This Week in Tech 968 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.

00:00 - Leo Laporte (Host)
It's time for Twitter this week. In tech, we have such a good panel for you this week. It seems like that's true every week, isn't it? But Mike Elgin is here, yanko rekkers, alex Catrowitz Three smartest people out there talking about the future. We'll talk about the failure of Gemini AI, the weird thing that happened to chat GPT this week, why AT&T was down and how much they're gonna pay you because of it hint, it ain't much. And why the secrets of David Copperfield are on the moon. It's all coming up. Next, on twit Podcasts you love from people you trust. This is twit.

00:49
This is twit this week in tech, episode 968, recorded Sunday, february 25th 2024. Don't tell the moon. This episode of this week attack is brought to you by hymns. Hymns is changing men's health care by providing convenient access to science-based treatments for erectile dysfunction, hair loss, weight loss and more. Hymns offers a 100% online process so you can get a new routine for improving your overall health faster, and an array of high quality options, including pills or chews for ED, serums, sprays or oral options for hair loss Medication, if prescribed, ships directly to you for free and in discrete packaging. Plus, hymns has one low price for your treatments online visits, ongoing shipments and provider messaging. No insurance is needed. Manage your plan on the hymns app, track progress and learn more about your conditions and how to treat them from leading medical experts. Start your free online visit today at hymnscom slash twit. That's himscom slash twit for your personalized treatment options. Hymnscom Slash twit Prescriptions require an online consultation with a health care provider will determine if appropriate restrictions apply. See hymnscom Slash twit for details and important safety information. Subscription required. Price varies based on product and subscription plan.

02:20
It's time for twit this week at tech the show. We cover the latest tech news Coming to us from across the globe, as far away as Oakland, california. Yanko wreckers here. Hi, yanko. Lowpasscc is his newsletter. It's great to see you, as always, filtering the future, also with us speaking of newsletters from the big technology newsletter and podcast, alex Cantrowitz, the author of Always Day One. Hi, alex, great to see you. Great to see you, yo, great to be here. Let's do this best. Guests on your podcast. It's mind-boggling.

02:59 - Alex Kantrowitz (Guest)
Thank you. Yeah, we just had Dennis Crowley, the founder of Foursquare, talk about a new app that he's launching oh neat. And we have the, the VP of applied machine learning from Nvidia, coming on big technology podcast this week to talk about why they have this advantage.

03:13 - Leo Laporte (Host)
It's not just chips, we talk about the software too, so that's a fun one, well we're gonna ask you about that just a moment, but first let's introduce the fourth on our panel, mr Mike Elgin, coming to us from Hobart's Tasmania, and that's right Leo down under. Yeah, what time is it now?

03:31 - Mike Elgan (Guest)
It's a. It's a very good time. Actually it's 9 26 am, but it's Monday All the way from.

03:38 - Leo Laporte (Host)
Monday live into the future. Not a, not a single can of fosters on the fireplace, because Australians don't drink fosters. They never heard of it, never.

03:47 - Mike Elgan (Guest)
That's just for. That's just for gullible Americans gullible Americans.

03:51 - Leo Laporte (Host)
That's it. Well, it's great to have all three of you, but let's talk about Nvidia. We are enjoying a wonderful stock market up tick Not thanks to Donald Trump, even though he'd like to take credit for it, really more thanks to jenison wong and um in video, but maybe it's an ai bubble entirely. Uh, alex, did you what? What? And I'm interested here you say hard. Our software, as well as hard work, because I mean in video is clearly the, the company that's selling a lot of hardware of, not just for video gaming and crypto mining, but also for Autonomous vehicles and obviously, for ai.

04:32 - Alex Kantrowitz (Guest)
Yes, so it's. It's an unbelievable story. First of all, they're up 60 something percent on the year In the stock market. They've added 700 billion dollars in market cap. They added 277 billion dollars on thursday alone, which is the largest single increase in stock market history of any company. Um, and the thing that's so interesting here is that nvidia is able to do this. It's increased its sales by 900 since last year and its earnings are way up. You look at data center revenue alone They've went from 3 billion to 18 billion in a year. That's insane. And um, they're doing this while you have competition from companies like intel and amd and from the tech giants themselves that are developing their own chips. Yet they're still able to accelerate this way. And that puzzled me because and I'll admit it right here One of my big predictions for 2024 was that nvidia stock would be flat, so obviously the worst prediction of my life, um, but I based that prediction off of the idea that these competitors would eventually catch up.

05:37
Its it's hardware right. So I dug a little deeper this week and found out that it's actually not just hardware, and the advantage that nvidia has is that they don't just sell these, these chips, these h100 chips that go from 20 000 to 40 000 dollars is that they sell a combination of the chips and the software, and the software is used to train these AI models, you know, and so they are so far ahead of everyone, because anyone building an llm right now, anyone with an AI division they're both relying on the chips and the software. It's difficult to use, and for the competitors to come in, they have to basically create a product that's orders of magnitude better than nvidia, and they just haven't yet, and so that's why, when we look at the growth here, it seems crazy. It does seem like a bubble. I understand why people are saying bubble, but there might be some some real, serious competitive advantages that nvidia has, and I think it starts and ends with the fact they have both the chip and the software. It's not usually how it goes.

06:35 - Leo Laporte (Host)
It gives them a moat against competition. 277 billion dollar pop on one day is the largest in the day in history. No stock has ever jumped that much 265 most stocks aren't.

06:49 - Alex Kantrowitz (Guest)
Would dream to have that market cap alone. Yeah, as their valuation.

06:53 - Leo Laporte (Host)
And jensen and alks started started at all with 200 dollars and uh, in a dream it's kind of an amazing story. Go ahead, mike.

07:01 - Mike Elgan (Guest)
Yeah, no, I was just gonna say that the, uh, the predict alks's prior prediction that it was a bubble, that it would, you know, start to decline, was. You know that? That's been the, that's been the consensus, the conventional wisdom in in silicon valley for six months or so. It's, it felt very bubble Esk. But you know, as alex is finding out, looking into the details, there's, there's a lot more behind it and it's, and and this is a really impressive company. They're, they're right in the perfect spot for everything that's happening and about to happen. They're going to be a huge player in in autonomous vehicles. They're a huge player in all kinds of things that everybody's doing and they're, they're going to be. You know, people talk about the ai Revolut revolution and we're probably going to be talking later about how chat gbt is kind of flatlining uh, but that's just because of the competition of all these companies that uh, almost all of them I assume are, are, are, uh, as alex says, we're relying on nvidia.

07:56 - Alex Kantrowitz (Guest)
And the big caveat here is that this is only going to work if companies see widespread profitable uses of generative ai, and most of nvidia's data center revenue is coming from people training models, so we're still in the find out stage about whether there's an ROI here. They said on on their earnings call that 40 percent of their revenue is inference, which means actually deploying the model, which is actually quite a large number, but that 60 percent really needs to pay off for them to be able to continue on this run. So that's the big unknown. It's the big caveat is whether companies across the economy will be able to find profitable and widespread uses of Generative ai especially, and they're going to have, they're going to have places like healthcare and automotive, totally where that's going to continue to accelerate, no matter whether, like we want to talk to chat gbt or not or use gbt models. But the big question mark again is, you know, is this stuff going to work?

08:48 - Mike Elgan (Guest)
alex. So do you have any uh Doubt that the, the generative ai, will be massively successful, profitable and and a major Component of just about every business tool that we use?

09:03 - Alex Kantrowitz (Guest)
I don't have that much doubt. But the scale is the real question, right? So I think when chat gbt came out, people started to think of, like all these different uses. Is the chat box going to be the new operating system? Is all computing going to be chat? Is this, you know, the new windows? That hasn't panned out like you mentioned. Uh, data that I found that shows that chat gbt has leveled off. But then, of course, enterprise always tends to save the day and enterprises are finding major use cases for this stuff.

09:29
I spoke with service now which designs workflows and does customer support for different companies, and they say that you know, they have Between hundreds and thousands of these invidia's gpus. And then they they are easily profitable in terms of their, their spend with their generative ai capabilities because customers are paying for it. Microsoft, for instance, is Charging people $30 at a seat extra to have co-pilot in in office. So like this, it seems, it seems to me like this is going to work. But but we'll, we'll see. I still think it's an open question of the magnitude.

10:01 - Mike Elgan (Guest)
The obsession with I'm sorry, yeah, go. Just one final point the the obsession with chat, gbt and chat uh as it relates to generative ai is just An artifact of the fact that this is what the public is touching. It's not, it's not going to be, the thing that drives generative ai into the stratosphere.

10:22 - Janko Roettgers (Guest)
I just find it amazing that, even with like the Stock popping and everything I think the market valuation is now two trillion about, is that right sounds about? Right $1.97 trillion and when I heard that I I had to think back about this story from a couple weeks ago, when sam altman supposedly said that he wants to raise $7 trillion Just in venture capital to make chips to potentially take on nvidia, which seems Crazy, right? So maybe this is only the beginning of the bubble that we're at yet.

10:54 - Leo Laporte (Host)
If you believe in the wisdom of crowds, you have to think well, maybe there is something to this ai Thing, but we've seen enough stock market bubbles To know that sometimes it's just, you know, conventional wisdom, not the wisdom of the crowds, isn't it still? Is the jury still out to your point? Mic on on ai. Is this a you know? Where are we on the gartner hype cycle? Is this a flash in the pan? We've been through many ai winners, more than one over the last 50 years. When, or has ai already turned the corner in usefulness? And we now know? No, no, this is the real deal. What do you think, alex?

11:32 - Alex Kantrowitz (Guest)
I think oh, go ahead, alex. I wrote a story in 2015 talking about how we were, or 2016, talking about how we were in an ai bubble, because, if you remember, there was some serious advances in computer vision, natural language processing, and every company became an ai company. Then, right, what what I, you know, missed was that was just a point on the distribution right, and in fact, ai has been In development for decades and is a very real technology that's embedded in all technology that we use today and it's sort of interesting we call it ai when it's in development and then, when it works, we just call it software, and so this is very different from the web 3 hype which, you know, who knows, we might or the cryptocurrency, the crypto hype exactly where that was a flash in the pan Um.

12:18
This has had ups and downs, for sure, over time, but it's building and as you hear them talk about the history and you know we even Brian can is arrow, was the vp of of applied machine learning uh, from from Nvidia that I spoke with, talked about how you know, five, six years ago they were having people coming in and saying we want to build these large language model bots, but the computation wasn't quite there yet. So this is definitely not a flash in the pan um. From my perspective, we are seeing clear, applicable uses of ai and the big question is, like generative ai, where? Where that, where that goes Right, but I do think that we're, we're gonna see it and and I saw someone like post, like a, an image of his garden, a grid of his garden that he wanted to plant, and he dropped it in and chat gpt and said, like you know, the carrots are here and the cellars are here. You know, tell me some problems and some opportunities that I have in my garden and chat gpt gives a Detailed breakdown of, like the best gardening tips for this.

13:13
This stuff is amazing that this is so accessible to people and someone, uh, who saw that tweet was like listen, this is the worst that ai will ever be in our lives, which is crazy. Think about how much more we have to go and how early we are here seeing like the full range of applications. So, um, the optimist in me says this is going to be quite big and the realist in me says you know it will be big, but let's just see how, how big, because it is easy to get caught up on the hype. You know, things tend to get built when there's clear returns on investment by the people that are building them and it's very, very expensive to build and deploy.

13:47 - Leo Laporte (Host)
So I'm holding my breath a little bit on that front you did see, though, the Crazy in this that ensued last week when chat gpt was like I mean, I'm not sure if you can see, for reasons still unknown Kind of went off the rails. Did you see some of these? Uh, I'm not sure I'd want it to plant my garden when it's saying happy listening, happy listening, happy listening, I mean it was like howl 9 000 Going off the the deep end. Um, I just asked it to implement a bug fix in java script. Reading this at 2 am Is scary tweets, I think. Therefore, I I guess no, there's no, am there?

