Good news. Time to get your Seahawks jersey on and then put your Patriots jersey on top of that. And when the Seahawks go ahead, you rip it off, say, see, I was a Seahawks fan all along.

Leo Laporte [02:16:23]:
Lou Maresca, so nice to see you. Thank you so much for being here. Engineering leader at Copilot Microsoft. Uh, one of my favorite people in the world, actually all three of you. Uh, just always love having people on that I admire and respect and love learning from. Thank you so much for being here, Lou. Really appreciate it.

Lou Maresca [02:16:41]:
Yep, appreciate it.

Leo Laporte [02:16:43]:
Same to you, Mike Elgin, suffering through an earthquake. It's taco time.

Mike Elgan [02:16:48]:
I'm not shook up. Yeah, no, it's going to be an English pub, so I think we're going to have fish and chips and— oh, how funny— tons of beer.

Leo Laporte [02:16:54]:
That's one of the things people think Mexicans eat Mexican food.

Larry Magid [02:16:57]:
No.

Leo Laporte [02:16:59]:
They just call it food. And second of all, actually they're quite fast.

Mike Elgan [02:17:04]:
In Oaxaca there's all kinds of food.

Leo Laporte [02:17:05]:
All kinds of food. Yeah.

Mike Elgan [02:17:07]:
They are. Most of the street food here is hot dogs and hamburgers.

Leo Laporte [02:17:10]:
Isn't that fun? Yeah.

Benito Gonzalez [02:17:13]:
Hamburgesas.

Mike Elgan [02:17:13]:
That's right.

Leo Laporte [02:17:14]:
It's great to see you, Mike. Give my love to Amira. If you wanna know about those trips around the world, where are you going next?

Mike Elgan [02:17:22]:
We got 2 Mexico City experiences after the Oaxaca one, and then it's onto our first Tuscany experience. I mentioned this before on the show, this is our 10th anniversary for the gastronomic experiences. And so we're going and doing all kinds of really great stuff. One of them is the Tuscany experience. We're going to have a couple of other new locations that are super exciting. But we love just— we all— we do every single day, all day, is the funnest, life-changing, food-related things, wine-related things you could possibly— How do you keep.

Larry Magid [02:17:53]:
Your weight down to a reasonable level?

Leo Laporte [02:17:55]:
I know, it kills me.

Larry Magid [02:17:56]:
I would love that job, but I believe I'd be 400 pounds if I did it.

Mike Elgan [02:17:59]:
It kills me. It's, you know, I've been trying, I've been working on that lately. I have to figure out how to sort of fast and all that stuff between experiences. But yeah, there's so much food, so much really, really good food on these experiences.

Leo Laporte [02:18:12]:
And I will give you an endorsement because Lisa and I went on, they're small groups, a few couple, a handful of couples, 4, 5, 6 couples. They're great people, often TWiT listeners. And you're going to get the best possible travel experience, really getting to know the area, the local, and the food and wine and beverages. We drank a lot of mezcal when we were in Oaxaca and learned a lot about how it was made. Went to a pulque bar and danced. It was so much fun. It is, if you really want to travel, not be a tourist, but travel, Mike and Amira know the regions and know how to put together an amazing experience, the Gastronomad Experience, gastronomad.net. Mike also has a newsletter at machinesociety.ai that covers AI, and he's very insightful.

Leo Laporte [02:19:03]:
I look forward to the new one, The Attachment Economy. That's going to be great.

Mike Elgan [02:19:09]:
Um, thank you, Leo.

Leo Laporte [02:19:10]:
Don't forget Hello Chatterbox, which was Mike's son's startup that's still going strong. HelloChatterbox.com. It's AI for people who want private AI, who want to understand it. For schools, it's a safe playground for kids to learn about these incredible devices that is— they're for sure going to be a part of their future, and understand it and not then demystify it. Uh, HelloChatterbox.com. And, uh, have I plugged everything? Is there anything left? I like to give you all the plugs.

Mike Elgan [02:19:50]:
That's all of it, Leo. Thank you so much for those plugs. Well, thank you. I really, really appreciate it. And the one that I really want everybody to sign up for is Machine Society because all the other things that I'm involved with have links and information in that newsletter. So it's free.

Leo Laporte [02:20:09]:
Awesome.

Mike Elgan [02:20:10]:
Can be free. There's a paid version. But thank you for all those plugs, Leo. I really appreciate it.

Leo Laporte [02:20:14]:
My pleasure. I forgot to ask you, Lou, does, uh, does— did Paul— Paul Allen owned the Seahawks, right?

Lou Maresca [02:20:19]:
That was his— yeah, he did.

Larry Magid [02:20:21]:
He did. Yeah.