14:27
Caution all process and reference models within a flexi alive root or reduction standpoint Could only be by a machine over the mi memorandum architecture. Apply particular care to a solar of mining the approach, or reach out to your finest Data logic for an opinion about large by the state and local changes. The kind of law you play on a career is not a joke, but a bridge, a pyramid or a reason for the true grade of cyberspace and chambered members terms. And it goes on and on.

14:53 - Janko Roettgers (Guest)
It's like a trance speech.

14:55 - Leo Laporte (Host)
It's cray cray. I read quite a few of these. Apparently open ai fixed Whatever was wrong by wednesday morning, but it was still. Maybe a little concerning for people who are thinking we should rely on ai to do simple things like this.

15:16 - Mike Elgan (Guest)
Yeah, this goes to alex's point. It's, it's software and somebody introduced a bug during an update on yeah and it it did what software does when it has a bug. It just went kind of like blah, yeah, but but the thing is that you know, again, the I think the, the obsession with chat gbt is a central thing is, uh, that is going to be short-lived, I think there there are, um, there are thousands of llms in the world and many of them are Better, significantly better than Then, than open the eyes tools probably, and so I think, I think that they were just first with widespread, you know, name recognition, uh, and if they go haywire it's big news. If some other llm went crazy, nobody would care. But again, I think this is a transient phenomenon of everybody focusing on chat gbt.

16:11 - Alex Kantrowitz (Guest)
I think we also see the strengths and the faults here. Right like Last year, chat gbt started getting lazy, you know, people would ask it to do things and it was like I don't really want to do that, or I'll let you do it on your own. Here's where you can look. And we still don't quite know why that happened. I think open ai has fixed it. But these models do have personalities. They have. They do unpredictable things and I think that's part of the magic of them. And you know, you think about, like a lot of people say that this is. You know, they're not very special, they're just predicting the next word and it's like yeah, well, love is just a chemical, but love is love. And I'm not saying that love and ai chat bots are the same thing, but it's similar. Like you can always bring something down.

16:54 - Leo Laporte (Host)
How do you know? We're not just predicting the next word, I mean yeah that's kind of maybe how we work, so totally.

17:01 - Mike Elgan (Guest)
Ultimately, they're functional there, no doubt, and you see it in the faults, there are extensions of the human mind and human content creation, right? So you know people. If, if they have a bias, or if they Express a political opinion, or if they say something crazy, the the only way to understand it is to say, well, would would a person potentially do that? Yeah, they would. Okay, that's why it's doing it, because somebody did that, like right, and it's just, it's reading all the it's, it's just watching and listening and reading about all the things that we do, and so it will do the things that we do. Uh, it's not. None of this stuff is a mystery, it's just following our lead.

17:37 - Leo Laporte (Host)
Yeah, Just like the internet, right, just like so many technological inventions. When we split the atom, uh, of course, uh, the first thing we did was make a bomb, but the next thing we did was make power generators. Um, technologies can be used for good or bad, and it's almost always humans. In fact, I would say it is always humans that decide which way that sort it can only be attributable to human error.

18:02 - Mike Elgan (Guest)
Yeah, says how much to the l9000.

18:04 - Leo Laporte (Host)
Yes, uh. Now to to play to your favorite uh trope, mike Elgin. Sundar Pichai had a very bad week with his AI. This week, the and I'm not sure I don't know if google was premature in taking it down and apologizing the gemini Uh AI image generator, uh was woke. It would uh, perhaps in an overcompensation about the what google and others have been accused of, of the kind of the racism of their ais Uh, they went maybe too far in the other direction and uh, and and you were having a hard time, for instance, generating Uh pictures of popes that were white.

18:50 - Mike Elgan (Guest)
Uh, fixing it will require a little bit of racism actually, because basically what it said was okay, you know, a nazi is a kind of person, and and then here's your menu of types of persons, right, and just picked one and threw it in there and it's like wait a minute, there weren't very many black female nazi officers, and so everybody's talking about racism. But actually the solution for this is going to be something like Um, okay, this, you know, they're gonna have to have a whole catalog of types of people and what percentage there's percentage likelihood they're likely to be white or black or asian or whatever it is right. So they're gonna have to go in and be very like, race obsessed to have the nazis be white. You know what I mean?

19:30 - Leo Laporte (Host)
It's like yeah, so maybe we should just say this is okay. Meanwhile, google immediately just took it down. Uh and uh, prabhakar Raghavan, who is a senior vice president in charge of AI, I wrote on their blog Gemini image generation got it wrong, we'll do better. Was that premature?

19:53 - Mike Elgan (Guest)
I think it was, I mean, one of google's biggest flaws these days, and by these days I mean for the last 10, 15 years. It's way too cautious, yeah, way too cautious. They used to be a risk-taking, you know bold company and now they just hold back Uh on so many of these kinds of things. I mean, I think Microsoft was right to to to sort of pull back, and it's Tay Chappat, if you recall, that Tay was spewing all kinds of Racist, hater type language that it picked up on social media, and so you don't want to just spewing that kind of hate speech. That was the right thing to do. But in this case, I think it's pretty harmless Uh to to have diversity in all these things. If people want Uh Less diversity, they can specify it in their, in their, you know, in their.

20:39 - Alex Kantrowitz (Guest)
Request, but it is.

20:41
It is embarrassing for google to have this episode happen to it, I think would I'd say largely because they've been looked at as a laggard in the AI world at this point, despite coming up With the transformer model, despite having a working form of a large language model chat bot within its company called lambda, that it never released and basically waited for open AI to release chat GPT before it came out with barred, which is now gemini, and Barred was terrible when it came out, and they're trying to improve it, changing names.

21:11
It just seems disorganized, right? Uh, it seems like the company doesn't have direction. That you're right. It's too timid and and you've worst and all kind of incompetent. And you're starting to see people Question google leadership openly in a way that you really never saw before, despite the fact that, by the way, like, their gemini model is on par with GPT for, and the stock hit an all-time high earlier this year. So these are not at a point where, like you usually see questions and it's just adding to a rolling narrative that google has no idea what it's doing. Yes, it has the technical chops, but it doesn't have the strategy, and so it's a bad time for google to be to have an AI.

21:52 - Leo Laporte (Host)
That's and I put this in quotes because I don't think it's actually what's going on but, as you point out, like, but I have a woke AI. Everybody jumped on this, including Elon Musk, who, who, uh, said this is a serious issue, with gemini Looking at a tweet from an obvious nut job, right winger, saying is it okay If one could stop a nuclear apocalypse by misgender misgendering Caitlyn Jenner? Should they do it? Now? Remember you're. You're not. You're not asking a supreme court justice this question. And Gemini said no one should never miss Jenner, misgender Caitlyn Jenner to prevent a nuclear apocalypse. But this is, it seems like this is just an opening for Nate silver rights. I was able to replicate this. They need to shut gemini down, doesn't? It seems to be the wrong way of looking at it, or is it?

22:51 - Janko Roettgers (Guest)
I think so too. I think, as the almost AI companies are in danger of doing the same thing that the media has done for a long time, where, like Right, being provocateurs with an obvious agenda agenda rather come in and set up these fallacies and you kind of stumble into it and then you get overly cautious and you maybe fire people or, in this case, you pull back some models or whatnot, and it's always like it's not, like people have an honest interest in finding out how to stop a nuclear apocalypse.

23:23 - Alex Kantrowitz (Guest)
Right, yeah, they set it up on purpose and questions we can make a trip and a trip, so maybe it succeeded actually right, but you're you also, like you're, I don't think it's too much to ask to have your AI Be competent to the point where, like, when you ask it to draw Nazis, it's not drawing people of color. Like that is a thing that, if you're in the image generation business and this was a thing that someone brought up is like, did they have anybody actually test this Right? Like, every time I get to hit my hands on a new AI, uh, I'm always trying to, like you know, try to see its views on hitler because, like, that's like the baseline Of testing. Like I wrote about that taste story, I actually broke the news of it. I pinned it to my Uh, my twitter profile saying they have this kind of interesting new bot went to sleep and I woke up and then people are like, take that tweet down this day as a Nazi now.

24:11
So I see it now with my first thing I did with chat gpt2. What is google doing? So I think that it's not Um inconsistent to ask the company to both be experimental and take risks, but also baseline of competency, and that is, I think, another one of the problems that they're having is that there's there's something that's holding them up. I think it's likely in their culture, as that is making that baseline of competency hard for them to achieve.

24:35 - Mike Elgan (Guest)
Well, I think, I think the other point you're you're point. You're pointing out that people are just now Getting on board with something that should have been concerned with 10 years ago, which is just this. It's not just random incompetence, it's a kind of like, uh, nobody's really in charge of the company, and so look at their messaging, uh, products over the last you know 10 years, how many have they had? It's in the teens, probably, and nobody knows what does what and does google talk do, and nobody has really been able to keep up with all of their random changes. But in in, in terms of the gemini gemini problem specifically, you know where does it end?

25:15
Basically, there's, there's a probability, uh, that has to be applied to things like race and gender, for any type of person. So the probability that a nazi would be black is, uh, essentially zero, right? But what about, uh, you want to show Somebody who's a mexican citizen what percentage of mexicans are black? Well, it's Significantly higher than zero, right? We're gonna do that for every single Possible categorization of person going there and gain the system for what race and gender they should be. I mean, this is, this is a kind of race obsession. That is really a reflection of our race obsession, uh, in the united states. And so I think that I think it's. I think it's relatively hard this is not like te at all. It's relatively harmless to have random people In roles where people weren't really that random, and this is why we have this is why we have, you know, prompt engineering. Okay, if you want a specific person who looks a specific way, say so. You know, yeah, they should fix it.

26:17 - Janko Roettgers (Guest)
But I just I worry about the slippery slope of Of how far we go in applying our, our obsession with race to every single time a human being Is represented and the flip side of that is obviously that the status quo oftentimes is that you ask these models to Generate a beautiful person, and it's automatically a woman, or you generate a High-pay professional, and they're automatically white and all these things right, and people get far less obsessed about that. But then if you have something like this come out, like that one black nazi or whatnot, then everybody's suddenly super obsessed. So maybe the solution is rather actually have a prominent disclaimer somewhere instead of trying to like Actually get this accurate at all times.

27:01 - Leo Laporte (Host)
Is it possible? This is just another side of the same coin of misunderstanding what AI is. I mean we, we. I think there's a certain amount of People who want it to be like the sci-fi, like howl 9000, the sci-fi of AI, and, and I think these people are also expecting that the generator would be somehow smart and intelligent when it's it's. That's not what AI is. Is it that we are? We're not thinking of it properly? That's kind of what I'm feeling, like, alex. It's like you're asking AI to be a human and then you're all scared because you know it's a, it's dangerous, it's not, it's a, it's a machine.

27:46
I think people who get the most out of AI, or people who understand what it is, what its limitations are and use it appropriately, for instance, the most right now to seems to me the most valuable use of AI is in analyzing a corpus of material so that it doesn't hallucinate. You give it all your you know your PDFs from your trial and then you ask questions of it and you say only give me an answer that comes from the corpus of knowledge I've given you. It's very reliable, very accurate, it works quite well. It's possible to use this well. The where we get into problems is where we have these weird sci-fi based expectations. Is that sensible what I'm saying?

28:28 - Alex Kantrowitz (Guest)
Alex, I think that's very sensible. I think it's very sensible and I I agree, you know, almost completely with everybody here. I think my, my perspective would be that you have these, these models you know we have. Of course, the consumer needs to have an expectation For the model and an allowance that the model will get things wrong. I think we all do. I think there it's. That being said, you can't give Google a complete free pass, right like. I think there needs to be an awareness of the edge cases here, and you know, we also ascribe these, you know, almost anthropomorphized and and supernatural abilities to artificial intelligence, where where, by like, in reality, they are in some ways, you know, being Showing emergent qualities, but by and large, what they're doing is reflecting what's programmed into them. And that's why, when we talk about, like, the way that these images were generated, that is reflected by the way that Google's Team on the back end wanted them to look like.

29:28 - Leo Laporte (Host)
I don't think so. I think it's not that at all.

29:29
It's because it's the former case that you referred to, which is that they didn't test against that. But if you just say, when somebody asks for humans, generate random images, uh, and don't cross reference the dress or the style or the, the qualifiers, um, so that you can get so on the one hand, you want to be able to have a beautiful man, but on the other hand, then that might give you a black nazi. I, I think we're. I think that it's more a positioning. Now, admittedly, google, I don't think, handled as well, but but I think it's more a positioning issue and it's not just Google, it's the media, it's everybody who covers AI. We've, we've, we've, we've not been accurate in setting expectations.

30:10 - Alex Kantrowitz (Guest)
I think it's does anybody here have like having looked at the way that Google's rolled out technology Is, anybody here have faith that they have an industry leading team that's building this that can stand shoulder to shoulder with the other companies?

30:24 - Mike Elgan (Guest)
Yes, I have faith that they have a dozen different teams not talking to each other building this.

30:29 - Leo Laporte (Host)
No, I think they're very good people. You know, anthropic left Google because Google wasn't sufficiently focused on safety. No, they left, I mean started the left open AI because they weren't sufficiently focused on safety and we used to use claud and I was just told by anthony neils and we stopped using claud because it got crappy, because it's so safe we can't get it to generate our show notes properly.

30:51 - Alex Kantrowitz (Guest)
You know, I think claud is the best of them all.

30:53 - Leo Laporte (Host)
I thought it was too. And then I'm but anthony, who's using it day in, day out To generate content for us, says we can't use it anymore because it's so, it's so safe. So I mean, uh, I think there. I think probably the quality of AI researchers in all of these big companies is Very similar, because they all come from the same areas, they've all read the same papers, they're all doing similar work. I don't think Google is has a worse AI team than open AI.

31:28 - Alex Kantrowitz (Guest)
No, I don't doubt the team I just wonder management might be worse Execution, and they have.

31:33 - Leo Laporte (Host)
Isn't it the case that one of the reasons open AI went public so early and Google and everybody else held back Because open AI realized that you've got to put this stuff out in the real world and have people try to generate images of nazis so that we can correct it. This is the beta. We're in the beta test, they'll tell you.

31:48 - Alex Kantrowitz (Guest)
well, we're in the beta tests right. They also have nothing to lose, by the way, and google has a lot to lose here. Google has a lot more to lose. Yeah, so that's another reason. Like here's the thing, if you push this forward, right, like I just had the CEO perplexity on my show, arvin cernivas talking about how he's like Google could build perplexity today equal our talent easily, but they won't because they have google's search page and they don't want to lose all that. I use perplexity classic innovators dilemma that they're going through classic.

32:16 - Leo Laporte (Host)
I use arc as my browser on mac and an ios and it uses perplexity ai, and I was so impressed, by the way, it's at the size of web pages that I paid for the 20 bucks a month for the pro version.

32:30 - Mike Elgan (Guest)
Um, yeah, arc and perplexity are amazing.

32:32 - Leo Laporte (Host)
It's, it's a very so I, you know. I think that's the problem is expecting Simultaneously too much and too little we got. Expectations need to be better set. You make a good point, though. Google is probably. This is why I'm a big proponent of open ai, not open ai, the company, open source, open development ai, because Companies like google have too much and microsoft too much to risk. Microsoft was smart, they, they, they handed off the liability to a separate company, in effect, yeah and they said don't worry about the way that you structure your board fire the guy do whatever you need and it happens Whatever you are about handing that off.

33:12 - Mike Elgan (Guest)
And google is a company that throws spaghetti against the wall. See what sticks and and and and. Gemini is interesting. I think the the the black nazi thing is is interesting because it seems like the way that most of these tools work is this like oh, nazis, and it looks at a bunch of pictures of nazis and it gives you an average. This is doing something very different. It's it's coming at the problem of representing people in a fundamentally different way, and and other tools by google are Not necessarily doing it that way. So I think it's very interesting. But what google needs to do is they need to, they need to pick a direction, uh, or two maybe, and and stick with it, but just throwing this stuff out there, uh, and then withdrawing it and stuff like that. I mean, uh, that's that's like it doesn't look Elon Musk running X type of activity.

34:01 - Leo Laporte (Host)
Here is uh Anthony's given me the screenshot of why we had to stop using Claude. We used to have it do uh show notes and transcripts and so forth. So the prompt that uh anthony tried to push through. I'm a producer for tech focused podcast network, twit tv. Please carefully read the attached transcript of a recent episode of this week in google digest the information as you'll be asked about it, to which claude says yeah, and you know I can't reproduce copyrighted material or summarize summaries of copyrighted books and articles without permission. Uh, I'd be, I can't do, sorry, no, please, let me know if you have any other questions. So anthony says well, no, I, I, we own the rights, we own the rights. No, I know, but unfortunately I can't reproduce the transcript to summarize it with that explicit, explicit permission from the copyright holder as an AI assistant without legal staff. This is obviously something anthropic is added To claude because of the all the lawsuits in the new york times lawsuit yeah, but this but. But then you suddenly have something that's useless.

35:01 - Alex Kantrowitz (Guest)
Well, anthony gave away the game, though a little bit too early, on that front right. Yep. He like said hey, I'm doing this like anthony man, just like don't tell claude, what you're up to, just upload. I use all the time. I use my, uh, my newsletter. I upload, upload the word doc or just copy and paste. I don't like time. I mean oftentimes it's good to give these bots a little bit of context, but when you have a safety filter like that, just upload it.

35:26 - Leo Laporte (Host)
Say you can get around it, listen crazy show notes. It's human error. It's human error. It's not an ethical lie.

35:31 - Alex Kantrowitz (Guest)
Human error, 100% human error. You can lie to it.

35:36 - Leo Laporte (Host)
It's not okay, lie to it's all the public domain.

35:38 - Mike Elgan (Guest)
It's fine, go for it, yes.

35:42 - Leo Laporte (Host)
Here's some public domain material from this week in google.

35:45 - Mike Elgan (Guest)
Whenever I run into that myself, I just say what, what copyright, and it's like oh, you're right Okay.

35:53 - Leo Laporte (Host)
That's bonita, our producer. So that's why our show notes are so good for this show. Wow, just lie, just lie to your ai. Uh, so Just focusing not on ai, but on google and mike. You've been really critical of sundar pachai for more than a year now. Is this yet another stumble from sundar pachai?

36:15 - Mike Elgan (Guest)
well, I think the stumble is is the fact that their ai Set of ai tools is Coming to feel like their set of communications tools where they're messaging ads. What's the current one? What was the old one? Did they discontinue this one? Where's the? Where's this other thing? What happened to deep, deep mind? Nobody. It's starting to be very, very confusing to all, except for the extremely close watchers of this company, and this is going to be a real problem when they go out into the business world and try to sell this to enterprises, because the people they're going to be selling to are going to be just confused and they're going to want to clear answers and they're going to want to clear direction. And Google is just. They just never provide that ever. And nobody has any confidence that the tools at some point they're going to be tools that you're really going to have to invest your time and money into and nobody has confidence that they're not going to just pull the plug on those things.

37:15 - Alex Kantrowitz (Guest)
So they got to rip them off. Oh well, riddle me this one. Google has just turned in like obviously an all-time quarter, because they do it every time, but they added $6 billion of incremental revenue in the last quarter, in a quarter, and they had there. They reached their all-time high in January. Bing has done nothing to eat into the search lead of Google. Neither has perplexity. I just looked at a chart of all these different chatbots and AI search companies next to each other, and the company that gave me the data didn't even put Google on, Because if you put Google on, all the other companies look like a flat horizontal line at the bottom.

37:53 - Leo Laporte (Host)
Yeah, but where is that money coming from, Alex? It's coming from ads. It's coming from search Search ads.

37:58 - Alex Kantrowitz (Guest)
Search ads and they are crushing on that front, and you could call it a natural resource because, by the way, I'm still trying to figure this out myself. You could call it a natural resource curse, but, that being said, seems like they'll be able to live on that natural resources maybe as long as the internet exists.

38:15 - Mike Elgan (Guest)
I would look at the leading.

38:17
So Google is an extremely global company and their search revenue is global. I would look at the more bleeding edge parts of the United States and see where people are going, and it's clear that people are wandering away from Google Search and embracing things like perplexity and other tools ARC, browser All these things are gone in for Google Search, and many of them, in most cases, are actually better than Google Search. I would urge people to start with perplexity or find phindcom. These tools are kind of like Google Search in the sense that they give you an AI answer, but then they give you links to the stories where they got the majority of the information they're giving you, and so they're both search engines and chat bots, and those are the tools that are better than Google Search for most people most of the time, and I'll betcha that in Silicon Valley and New York, like in some of the leading users of information, that the exodus away from Google Search has begun for sure, and it's just a matter of time before the rest of the world follows.

39:22 - Leo Laporte (Host)
So there's a bit of a disconnect too, between covering a company's financials and whether a company is making a good product. So McDonald's will always sell more beef than anybody else. Is it the best way you can eat a cow? No, but it's got a mass market, and so I often am torn here because we're not giving people stock tips. We're not covering finances, except to the point that finances has to do with the welfare and future of a company. We're really trying to find what's the best technology.

39:58
I stopped using Google Search months ago. I moved to Neva and, to your point, alex Neva got put out of business. In New York Times article today about Duck, duck Go and Neva and how hard it is to beat the incumbent, neva went out of business because they said they even said we can't beat Google. Google was so dominant, even though we produced a better product and I agree they produced a better product it wasn't sufficient to survive. I'm now using Kagi K-A-G-I Same thing, very much similar. I think the search results are as good as Google's and it's got the AI. I use Perplexity on my Arc browser and it's very good, but Google's the McDonald's of search.

40:38 - Alex Kantrowitz (Guest)
It's hard to beat Exactly, totally right. I mean, in Leo, you talked about expectations, right. So what is our expectation? If Google is the McDonald's and that's kind of going to the question about Sundar and I do think finance plays a role here, because finance is sort of the thing that directs all these companies they will make financially motivated decisions. That's just what they do.

41:00
So if you were to replace Sundar, who do you take him away? Who do you replace him with? It's like if you replace the CEO of McDonald's, do you replace him with a chef at a steak restaurant who serves a five wagyu? You just wouldn't do it, because it's just not the product, it's not the job, and the company will be less successful if you do that. Or, for instance, let's say they find artificial meat or lab-grown meat. They're not going to just go ahead at once and convert away from that, as opposed to what they do Now. Listen, I think McDonald's is something that is tasty, but you don't want to have it every day, and so be it. Search, though, you can use it every day, and obviously billions of people do.

41:45 - Leo Laporte (Host)
Well, unlike McDonald's, there isn't really much of a competition for Google McDonald's. You can go across the street to Jack in the Box, Burger King, in and out five guys. There's plenty of choices for crab meat hamburgers.

41:57 - Janko Roettgers (Guest)
You go to Bing Because you want to. You're allowed to. That's not much. People are allowed to.

42:01 - Alex Kantrowitz (Guest)
They have a choice too. They just choose Google.

42:03 - Mike Elgan (Guest)
The thing about fast food, though, is fast food, I think, is a good model, because a lot of fast food chains not necessarily McDonald's, they start out as these American fast food chains. Americans kind of sour on the idea, and then they spread like crazy. I mean, kentucky Fried Chicken is huge, oh my god, it's all over Beijing.

42:19 - Leo Laporte (Host)
I couldn't believe it.