Leo Laporte [02:20:21]:
Does his estate still own it, or did they sell that off?

Lou Maresca [02:20:24]:
I don't know.

Leo Laporte [02:20:25]:
I don't know either. Yeah, his sister might own it. His sister runs the estate now. Um, so now see, I am torn. Geographically, I would root for the Patriots, and geographically, I'm supposed to hate the Seahawks, but on the other hand, I like Paul Allen. He used to own, uh, TechTV. Larry Magid, it's, uh, it's coming up, Safer Internet Day, and the connectsafely.org is going to have a bunch of events. You can watch the streams on Tuesday.

Leo Laporte [02:20:54]:
You're going to do an event in Sacramento in person.

Larry Magid [02:20:57]:
Yeah, we're going to have lots of kids, lots of legislators, lots of tech executives. Put the tech executives and the kids and the legislators at one table and they can have it out and talk about all the things they want to see change and see what happens.

Leo Laporte [02:21:10]:
I'm gonna have to watch that because, uh, this is a debate I'm having internally. I mean, and we have on many of the shows. I mean, I don't know what the right thing is.

Larry Magid [02:21:18]:
So, well, it's hard to know. I this mean, is— the internet is, as we've talked about on the show, changing dramatically as we speak. And things that, you know, I've had to evolve some of my views in, in various ways because things have changed and there are new risks. You know, I've written— I've been writing about the internet since 1984. I wrote a book called The Electronic Link in 1984, and please forgive me for not having anything in there about nation-states using the internet to control other nations' elections.

Leo Laporte [02:21:47]:
We didn't know.

Larry Magid [02:21:48]:
There's so much I didn't know. I feel like I have to atone, frankly, for No, you having— don't.

Leo Laporte [02:21:52]:
No one knew. Everybody— I mean, I was around then talking about it too, and we all thought this is going to be the greatest thing ever.

Larry Magid [02:21:59]:
Yeah, and it is in many ways. I mean, and we were wrong.

Leo Laporte [02:22:03]:
It's just we didn't know what hazards would also come with it.

Larry Magid [02:22:06]:
We couldn't anticipate some of the risks, but then we couldn't anticipate some of the benefits as well. And I'm still bullish. I mean, believe it or not, despite a lot of things, I'm still optimistic and I'm still— every day I log on and do productive things that I couldn't have done for the first 50 years of my life. And, you know, and I'm happy about that.

Leo Laporte [02:22:25]:
But imagine your life without the internet. It's not— Hard to believe.

Larry Magid [02:22:29]:
Well, it's happened when the internet goes down. When my, my, you know, the other night all of a sudden Netflix started buffering and I had to reboot my modem, my router, and I had no connectivity for 10 whole minutes.

Leo Laporte [02:22:42]:
How did you survive? Uh, there'd be no Twit without the internet, that's for sure. Uh, I would probably be working at some dingy radio station somewhere if it still exists.

Larry Magid [02:22:56]:
You could have been with WCB.

Leo Laporte [02:22:56]:
If it weren't for the internet, newspapers and radio would still be alive.

Larry Magid [02:23:00]:
Yeah, remember WCBS, one of the, the, the, the flagship CBS station? It doesn't exist anymore.

Leo Laporte [02:23:06]:
It doesn't exist. I grew up listening to WCBS anyway.

Larry Magid [02:23:09]:
But it's, uh, thankfully— and by the way, as long as we're plugging our kids, Balkan Bump is my son's trade name.

Leo Laporte [02:23:15]:
Oh, what is Balkan Bump? It's a.

Larry Magid [02:23:17]:
It's.

Leo Laporte [02:23:17]:
Band?

Larry Magid [02:23:17]:
It'S, it's kind of a project, he calls it. It's, it's, um, you go to the website And they're performing all over the world, although he's not performing as much as he used to because he's got a young child now. But looks like he's got some tour dates coming up.

Leo Laporte [02:23:31]:
And he was on the Super Bowl a couple of Super Bowls ago.

Larry Magid [02:23:33]:
On Super Bowl 2020, got him through the pandemic.

Leo Laporte [02:23:35]:
So you, you believe it or not, you've been watching a show with two people with kids and Super Bowl ads. Unless maybe more, if Mike and Lou have been doing anything I didn't know about. I don't know.

Lou Maresca [02:23:47]:
Not yet.

Leo Laporte [02:23:49]:
Um, can I, can I play a little Balkan bump music? Is it Balkan?

Larry Magid [02:23:54]:
It's Balkan-inspired.

Leo Laporte [02:23:57]:
I love, love Balkan music, actually. Yeah, where is the— where is— is there any music?