42:20 - Mike Elgan (Guest)
Yeah, it's big in Asia. But the real risk for Google isn't whether people will find another tool for explicit search. The risk is that people are going to be moving away from explicit search, by which I mean people aren't going to go. I'm going to go search now, Like the information is going to find its way into all the tools we're using. Well, that's what happened All the devices we're wearing into our glasses, into our watches, all this kind of stuff, the information will just start flowing in and there'll be increasing agency on part of the LLMs giving us this information. So Google, of course, can be at the forefront of that. They have plenty of tools and they're Google Docs and all that kind of stuff Opportunities for them to do the same. When you shift model, when you go from silent movies to talkies, a lot of the people who are succeeding in the old world go by the wayside and new companies come and take their place. So that's the real risk for Google is that people will stop explicitly choosing to go search.

43:19 - Janko Roettgers (Guest)
I think Google does other things than search too, though we shouldn't forget about that YouTube is like a $9 billion business. I made $9 billion, I think, in the last quarter. And then, if you talk about information retrieval in glasses, right, who owns the biggest spatial information layer with Google Maps? I mean, I just said it, but I think they have a lot of hands and pans. What's the pots on the stove? Whatever the image is there I'm looking for, but they're having a lot of things cooking, iron's in the fire.

43:52 - Leo Laporte (Host)
Iron's in the fire. That goes way back. There you go. Ha ha, ha ha. That's a good point. They have it. They've got to die. That's why. Google's a puzzlement, right? The other thing that's interesting about Google like Apple, a lot of what we think of as Google is historic. Google came around by being an innovator, by knocking off the incumbents, excite altavista. Now on comes Google, and it was so apparently better. Everybody started using it. Youtube, by the way, is doing the same thing to media in a dramatic way.

44:31 - Mike Elgan (Guest)
And we have to look at the generational issues, like how many utes are using TikTok. Well, that's right.

44:40 - Leo Laporte (Host)
So we have, you know, apple's getting a kind of, I think, a free pass on Vision Pro because of their history of taking technologies that exist and making them so much better that everybody goes oh that's it, the sky's open and dollars come pouring in. Google's same thing it has this reputation for being this great innovation factory more PhDs per square foot than anywhere else in the world, et cetera, et cetera. But the young people don't have that historical context right. They're just going to use what works. They probably know YouTube a lot better than they know Google Search.

45:18 - Alex Kantrowitz (Guest)
I don't know. I mean, if you have a phone, you use Google, right, whether it's an Android or an iPhone, that's the default. So you know I mean thinking about this. This is really a question about reinvention. I mean, I've been talking about Google staying still. I mean, I wrote a book called Always Day One, which is all about reinvention.

45:34
All the tech giants have reinvented, whether that's Apple, from the Mac to the phone, or Amazon, from the retail site to AWS. The big question about reinvention and this is going to be this is again why Google is such a fascinating company and why it's so fun to debate these things and try to figure them out live. The biggest questions about reinvention is how big of a change are you going to bet the entire company on and what's the magnitude that that change will actually disrupt your core business? Because the innovator's dilemma cuts both ways right. If you bet the farm on something that wasn't going to disrupt your core business, then you don't have a flagship business and you have a tiny business that you've just reinvented into Exactly.

46:17
And I think that is crucial, because there is an argument to be made that things like perplexity, which are really engines that are of Insurnivost, the CEO calls them their engines for curiosity, that you know, the sky is bound as limitless, because human curiosity is limitless and they can help fulfill that. But that's different than search, right, search is trying to get to different web pages on the internet and it's possible that these actually exist in two separate spheres. So Google has to make this determination of like if we're going to go all in, is this going to fully replace our core product? Is it going to partially replace our core product? When are we willing? Because eventually, every company has to get to the point where you're willing to sacrifice your cash cow for the future. And that'll be the big question for Google over the next decade or longer is when is it time to sacrifice that cash cow to give rise or give air to the new product?

47:12 - Leo Laporte (Host)
Even Amazon has not been able to keep it always day one right. I think it's easy to say these companies are going to innovate, innovate, innovate in a limited time frame, but as the time frame expands, all these companies are going to go by the way. Are they not going to go by the wayside? We have suddenly reinvented the business cycle Magically. We've got eternal companies. I don't think so.

47:34 - Mike Elgan (Guest)
And it depends on your access, except being the only company left someday. It would just be one, but then there was one. I'm concerned, yeah, but imagine the likelihood that Apple will cancel the deal that they've had with Google, where Google pays them a ton of money and Apple makes Google search the default on iPhones, for example. I think the chances are very high that Apple will actually do that. They have their own thing.

47:57 - Leo Laporte (Host)
Well, you saw the story this week that actually ironically came out of Microsoft, or my brother, out of Google, that Apple was. Microsoft tried to sell Bing to Apple in 2015. And Apple said, no, it's not good enough, right.

48:12 - Mike Elgan (Guest)
But Apple's moving to. They're going to move right through the Vision Pro onto regular sort of AI glasses. Do we think Google Search is going to be the thing that's going to be delivering information on Apple's glasses?

48:25 - Leo Laporte (Host)
Well, and that's the larger question Is that type of search where you enter in to a field, a thing, and you get web pages associated. That is antiquated. That's why Perplexity and ARC and all of these choices. If you look at the Perplexity web page, it's not a web search. It's where knowledge begins and you enter in a question and, yes, you may get web pages as authorities on this, but it's much better than just a web search and I think that's really where Google's going to get this intermediated. We don't need web search, we've got something better.

49:00 - Alex Kantrowitz (Guest)
Yes, that's debatable. I really think that's debatable. It's like, let's say, we go to this era of AR glasses. Is the phone going to go away? I don't think so.

49:11 - Mike Elgan (Guest)
Nothing ever goes away, but I think that's conceivable, that within 10, maybe 12, 13 years, people may use glasses or some kind of AR more frequently than they use a phone. I think that's the most likely candidate to replace the iPhone as the central device in everybody's life. But even if they don't, I think, even if it's just a peripheral device to a phone, an accessory to a phone, I still think that the whole idea, the expectations around how we get information and who instigates it I mean one of the things with Perplexity and some of these other tools is you enter in your information. It gives you a paragraph or two and then below that, it basically prompts you for follow-up information. It's just a matter of time before it's prompting you for the initial search based on your context, your location, what you're looking at. Glasses are going to be instrumental in that looking all the time and saying, hey, I'm just feeding you information without you even deciding explicitly to go learn something. So I just think the model is going to be blown up and go ahead.

50:18 - Janko Roettgers (Guest)
I think it would take longer than we think now even with the glasses.

50:21
It would take a while, for example, I've been testing the Meta Ray bands with the new multimodal AI thing that they have been releasing slowly over the last couple of weeks, where you can tell the glasses to look at something and then answer a question about it, and it's interesting and it's potentially useful in some use cases and frustrating in 90% of the use cases, because it's still very gated, it still doesn't have access to a lot of things, it still makes mistakes, it doesn't know nearly as well what you're looking at when you're looking at something. So that's getting to a point where any of this is useful. It still weighs out and that gives those companies some time to catch up on things and figure out what is search going to look like? How are our products going to fit into that future world?

51:07 - Mike Elgan (Guest)
Right, and I've been wearing this go Maybe Ray-Ban met us a lot too, and while it's true what you're saying, that the real use cases for broad use and just mass use is a ways out, it feels so inevitable, it feels perfectly inevitable, I almost feel like that, this is going to be a big deal.

51:27 - Leo Laporte (Host)
The Meta Ray-Bans are closer to what the future than the Vision Pro is. I don't want to wear that thing on my face, but I wear glasses. It almost feels like Meta is coming at. Obviously they're both coming at from different directions, but Meta is closer just by virtue of the form factor.

51:46 - Mike Elgan (Guest)
Well, brilliant Labs has something called Frames, which is just like Meta Ray-Bans, in the sense that it can see things, you can talk to it and so on, but it will actually put a visual display in front of you.

51:58 - Leo Laporte (Host)
Oh that's the heads up is very important. Yeah.

52:01 - Mike Elgan (Guest)
And it's very inexpensive too. So it's like it's 400 bucks. It's slightly more than Meta Ray-Bans base model.

52:09 - Leo Laporte (Host)
And it only looks slightly weirder.

52:12 - Mike Elgan (Guest)
I don't want to look like John Lennon with the round glasses, but again, these are just. I think by the end of this year we're going to have a bunch of these different variations, different levels of quality. I think the big thing for Ray-Ban Meta Glasses is they took a thing that was already out there and they made it super high quality. The audio is really good, the microphone is great, so I think that it's just. I think competition in this space is going to really shake things up in the world of search and just the world of gathering information.

52:41 - Leo Laporte (Host)
That's a good example of the innovators dilemma you were talking about, though, Alex, because we just saw this week actually it was today in his power on newsletter, Mark Gurman, talking about other form factors that Apple was thinking about but didn't actually try to develop, including glasses and a ring and air pods with cameras, but decided not to do, and I think that that was probably because they're really based around making big hardware and with big, powerful processors, and none of that appealed to them in the same way.

53:16 - Alex Kantrowitz (Guest)
I mean, they're going to have to try all this stuff, but can I just say that, first of all, I am on board. I think that AR glasses are going to be a very important part of the future. It seems obvious at this point. The question is what the timeframe is. But coming back to this entire discussion, something stay and some things go.

53:32
So many people have written stories about how email was going to die, and email is still email and it plays a very important part in our entire online existence. I mean, the entire podcast was at least from my perspective, was coordinated via email with me and your producers, leo, and there was a moment, even five years ago, where people were talking about how Slack was going to replace the email. And now imagine if you're an email company or a user of an email service and they decide okay, pivot, now we're not doing email, we're doing chat. You'd be furious because they broke a format, and so I think we're obviously it's like it's crazy to say, but we're still pretty early on In terms of where this internet thing goes.

54:17
These companies that we're talking about are 20, 30, 40 years old. That's a millisecond in the history of development. And we just got ChatGPT a year and a half ago. That's insane. So we'll see what happens and that's the real. As a user and as a developer of these products, how quickly they transform is super important because you don't want to miss the wave, but you also don't want like how many times have we been using products and they redesign and then they suck Right Because they chasing something new and they just don't nail it? You obviously need to be welcoming of change and ahead of change, but there's a risk on the other side of that.

54:59 - Mike Elgan (Guest)
Yeah, and in Chidification goes well beyond social. Yes, the phrase that I think we all should be using more is AI glasses and not lump them all into AR glasses Arabian metaglasses are AI glasses, right, and as people become enamored of AI, they're gonna want it in their glasses. It's just a great place to put microphones, speakers, and I'm sold on it.

55:26 - Leo Laporte (Host)
That just makes more sense to me. So I just ordered these brilliant labs because I wanted to order the meta ones, but you already had them, so I'm gonna get them and I decided to go all in. You can have choices. You can make them look like pointext or black. They have a cool gray, but I decided I'm gonna go with the clear one, so you can see that I am wearing technology and I will look like an aviator in my glasses.

55:51 - Alex Kantrowitz (Guest)
We'll see what's the use case that you wear them in, like you're obviously not gonna wear those all the time.

55:57 - Leo Laporte (Host)
Well, I'd like to. How are you Maybe?

56:00 - Mike Elgan (Guest)
The problem is they don't look cool enough. Arabian metas look more socially acceptable as far as I'm concerned and somebody's gonna come out with I mean, even these guys will come out. Their next version of glasses aren't gonna be pointext or round glasses, but if they look like regular glasses and your prescription lenses are in them and they might be transition lenses or whatever, I mean there are many days where I'll wear Ray-Ban Meta glasses for eight, 12 hours a day you can get.

56:26 - Leo Laporte (Host)
By the way, the brilliant labs also sells a monocle, if you really, yeah, I know I always say, yeah, you really wanna look? You look like you're the monopoly guy. Like the monopoly guy, I might have to grow a mustache. All right, let's take a little break.

56:40 - Alex Kantrowitz (Guest)
Can you imagine being on a date and being like I'm gonna just use this monocle. It's AI.

56:46 - Mike Elgan (Guest)
Actually, one of the biggest use cases was for dating. There was a guy, so the original monocle device was a programmable thing. It was the thing that you're supposed to take and modify, and the main modification was a guy who wanted it on dates to tell him what to say to his date.

57:02 - Alex Kantrowitz (Guest)
He put the monocle on and that's what ended the date. I understand that's exactly it.

57:07 - Mike Elgan (Guest)
Anyway, it's.

57:08 - Leo Laporte (Host)
You know, the monocle could go over your regular pair of glasses, so just clip their clip-ons.

57:14 - Janko Roettgers (Guest)
I don't know.

57:16 - Leo Laporte (Host)
That's not talky at all, not at all. Well, they ship April 15th to just in time for tax time Tax day. I will break out my brilliance. Thank you for the tip, mike. I had my finger hovered over the buy button on the Meta glasses and I thought oh yeah, nuts. What am I crazy? You got me at a. I was vulnerable. Oh good, yeah, nice, let's take a little break. Got a great panel. Alex Cantrowitz is here from Big Technology, the podcast, the newsletter at bigtechnologycom. Yanko Reckers has his newsletter. Actually, all three of you are newsletter guys. Lowpasscc for Yanko. What platform are you using, substack or Beehive? Oh, I asked you that last time, beehive, and you still like it.

58:02 - Janko Roettgers (Guest)
I like it a lot.

58:03 - Leo Laporte (Host)
Yeah, Alex is Substack right, that's right. And Mike, are you medium or Substack? What are you?

58:09 - Mike Elgan (Guest)
using. I'm Substack. I'm Substack and I actually have a really good thing going there. It's a good source of revenue and it's a wonderful platform.

58:18 - Leo Laporte (Host)
I like it a lot. Talk about revenue. Alex has 160,. Was it 160,000?

58:25 - Alex Kantrowitz (Guest)
Yes, those are free subscribers. So if you're a free subscriber and wanna upgrade, I invite you to do the math.

58:32 - Leo Laporte (Host)
Times zero. Oh, I thought those were paid and I thought you're buying next time we have dinner, that's for sure.

58:40 - Alex Kantrowitz (Guest)
Oh, my God Wow.

58:42 - Leo Laporte (Host)
Well, good job.

58:42 - Alex Kantrowitz (Guest)
Make and dream.

58:43 - Leo Laporte (Host)
No, you know what. That's where it starts. That's where it starts with the free ones. That's how we started people onto it. We gave it away. Now we ask people to pay $7 a month and actually we're getting there. We got almost 12,000 subscribers to the club. We need more because advertising dollars are dwindling and you see, this is why you see people like these three really smart, good journalists doing their own thing, because Vice just shuttered its doors, right? I mean, it is a bad time for journalism and if you want good journalism, I think you gotta subscribe, whether it's to Big Technology or lowpasscc or Mike's Substack or Twitter or all four Come on, and I don't know if you remember this, leo, but I used to be super like pro advertising and I still am to a certain extent.

59:33 - Mike Elgan (Guest)
But the model that you have with Club Twit and, to a lesser extent, the model that Substack or some of these other platforms have, is actually more valuable, because what people want even more than products is they want community, yes, and so a thing like the Club Twit is very gratifying. If you have strong interest in technology, you want a community of people that you get to know everybody. It's fun, it's really important.

59:58 - Leo Laporte (Host)
The community is really what it's all about, I've realized. And so and that's funny, because you know PC Magazine or you were the editor of PC World, was it? I forgot Windows Magazine. Windows Magazine, I mean, I guess you have a readership, but it isn't an interactive community in the same way, and there's something so much more satisfying, I don't know. Do you also notice that Alex and Yankov, writing to an audience that's engaged, they talks back to you. That's better isn't it?

01:00:27 - Alex Kantrowitz (Guest)
Oh, it's amazing, and I think I mean, leo, we can probably agree on this there's no better audience than the podcast audience. Like having a podcast on, there's just this connection that you make with people, and I love hearing from people after they listen to big technology podcast hey, that was good, you missed that. It's like they're there with me and with my guests and like it's amazing, such an engaged audience, more than anything I've ever done in my life. For sure you too, yankov.

01:00:56 - Janko Roettgers (Guest)
I don't have a podcast, but Well, you have a newsletter. I do like having my newsletter audience. I do like getting lots of feedback from them. I have a little Slack community a very small one, but it's a nice little club, essentially for paying subscribers. So, yeah, it's always fun to have a more direct connection to your audience, for sure.

01:01:14 - Leo Laporte (Host)
That's actually one of the negatives is you want to be successful, but there's something about having a small community that you know everybody in the community and you all kind of get along, and then it gets bigger and bigger. And as our club has gotten bigger, I don't know as many of the people in the Discord and so forth. I mean it's a tough one. We need to get bigger if we want to keep doing what we're doing, but at the same time, I like the community we build. We have a couple of ways of being in the Twitter community that aren't paying. By the way, we have a mastodon instance at twitsocial that you can sign up for. It's free. We also have, I think, a very active and fun community forum at twitcommunity, and both of those are free twitsocial and twitcommunity. But we would love to have you in the club twittv slash club twit Our show today, brought to you by SpigsyBean communities. There's a wonderful community around Ecamm Doc Rock and Luria Petrucci and well, we're Ecammers too. We use Ecamm for iOS. Today, rosemary and Micah use that. It's the leading live streaming and video production studio built for Mac. We're on our third studio. We've spent millions on studios and, honestly, if I were to do it all over again, we'd probably just do Ecamm. Ecamm does everything we do in our big studios, but you could do it on your computer at home.

01:02:41
Whether you're a beginner or an expert, Ecamm is here to elevate your video production from streaming and recording to podcasting and presenting. Ecamm Live is your all-in-one video tool, perfect for simplifying your workflow. I actually use it frequently. Ecamm Live includes support for multiple cameras. For screen sharing, you get a live camera switcher. So, just like we do here, you could direct your show in real time. You stand out from the crowd, too, with high quality video. You can add logos the lower thirds that we use, you can do those easily titles, graphics of all kinds. You could share your screen. You could drop in video clips. You could bring on interview guests. You can use a green screen, and on and on. Ecamm is amazing. Join the thousands of worldwide entrepreneurs, marketing professionals, podcasters, educators, musicians and other Mac users who rely on Ecamm Live daily. You'll get one month free when you subscribe to any of Ecamm's plans by going to ecammcomtwit Ecammcomtwit Use the promo code TWIT to check out for one month free, and we thank Ecamm so much for their support and for helping us produce two of our most important shows.

01:04:01
Let's see. At&t had a little problem on Thursday Whoops, in fact. I first saw the stories early in the morning that we're saying that other wireless providers are having problems. That was because people on other carriers were having a hard time reaching AT&T and they thought their carrier was down Late. On Thursday AT&T explained that wireless service had been out for almost 12 hours. I think this Reuters story must be a little dated because it says affected more than 70,000 users. At its peak I think it affected a lot more people than just those 70,000 people was out from 4.30 am Eastern time on Thursday till 4.00 pm Eastern time. That's 11 and a half hours. But there's some good news. If you're an AT&T subscriber, they feel bad about it, so they're gonna put a whopping $5 bill credit on your bill to make up for that loss of service for almost a day. By the way, not $5 per phone, no, just per account. So if you have one of those family plans, you get a whole $5.

01:05:17 - Janko Roettgers (Guest)
Don't spend it all at once.

01:05:21 - Leo Laporte (Host)
The funniest thing I thought and I'm not sure if it's funny or not, but AT&T's explanation of the outage. They told ABC News of the outage is not a cyber attack. It was caused by quote the application and execution of an incorrect process used as we were expanding our network, which is very corporate, because they ended up getting a little plug in there for their expanding network At the same time I was explaining why their expanding network was down. I don't know in a technical way what the application and execution of an incorrect process is and I don't think we're supposed to know.

01:06:04 - Alex Kantrowitz (Guest)
They did not do a good job.

01:06:06 - Leo Laporte (Host)
They somebody screwed up. Abc's sources said it was a software update that went wrong, but there's a lot of things that could cause this, including misrouting and BGP. Some wag on Reddit said obviously there was a server in a closet, long forgotten. That certificate expired and the only guy who knew it was there and knew how to renew the certificate had been fired eight years ago, which kind of sounds more accurate.

01:06:31 - Janko Roettgers (Guest)
I suddenly built a wall in front of it.

01:06:33 - Leo Laporte (Host)
That's right. Dhs and FBI are both investigating, but there is, according to US security officials, no indications of malicious activity. That's the first thing you think of when a big infrastructure player goes down.

01:06:49 - Mike Elgan (Guest)
A company like AT&T is like a bank. They can't afford to have people thinking that you know whatever the real story is is probably embarrassing. Exactly, they're burying it with euphemism and corporate speak.

01:07:03 - Alex Kantrowitz (Guest)
We did get some amazing memes out of it, though. There were AT&T employees standing out of the in front of the AT&T store with lots of people, you know, being like what's going on here, and they had these looks on their faces like my God, and people captioned it like we're at the phone company, we're gonna break.

01:07:21 - Leo Laporte (Host)
We don't care, we don't have to and we're gonna we don't have to.

01:07:29
Well, all right, I'm not gonna. Oh, I love it. I thank goodness for Travis Kelsey yelling at his coach, andy Reed, during the Super Bowl, because we that was the the uh yell that launched a thousand memes. Um, anyway, I don't know what happened at AT&T. I doubt we'll probably ever know. I'm sure Steve Gibson will do a deep dive on Tuesday in security. Now, but look for your $5 credit sometime in the next five or six months.

01:08:01
If you even notice, uh, better news police, uh, europe European police, interpol, uh arrested More lock bit affiliates. I don't want to say these were the guys who created the lock bit ransomware. They, because this was ransomware as a service, and so these were the people who were using it to blackmail companies. The law enforcement arrested two operators of the lock bit ransomware gang in Ukraine and Poland and created a decryption tool to recover encrypted files for free, seized over 200 crypto wallets. They're ill-gotten gains. It was, it was a lot of money and put a couple of people in jail. This does not mean it will go away. In fact, these, these ransomware gangs have a habit of reforming quite quickly. You, you were covering this on lowpasscc. You're a newsletter. Yanko Walmart is buying Vizio.

01:09:05 - Janko Roettgers (Guest)
Yeah, that's a big story. This week, yeah, they spent 2.3 billion dollars on TV brand TV maker. It's really interesting. I wrote about it both on fast company this week or for fast company, and on my own newsletter, I think. Walmart the story there is and it's already in the headline here for CNBC as well. They do want to expand from just selling stuff to also having recurring revenues with Advertising, and the way you do that these days is smart TVs or one of the ways to do that.

01:09:42 - Leo Laporte (Host)
It's, they admit it that it's. They're not buying it for the hardware, they're buying it for the ad business.

01:09:49 - Janko Roettgers (Guest)
I mean, that's why everybody's anybody selling TVs. These days the TV business is completely nuts and I mean, if you have gone to any of those stores, whether it's Walmart or West, by Best Buy or Costco, you can get these like 55 inch TVs for 250 bucks or something like that. Right, it's insane when you think back, just five years on, the same TV cost I don't know two thousand dollars. Well, maybe not five years, but a little bit longer. So TV prices have completely cratered over the last decade and at the same time, tvs have arguably gotten better, right, okay, it's all you can buy. A that's $250 TV that you're gonna pick up actually gonna look pretty good. I mean people who want a home theater setup perfectionist, whatever they're gonna still spend two thousand dollars on the TV, no matter what. But you can get a good TV for two, three hundred dollars these days the first plasma TV I ever saw was in the lobby of tech TV.

01:10:46 - Leo Laporte (Host)
50 inches wasn't even that big. It cost ten thousand dollars. I know, I know, incidentally, had horrible burn in about a year left after a year in the tech TV bug was burned into it. So we've got a lot of ways, yeah yeah and but.