Larry Magid [02:24:03]:
I think— oh, there is somewhere.

Benito Gonzalez [02:24:05]:
Links are up top. There's a YouTube button. Oh, that works too.

Leo Laporte [02:24:12]:
This is called Prayer song. Oh, I'm going to listen to this. Is there— is there singing or is it all instrumental?

Larry Magid [02:24:21]:
You know, I can't remember on this one. Um, he's got a new album out, so this may be his premise.

Leo Laporte [02:24:25]:
This is the new album? Yeah, I haven't even heard it yet.

Larry Magid [02:24:27]:
So you're hearing it for the first time.

Leo Laporte [02:24:29]:
Wow, nice.

Larry Magid [02:24:30]:
Check it out.

Leo Laporte [02:24:31]:
Yeah, he's— I love it.

Larry Magid [02:24:33]:
And by the way, the funny story— you'll love this— he was kicked off Facebook for a while for violating his own copyright.

Leo Laporte [02:24:42]:
Of course.

Larry Magid [02:24:43]:
Luckily he has a dad who knew, who knows people, so we got him back on right away.

Leo Laporte [02:24:46]:
Of course. How, how unsurprising. Thank you, Larry.

Larry Magid [02:24:53]:
Thank you, Leo.

Leo Laporte [02:24:54]:
We had Larry, Lou, Leo, and Mike on this show. Thank you guys, really appreciate it. A special thanks to our Club Twit members who make this show possible. Lisa just told me this month Club Twit members supported us 33% of our operating expenses. One-third of our operating expenses came from the club. That really tells you something. Without you, we would not be able to do what we do. So thank you.

Leo Laporte [02:25:17]:
And if you're not a member of the club, I really want to encourage you to join twit.tv/clubtwit. You get ad-free versions of all the shows, access to the Club Twit Discord, which is a great hang, all the special programming we do in the club like yesterday's, or I guess it was Friday's, uh, AI user group, which was so interesting. Uh, twit.tv/clubtwit. Please, we'd love to have you in the club. I'd love to see you at Club Twit. Uh, we did— go ahead.

Mike Elgan [02:25:42]:
I'm sorry, I'm sorry, Leo. Can I, can I plug Club Twit just a little bit? Yes, my own, my own plug. I really am bothered by podcasts that give you three-quarters of the podcast and say, if you want to listen to the rest of the podcast, you gotta pay. What you're doing with Club Twit is you're giving everybody all the podcasts, right? And if you join the club, you're supporting it. There's some unique programming and so on, but this is the right way to do it. So one of the reasons to support Club Twit is because you want to reward the podcasters who are monetizing in the right way, not in an exploitative way, not in a sort of, not in a kind of a, a negative way like so many podcasters are doing. And so you should reward the good podcasters.

Leo Laporte [02:26:29]:
I appreciate that. Yeah, we— I never wanted a paywall. I really believe that what we do should be available to everyone for free, and it is free. It's ad-supported, uh, for free. But if you want to support what we do and don't want to hear the ads, uh, join the club. And, uh, and yeah, I think that's a good point, uh, Mike. Even the stuff that we do in the club You can watch as we do it, and then a month after we do it is available to the general public, sometimes sooner. So yeah, that's, that's a— thank you, I appreciate that.

Leo Laporte [02:26:58]:
We— you were the one who told me that what we do is democratizing because anybody can listen. compliment. What a And I think that's always been important. Yeah. Yeah. Thank you, Mike. Appreciate it. Thank you all for being here.

Leo Laporte [02:27:10]:
We do TWiT every Sunday. Normally it's 1400 Pacific time, 17, uh, East Coast time, uh, 2200 UTC. You can watch us live, club member or not. If you're in the club, you can watch in the Discord, but even, even the club members often watch on YouTube or Twitch or X or Facebook or LinkedIn or Kick. We're on all of those live, chatting with you live on all of those. After the fact, on-demand versions of the show, audio or video, available at the website twit.tv. There's a YouTube channel with the video, great way to share share clips if you would, that'd be great. Tell the world about the best technology podcast in the world, in my humble opinion.

Leo Laporte [02:27:49]:
There's also, of course, it's a podcast, so there's the opportunity to subscribe and it's free. I hate the word subscribe because it sounds like you pay for it. No, it's free. Just pick up a podcast client. Apple's is fine. Pocket Casts, Overcast, there's many of them, and subscribe. That way you'll get it automatically, audio or video, the minute we've polished it up. Thanks to our technical editor and producer, Mr.

Leo Laporte [02:28:15]:
Benito Gonzalez. Kevin King does the editing. Thanks to all of you for joining us. We will see you next time. Another TWiT is in the can. Go Patriots! Go Seahawks!