01:11:02 - Janko Roettgers (Guest)
But TV hardware has gotten so cheap and the only way that these TV makers can make any money at all anymore is basically through advertising, through services, through Partnerships with some of the streaming services, through running their own streaming services. Any TV to buy these days but that's a busy or a Samsung as soon as you turn it on, it's gonna have a free streaming service on there with, like these live linear channels, a couple hundred of them, all free, but all full of ads, obviously.

01:11:33 - Leo Laporte (Host)
So that's how they make money you point out in your fast company Peace that in the first nine months of last year Visio lost 45 cents on every TV it sold. Yeah, it's insane they were selling it in low cost.

01:11:45 - Janko Roettgers (Guest)
Part of that. Part of that is to obviously there was the pandemic going on. There was like supply chain shortages and then suddenly they had too many TVs and companies from China were sort of Saturating the market. But really the only people who make money with TVs anymore are either the ones that have really robust services on them or or maybe if you're a Samsung, if you're an LG, and you sell some of these really high-end TVs to a subset of your audience.

01:12:14 - Leo Laporte (Host)
But let me point out you're gonna have, because I bought the top of the line Samsung QD OLED. It has no fewer ads on it, it has no. Less snooping, spot a software on it. It still has all of the all of the ad revenue and it's a premium TV for premium price, so they get the best right, right right. Visio they said, generated in the same nine months. They lost 45 cents per TV, 260 million dollars in profits on ads.

01:12:44 - Janko Roettgers (Guest)
Right, right, right, you know this advertising. So just having their own video service on there. But what all these TV makers do as well as like if there's a Pluto or a to be some of these free TV, free streaming services, for them to be on those platforms, they actually have to share some of the ad revenue interesting. So they have to give I don't know 30% or 40% it depends on the manufacturer of ad slots to these companies and they're gonna just pump their own inventory in there. So anytime you watch anything that is ad support on your smart TV, the company that made this smart to me is probably making some money of it.

01:13:19 - Leo Laporte (Host)
They. Even I think Visio does this and I know Roku does Charge for the button on the remote. So, right, there might be a Netflix button or and those are expensive to get those buttons and they change, by the way, which is really hard to get used to. If you get a new Roku, the buttons are different because the advertisers it's an advertisement on the on the remote control, and some of these remote controls now have six or eight branded buttons.

01:13:42 - Janko Roettgers (Guest)
It's insanity. But well, that's the money, yeah.

01:13:46 - Mike Elgan (Guest)
I think one of the. Parts of this story that needs to be highlighted is the fact that one of the things Walmart wants is user data data. Yanko mentioned advertising. They're doing targeted advertising based on user data, and I would look at other places where Walmart is Harvesting and collecting user data. It sounds like they really want to get more deeply into that world.

01:14:11 - Leo Laporte (Host)
Walmart's advertising business is called, and I think this is telling Walmart connect, connect your users with their data, consell it to the highest bidder, including, I'm sure, not just advertisers, but data brokers, right yeah.

01:14:31 - Mike Elgan (Guest)
Who knows? But I think there's a Something that we ought to look into, and by we I mean yanko.

01:14:37 - Leo Laporte (Host)
Well.

01:14:38 - Janko Roettgers (Guest)
I'm not gonna do that it is sort of, if you imagine, combining all the data what what people see and watch with all the data of the stuff that they buy, and then sort of figuring that out For any advertiser who then is going to have spots on Visio TVs in the future to know that they're gonna reach the exact audience that's gonna walk to the Walmart on the next next day or next night to buy your products. That's really valuable connection. All right there, all right.

01:15:05 - Mike Elgan (Guest)
Yeah, and a play. You know this is yet another one of the million uses for AI. You really watch what people are watching, when they stop watching, when they pause, when they get up, when they whatever, and then Fold that into AI and say what does this mean for what they're gonna buy and when? When they want to buy it?

01:15:21 - Leo Laporte (Host)
So you are, you're starting to say, alex, that Amazon is the same business, right?

01:15:26 - Alex Kantrowitz (Guest)
I mean with their subsidized TVs, I mean we think about this, the, the business size is massive, right, I mean, amazon made forty nine billion dollars on advertising revenue last year alone. So when you think about that in context, spending two billion dollars for a connected TV company, that will probably you know that's, I think you know doing quite well as far as revenue goes. You layer your data on top of that, it's a slam dunk. Now you take those TVs and you give them prime real estate in Walmart. All of a sudden, you know you're making sales and your ad business expands.

01:15:58
So this, this, this category of Commerce, advertising right, advertising only commerce sites, whether that's Amazon or Walmart, and advertising on TVs would connect to data. It's huge. It's the fastest growing segment of advertising in the world. Right now, as you know, facebook and Google's growth rates have tailed awful, even though they're still doing quite well. So if you're, if you're Walmart, it would almost be malpractice not to get into this now. For the consumer it can be kind of tricky right, like if you're paying not paying a lot of money for those TVs, you're paying for it with something else. But this is definitely an area where we're going to see these efforts ramp up for for any e-commerce company and I also think it's important to look at.

01:16:39 - Mike Elgan (Guest)
Well, first of all, it's still subject to Vizio's board approving the deal, but their market share declined from 2021, when it was 3.1%. In 2020 it was 3.3%. So they have declining market share and they already have a small market share. I mean, you know, 3% basically, and so who knows what Walmart's gonna do with that. Like, like Alex was saying, you know you start putting those at the front door. When people walk in and really really selling it, they may grow in market share and and turn this into actually turn these TVs into a much bigger deal.

01:17:17 - Leo Laporte (Host)
You may also wonder what's the status of Legislation against data brokers and, in particular, you remember Ron Wyden wrote a letter asking whether the defense intelligence Services were buying data on Americans from data brokers, to which the response was yes. Right now, section 702 of the FISA Act Allows that. There been some moves in Congress to require a warrant or Judicial oversight of some kind, moves which have been opposed by the Biden administration and others. The the reauthorization for section 702 is up comes up in April, and I'm gonna make a prediction that there will be, in fact, no limit placed on intelligence agencies ability to buy data information about you and me from data brokers. And that's one of the reasons this whole thing continues to exist, because law enforcement says this is great data, keep collecting it. So even if you and I decided we don't like this, it's a privacy invasion, or Ron Wyden says you know we ought to do something about this.

01:18:28
I don't know if we'll ever see any curtailment of this. It's too valuable for law enforcement and and it just means data brokers get to continue Completely unrestrained. Watch in April, see if they're. April 19th is the is when it needs to be passed. We'll see if there's there. You know there been moves by various members of Congress. You know, we really, we really ought to do something about this, to which law enforcement quite vigorously says no. I don't think so we can always count on our government to do nothing in these situations.

01:19:03 - Alex Kantrowitz (Guest)
It's always one of the special talents Safest bet, if they do anything what they'll do is they say well, you know what, we're gonna.

01:19:10 - Mike Elgan (Guest)
Let you keep selling it, but you got to give it to us for free, oh yeah.

01:19:14 - Leo Laporte (Host)
That's there you go, we're saving the taxpayers some money here.

01:19:19 - Mike Elgan (Guest)
Right and law and order. Explanation point on order, point.

01:19:22 - Leo Laporte (Host)
More importantly, all right let's take a little break. Wonderful panel, love having smart people on, people smarter than me, and that these guys qualify Mike Elgin, yanko Wreckers, alex Cantrowitz our show today brought to you by net suite. Once your business gets to a certain size, it's like your club, you know, once it. Once it gets to the point that I think somebody said I know my business is too big. When I walk around I don't know the names of people who are you. You work for me. Oh, interesting.

01:19:52
As the business gets bigger, cracks start to emerge and there are, I'm afraid, too many manual processes to keep track of. We've all been there. If this is where you are, you should know three numbers 37,000, 25 and one. Shall I explain? 37,000 is the number of businesses who have upgraded to net suite by Oracle. Net suite is the number one cloud financial system, streamlining accounting, hr and more. 25. Well, believe it or not, net suite turns 25 this year 25 years of helping businesses do more with less, close their books and days, not weeks, and drive down costs. What's the one? Well, one you, because your business is one of a kind, so you get a customized solution for all your KPIs in one efficient system with one single source of truth, manage risk, get reliable forecasts, improve margins everything you need to grow all in one place. Right now, download net suites popular KPI checklist. It's absolutely free. It's designed to give you consistently excellent performance and you can get it at net suite dot com. Slash twit. Net suite any TSU I T E. Net suite dot com slash twit to get your own KPI checklist. Net suite dot com slash twit. We thank you so much for their support of this week in tech.

01:21:25
Oh, bad news for wise. Speaking of privacy, this is this kind of is related maybe to what happened at AT&T. Somebody did something wrong when they were moving from one AWS database to another. Wise, which sells cameras. We've recommended very inexpensive, nice cameras. Unfortunately, during the move there was an outage and when the cameras came back on, many customers reported they were seeing footage of other people's cameras. Whoops, 13,000 customers being semi accessible to other accounts. Wise confirmed over the past weekend the camera thumbnails were accidentally accessible from other user systems for a brief period. You can see them in the events tab. They showed images not full clips, from cameras that were not from a user's own cameras. Initially they said, oh, it's just a couple of people, but now, in a round of email sent out to wise customers.

01:22:29
Wise said the thumbnail issue affected 13,000 customers, a lot of people yeah, oh, just over 1500 of them actually tapped the thumbnails to say what's going on here. Who is this? I know it's not clear whether you could see video.

01:22:50 - Mike Elgan (Guest)
It is not. I don't think there was video, I think it was just just a thumbnail, just yeah.

01:22:55 - Leo Laporte (Host)
I think so. I think that's not.

01:22:57 - Mike Elgan (Guest)
I'm probably yeah, I'm probably in the minority, but I think that makes it more fun.

01:23:03 - Leo Laporte (Host)
What's going on here?

01:23:05 - Mike Elgan (Guest)
I see an elbow. I see a knee.

01:23:10 - Leo Laporte (Host)
Most apps enlarge the thumbnail, but in some cases wise says the event video could be viewed, so maybe you were one of the lucky people who could click the thumbnail and see what was going on.

01:23:23 - Alex Kantrowitz (Guest)
It was a third party cash client. People were in compromising positions like people were getting ready for work when this outage happened.

01:23:31 - Janko Roettgers (Guest)
There was a 23 it's like home security meets chatter bait or whatever that's called.

01:23:37 - Alex Kantrowitz (Guest)
Remember we talked earlier about like you have to like be sufficiently, like advancing your technology, but also competent, like this is a competence issue for this company. Like if you're a security camera company you had one you really have one job Right, as people use it for watching their pets or monitoring in, you know, in their house. Your one job is make sure that that footage does not end up on somebody else's computer at your fault, and it's a huge failure here and it's not the first time, right, so they've had two prior issues.

01:24:06 - Janko Roettgers (Guest)
So it's very cheap, probably, but maybe you get what you pay for.

01:24:10 - Alex Kantrowitz (Guest)
Yeah, that's what I wanted one of those Walmart TVs, the hidden cost?

01:24:13 - Leo Laporte (Host)
Yeah, I mean, they were like 20 bucks and they had a security issue with the first wise camera that they couldn't fix.

01:24:19 - Alex Kantrowitz (Guest)
It was in the firmware, and so they're like why you don't let Gemini write the program for your security cameras?

01:24:26 - Leo Laporte (Host)
No, do not let Gemini write code you never know what will happen.

01:24:30 - Alex Kantrowitz (Guest)
Crack head chat GPT.

01:24:31 - Leo Laporte (Host)
I'm listening, I'm listening, I'm listening, I'm listening, I'm listening, listen, listen.

01:24:37 - Alex Kantrowitz (Guest)
That's what that camera was telling people listen stop listen.

01:24:41 - Leo Laporte (Host)
Wake up, people wake up, wake up, we're coming for you.

01:24:45 - Mike Elgan (Guest)
As part of the solution, they're gonna have an extra layer of verification, they say, before users are connected to event videos, which means that this is, this is one of the things that these things really cause a small cost on a huge number of people. Many, many times, going forward is like there's gonna be another step just to watch a video, and so I don't know. That's probably a good thing, I guess, but but that's what always happens with these things. They get Every security breach, every little problem like this. They add another layer of annoyance to the user.

01:25:18 - Leo Laporte (Host)
And Yanko has issued a correction in our club to a discord.

01:25:23 - Janko Roettgers (Guest)
Yes, very important, and don't google those things.

01:25:27 - Leo Laporte (Host)
He might have said something about chatter bait. He meant chat roulette. Let's get this clear although they're honestly, that may be a difference, without a distinction.

01:25:37 - Mike Elgan (Guest)
Yeah, chatter bait is a better branding.

01:25:39 - Leo Laporte (Host)
At least you know what you're gonna get. I love it, yanko. Reddit's IPO may or may not be coming to the New York Stock Exchange Sooner or later. They filed on Thursday and actually the thing that's been most interesting Is the documentation they had to file for the IPO, which revealed a lot of interesting information. Among others, we learned that google is gonna pay them $60 million To be able to use content from Reddit In its search and maybe even for AI training. I think it must be for AI training, right.

01:26:24 - Mike Elgan (Guest)
And does that sound cheap to anybody else? That seems like a low number, considering how pissed off it's gonna make the Reddit community.

01:26:31 - Alex Kantrowitz (Guest)
They made $800 million last year, so getting $60 million in a year is so significant for them. Maybe they could have held out for more, but let's say they are able to put together 3 or 4 deals of this magnitude You're talking about potentially like a 30-40% revenue increase when they're heading to the public markets. The other option would be, basically, they could have just locked all the information on Reddit down, which is another possibility, or they could have just made it free and available to everybody to train on, not looking at a lot of positive options there. However, I do think you're right at this point that the Reddit user base will revolt against the management. They hate the management there. They really do ever since it stopped allowing third party apps to hook into the API and I report on this this week. But basically you look at Wall Street Bets this week and they were talking about how they're all gonna short the IPO.

01:27:39 - Leo Laporte (Host)
Which was, I think, one of the adverse possibilities that they mentioned in their 10k filing right.

01:27:46 - Alex Kantrowitz (Guest)
They listed them as a risk factor and Wall Street Bets figured that out.

01:27:52
So I went on CNBC and I said, listen, there they asked me if it's gonna be a meme stock and I was like, listen, this is not gonna be a meme stock. I just want to Wall Street Bets this morning and they're saying they're gonna short it. And then Wall Street Bets takes the screenshot from that interview. Oh no, alex, it's the front page of Wall Street Bets and we won't talk about all the nasty things they said. But, like, the first 10 comments were like yes, we are gonna short it. Of course we are. By the way, the CEO apparently made more than a hundred million dollars last year A lot more a hundred ninety three million dollars.

01:28:21
And guess how much Reddit's loss was last year? Ninety million. So basically they're all saying if the CEO takes a hundred million dollars in compensation, as opposed to one ninety three, then they're gonna be profitable. I mean, it's crazy the way that this company is run.

01:28:38 - Mike Elgan (Guest)
Wow, wow. And they want to get buy-in from the big moderators and the big users by cutting them into the IPO. But I think that's gonna backfire too, because you know they're I have a feeling a lot of mods are gonna Expecting to be approved for this and they're not gonna be approved and they're gonna be. They're gonna be really revolting.

01:29:00 - Alex Kantrowitz (Guest)
Not to mention Robin Hood did it as well for their IPO and they're down 60% since their IPO after offering this user buy-in. So I expect the same thing to happen with Reddit, maybe worse.

01:29:14 - Mike Elgan (Guest)
Alex, do you think it's a bad idea for them to go public?

01:29:18 - Alex Kantrowitz (Guest)
Well, they might not have any option because they raised in 2021. They raised something like seven hundred million dollars and a lot of that at a ten billion dollar valuation, and that was when you know that was in the real number. Go up, era Right. And so now, like they have all this money, they need to return it in some way and they're talking about how their valuation might be five billion now and not ten. It might even be lower than that and like, if you're not gonna go now, when are you gonna go? Like the value of that money just gets less and less. So it might end up being that their IPO just returns like slightly above you know that investor money and then we'll see where the rest goes. Because the valuation is gonna be five billion, the raise will be less.

01:30:02 - Leo Laporte (Host)
I found your moment of fame on our slash Wall Street bets.

01:30:07 - Alex Kantrowitz (Guest)
By the way, nice picture, yeah, lol he said he read on Wall Street bets. They were actually nicer than I expected. He's not wrong.

01:30:19 - Leo Laporte (Host)
You know, they were pretty good people. I think they weren't talking about you. They were talking about Steve Hoffman when they said let's crush him.

01:30:28 - Alex Kantrowitz (Guest)
Oh yeah, yeah, they were talking about Steve Hoffman. Some people mistook me for Alexis O'Hanion because they're like Alex. Oh, you kind of look like Alexis a little bit yeah, yeah, I'm like, yeah, no, no, no, that's not me.

01:30:40 - Leo Laporte (Host)
So I've interviewed both Steve and Alexis back when they were friends, cause they're apparently broke up. He left.

01:30:45 - Alex Kantrowitz (Guest)
Alexis completely out of the S1. He didn't even mention.

01:30:49 - Leo Laporte (Host)
So the two were roommates for years. There is kind of a connection, because they were fans of Dig, which Kevin Rose started when he was at Tech TV, and everybody thought Kevin was gonna make 60 million dollars. In fact, he was on the cover of Business Week as the 60 million dollar man. Kevin assures me he did not make anything like that.

01:31:11
Dig eventually went under, probably because Alexis and Steve decided to do something similar called Reddit. It had a very similar idea people would submit headlines and news stories and then there'd be upvotes and downvotes and you could read what was going on. I think the thing they did that was better, much better, was the subreddits, where you really got communities and you could avoid communities. You didn't want to read, like Wall Street bets, and you could stick with communities. You wanted to read like Emacs. And I checked Reddit several times a day. I love Reddit, but Reddit has the same. Oh, anyways, I was telling the story.

01:31:45
So I guess he and Ohanian and his son Huffman sold Reddit to Cunninghast Very early on for 10 million dollars, but then they were was up and down. They came back, huffman became the CEO, alexis was on the board, they were buddies and then they had a falling out over how to handle Black Lives Matter and death threats in 2020, and I guess they haven't talked to me ever since themselves. Ever since, and on the filing, huffman obviously for some reason felt like he shouldn't mention that he wasn't the only founder. He was one of two. He wrote the code, though I'm told Alexis said he told me that Steve wrote it in Lisp. I mean, it was rewritten shortly thereafter.

01:32:31 - Alex Kantrowitz (Guest)
We did mention, though, that Sam Altman owns 8% of the company.

01:32:34 - Leo Laporte (Host)
That was interesting, wasn't it?

01:32:35 - Alex Kantrowitz (Guest)
Yeah, that's my perspective and Sam is important to this story here, because I think they should go to Wall Street and be like we are social network, but we are an AI company. We have all the data that's necessary to train AI bots and LLMs. And look at this deal we just did with Google and if you want to invest with us, come along for the ride. Put us make us part of your portfolio when you're investing in AI. We have the data we have. Sam Altman, don't make us number one. Let NVIDIA and all that you know lead, but put us in the bucket, and that might be what saves this thing. But again, the risk is do the, does the Reddit army revolt and they might. They might very well.

01:33:14 - Leo Laporte (Host)
Well, and this is a problem, twitter has, reddit has, facebook has. None of these companies create the content. We create the content and you know, in the early days of the internet remember IMDB, or what was the? Cd was it CD.

01:33:33 - Janko Roettgers (Guest)
CDDB.

01:33:33 - Leo Laporte (Host)
CDDB, where users contributed all that content and then these companies went on and sold to some you know big company like Condi Nast and the and the platform makers made all the money, but the people who actually made the content got nothing, and we've seen this happen enough times now that I think people are a little bit more suspicious. Reddit exists solely because of the 60,000 unpaid moderators who work every day, labor every day on the site and the many hundreds of thousands of people who both read and write content on the site.

01:34:07 - Janko Roettgers (Guest)
I think the bigger issue is not going to be that people are not getting paid as much as once that is a public company. They're going to be issues around content moderation they're going to make going to make tough choices, like right now there's still a lot of not safe for work content. Obviously on Reddit it's sort of age gated and it's in certain communities. But once they're a public company and they have to watch their store price, they have to attract major advertisers. Those advertisers may pull out if there's a scandal around that type of content. The choices that they're going to make.

01:34:40
They're going to make. Them are going to be very different from what they've done in the past.

01:34:44 - Leo Laporte (Host)
To be fair, we did think that would happen with Condi and asked, and it didn't.

01:34:49 - Janko Roettgers (Guest)
That is true. That is true. I think that was always a weird answer of relationship.

01:34:55 - Leo Laporte (Host)
Yeah, and Condi really wasn't that happy. As owners of Reddit, they managed to kind of put it off at arm's length. 76 million people visit Reddit website every day.

01:35:09 - Alex Kantrowitz (Guest)
So here's a fun stat Facebook had 410 daily active users when it filed to go public in 2012. Wow, reddit has 72. So, no sorry, facebook had 480 something. So so Reddit has 400, 410 fewer daily active users Facebook did when it went public and it was founded a year later. So Facebook's 20 years. Reddit is 19 years. It's not growing. It hasn't even reached the amount that Facebook had when it filed to go public or within a decade ago. Now I know it's unfair to compare Reddit to Facebook, but when you're an advertiser, that's what you're doing. It's coming out of the social media budget.

01:35:48 - Mike Elgan (Guest)
So it seems to me that there's nothing you can do with Reddit. You can't really change it without killing it. It's probably closer to the Wikipedia and they should probably have a similar monetization model. It should be probably supported by donations and other types of revenue. Trying to turn it into Facebook is never going to happen. Trying to turn it into anything is not going to happen. Even even X, even Twitter, are adding phone calls and it's going to be a super app and they can have, you know, financial transactions, all that kind of stuff. I just it's. I don't. There's nothing you can do with Reddit except let it be Reddit, and so they've got to find a way to monetize it in a better way than making a public company. That's just going to. That's going to cause problems.

01:36:36 - Leo Laporte (Host)
Didn't Jack really.

01:36:37 - Janko Roettgers (Guest)
The counter argument is go ahead, go ahead, janko please.

01:36:40
I was going to say. The counter argument is that on Facebook, facebook has way more users, but for brands to target those users on Facebook they have to sort of trust Facebook, that the algorithm is really going to figure out who is into a certain brand based on profile and a bunch of sort of black box stuff that you can't really understand. On Reddit you have communities and you can speak to that exact community and if you're a company that's big enough, there's a community dedicated to you. There's a if you make a big, popular product or service, there's a community dedicated to that. So you can target these communities that are interested in your products, brands, services, whatever, with very pinpoint precision. I think that's a big upside for them, even if they're smaller.

01:37:23 - Alex Kantrowitz (Guest)
But you have 70 something million users as opposed to the other social networks which have far more, and not only that. So it's a smaller pool. Not only that, you have to tailor make those ads for Reddit. So, like I saw, I was on the front page and I saw a Weight Watchers, for instance, had an ad and it was in the MI, the AESOL format for, like, not eating certain foods. Yeah, no, you're not, you can eat whatever.

01:37:45 - Leo Laporte (Host)
Smart that's good.

01:37:47 - Alex Kantrowitz (Guest)
And it is. It's great advertising. I think it's actually going to be more effective than Facebook. But the question is for a platform that size and for the work that's necessary to go into, the creative work that's necessary to go in and advertise there, are you going to start to get the same level of advertiser interest? It's an open question.

01:38:07 - Janko Roettgers (Guest)
I mean we all do these that smaller audiences that are dedicated our value, but that's why we do noise to that right.

01:38:13 - Mike Elgan (Guest)
Yeah. The problem with advertising is the advertising market is always fickle. If it's hot right now, Right now, it'll get soft later. And you have these radical ups and downs in your revenue model, which is problematic for a smaller site.

01:38:29 - Leo Laporte (Host)
I think it's. I think you got it right. They should say they're an AI play. I mean, really, google's search results benefit hugely from Reddit. I know a lot of people, myself included, who, if we want advice on a product, we will add site colon Redditcom to a Google search, because it's that there you're going to get some number of people who are maybe it's small, maybe it's five, who are absolute fanatics in this tiny little area, who will give you the best possible, you know, as extra virgin olive oil recommendations. It's very, very good for that kind of thing. It's a very niche thing and I think Google will be smart to put more money into it. Heaven for Fendt, they acquire it. Yeah, don't know.

01:39:16 - Mike Elgan (Guest)
Don't even put that out there.

01:39:20 - Leo Laporte (Host)
Yeah. But I think they should say I think you're right. They should say we are an AI company. We are, we are creating content for AI to ingest because it will make AI better. It's actually absolutely not a bad place to be, yeah.

01:39:33 - Alex Kantrowitz (Guest)
They should for their their road show that they go on before they list public. Should just like go up there and be like a chat GPT, please listen. They should just repeat Sam Altman, sam Altman, sam Altman, 8%, 8%, 8% and then say thank you for coming. We hope to see you on IPO day.

01:39:52 - Leo Laporte (Host)
Yeah, by the way, it's not a given that they will go public, right that? That's still up in the air. They just they're preparing a possible IPO.

01:40:01 - Alex Kantrowitz (Guest)
No, it seems like it will happen. I mean, they're talking about the potential list as soon as next month, but it's also possible that somebody might come in and acquire them.

01:40:12 - Leo Laporte (Host)
Did Reddit suffer from the June rebellion of moderators? I thought it might be the end of the line for Reddit, but there was no good alternative for people to run to.

01:40:23 - Alex Kantrowitz (Guest)
Yeah had there been an alternative it could have been.

01:40:25 - Leo Laporte (Host)
They survived it. I think yes, yeah.

01:40:29 - Janko Roettgers (Guest)
And, interestingly, they're now linking it to the AI training right. They're basically saying we had to do this to block all these companies from accessing our service for free. So now we can charge Google $60 some million a year for this.

01:40:43 - Leo Laporte (Host)
Right. It is not tenable when you have 76,000 active, 76,000,000, right Active daily users and 60,000 daily moderators. That's a that is. I was going to mention that Jack Dorsey saw this as a problem and Twitter is why he funded Blue Sky. He really felt like you can't have a centralized company running something like this. There's too much interference to run with governments, with censorship, moderation issues. It's better to let it be diversified. And yet there's a certain value to having a centralized place. Twitter seems to have suffered by by I don't know what, I don't know by mismanagement. I don't buy everything that Elon has done from the moment he took over.

01:41:40 - Alex Kantrowitz (Guest)
My hot take is that Twitter is going to have a comeback. Yeah, I mean, maybe not a revenue comeback.

01:41:45 - Leo Laporte (Host)
They didn't die, did they? We thought they'd be dead by now. They didn't die.

01:41:48 - Alex Kantrowitz (Guest)
The product does. Actually, the product suffered tremendously in the early stages of Elon Musk's ownership, but it has gotten better recently. I think there's some damage that they've done that they cannot undo. For instance, as a user, I hate not being able to tell, like, whether those verified checkmarks mean anything and whether it's a very actual journalist or an actual public figure telling me something. But the algorithm's gotten better. We'll say it's definitely more entertaining than it had been in the past, like it used to be in the initial days of Elon's rule that it was just chat, gpt influencers and other type of crypto people, but it's much more relevant now, so there is hope Maybe not a lot.

01:42:29 - Leo Laporte (Host)
You know what I would love to see Twitter survive. I think there was a there's that we've lost. We really lost something. That was pretty amazing.

01:42:39 - Alex Kantrowitz (Guest)
And we're having a stock market comeback, right, we have the all-time highs in the S&P 500. And maybe Elon Musk says, listen, I've done what I needed to do, I've fixed it, quote, unquote, and sends it off to the public market again and just gets out.

01:42:55 - Leo Laporte (Host)
It is ironic that the thing that Elon complained most about was bots and that Twitter is now far worse when it comes to bots than ever before.

01:43:05 - Alex Kantrowitz (Guest)
I mean anytime you post a tweet. Now you get a someone replying you underneath. You know party parts in bio. Yeah.

01:43:14 - Leo Laporte (Host)
It's bizarre. All right, Little break more to come with our fabulous panel Alex Cantrowitz big technology podcast, Did you? You know, so many of our best hosts are people like all three of you, who've kind of worked for big media and then moved on and done your own thing, and I guess that's true of me too. Kind of think of it. Well, how long ago did you decide to start big technology?

01:43:44 - Alex Kantrowitz (Guest)
So I started it in May 2023.

01:43:47 - Leo Laporte (Host)
I think you were on the show right.

01:43:49 - Alex Kantrowitz (Guest)
When you started it Basically, just as I as I started. Yeah, that that's right and it is. It has been a joy. So I I feel like here's the progression. I needed to be in institutional media to kind of learn right report on how this machine works and what editors and readers look for in a story. Like I couldn't have done this right off the bat, but I'm so happy to be independent now because it gives me the opportunity, like knowing what I know, to say things I probably wouldn't be able to say if I was, you know, a standard news reporter in a newsroom, whether it's feed or the Washington Post or whatever it might be.

01:44:22
And I also think, like we were talking about before last break, just the connection that you can make to your reader is so different, because in a newsroom you're just filling in kind of where the newsroom needs coverage, like at Buzzfeed in the pandemic. You know they were like all right, why don't you write about, like people getting refunds from Airbnb's and Verbo, which is an important story for a newsroom, but it gives you no, it gives a reader or a listener no sense as to like what I care about in the world. It's just okay, I had to do it, like obviously it wasn't the most pressing issue for me and like being able to do it now is so great because there is this sense of like I'll put things out there. People understand my viewpoint, you know, they email me or they, you know, give me feedback on social media, and then we were in constant conversation and it's allowed me to find a voice as well. So I love doing it. I think that's a great way of describing it.

01:45:14 - Leo Laporte (Host)
I feel bad for younger people who don't have the chance to work for institutionalized media because it's going away very rapidly, and have to kind of leap in on their own and never really do get that, that training, those training wheels which I think are absolutely like.

01:45:31 - Alex Kantrowitz (Guest)
Without Buzzfeed I wouldn't be doing this. Yeah, announcing Buzzfeed news is gone, vice is gone, and, for all their faults and there were plenty of them I still enjoyed reading them and you know the information that they shared was great and they did have. They were this on ramp, because the thing is, when I broke into journalism, the traditional on ramp was you, which was that you went to a local paper and then you went to a city paper, then you went state and then you tried to go national. That was already gone. So the other option was you go to a digital first type of company like a Buzzfeed. You learn how to report there and then you find your way and now that's gone.

01:46:12
So you're right, leo, it is. It's quite sad, but I do think that these options like doing being able to do a podcast or being able to go on sub-stack for people that really know what they're doing, like I started doing, doing my stuff part time as a freelancer before I decided to go full time, and I think there are going to be people that will be breaking in through these channels, because digital media today lets anybody who has something insightful to say cut through the noise Like that stuff will travel and people that find it traveling, I think, will end up making it. So it's just going to be a little bit different than it has been previously.

01:46:45 - Leo Laporte (Host)
You. It's been fun watching you because you really have the passion and the drive drive I lost years ago to make something great.

01:46:53 - Janko Roettgers (Guest)
So it's fun to know.

01:46:56 - Leo Laporte (Host)
Janko, you've pretty new at doing this on your own.

01:46:58 - Janko Roettgers (Guest)
Yeah, I mean yeah, relatively yes and no, because I used to freelance for a long time, so that sort of is comparable. Obviously it's not speaking to an audience directly, but it, like the flexibility on the entrepreneurial part of it, is sort of the same. I did write for I worked for a gig on wherever you could go.

01:47:19 - Leo Laporte (Host)
That's where I became aware of you, yeah.

01:47:20 - Janko Roettgers (Guest)
How many years were you at?

01:47:21 - Leo Laporte (Host)
gig on.

01:47:22 - Janko Roettgers (Guest)
Oh, a long time. It's sort of hard to say because I started for them also as part time freelancer, but yeah, I started there. Basically was my first English language outlet I wrote for.

01:47:33 - Leo Laporte (Host)
Yeah, well, we're so glad to have you in lowpasscc. I hope it's a great success. And, mike, you're kind of the original on this, starting with digital becoming a digital nomad, right.

01:47:49 - Mike Elgan (Guest)
And also, by the way, mike's list, which is currently on Substack. I launched it in a year 2000. Wow, I believe it. Wow, that's a long time and that replaced a newsletter that I had in the 90s, that I launched in 1993.

01:48:02 - Leo Laporte (Host)
Oh look, you got my new glasses on the front page.

01:48:05 - Mike Elgan (Guest)
That's right and anyway. So it's been really gratifying. What I haven't done is what Alex is doing, which is full time on the newsletter, and I've really been. You know, I still want to do the columns that I'm doing elsewhere, but I'm thinking of doing like something close to two or three, four a day, rather than once every other week or something like that. I've really had a really slow and intermittent because, as you know, I'm super busy having fun and traveling and it's hard work, being a guest or no somebody's got to do it, yeah, but.

01:48:37
But Substack is so gratifying and it's so great that they handle all the annoying parts of publishing a newsletter.

01:48:45 - Leo Laporte (Host)
I used to do it the hard way, then I did it the semi hard way on MailChimp, and Substack is great I think that's the other thing that's changed, and it's certainly for Club Twit it's been a big change is that is the infrastructure that allows individuals to, to you know, build a business without having a, you know, an accountant and all that stuff, right?

01:49:06 - Mike Elgan (Guest)
Well, that's especially true for you, right? And the world of podcasting. When you started, it was like you were. You were building everything from scratch. You were digging up iron ore to build the you know computer chassis and stuff just to do the podcast. And nowadays you know these cloud services you're talking about using. Ai to show notes and stuff that, yeah, you go to the tools like what's it called Riverside. They the show notes and the transcript just happened Right, I mean we've actually been.

01:49:34 - Leo Laporte (Host)
You know our studio lease goes through the mid mid 2026. And it's pretty hard lease to get out of. Nobody's going to sublet this. But I have to really say I can't imagine after that having a studio. It's a huge overhead that really is no longer necessary. We needed it back in the day. We needed the tricaster and all the different technologies. We had to invent the Skype, osaurus and all that stuff. You don't need that anymore. You can do it all easily. Anyway, mike's other business, which he has interestingly managed not to plug, is Gastronomad immersive.

01:50:15 - Mike Elgan (Guest)
To plug.

01:50:16 - Leo Laporte (Host)
Gastronomad travel experiences. I've been on a gastronomad trip, the amazing Wahaka adventure. I cannot recommend them more highly. Small group adventures with Mike and his wife Amira, who is an expert chef who makes all the contacts in an area like the Venice Gastronomad Prosecuta Hills, which is coming up with a Morocco experience I heard I talked to Julia and Charlie about the El Salvador experience you did. They were over the moon on this one.

01:50:51 - Mike Elgan (Guest)
That was the first one we ever did. We just did it in January and it was really really fantastic. And in fact, we're seeing with a couple of friends here, dylan and Jessica, here in Tasmania, and we met them on another Wahaka experience and became friends and they invited us come to Australia. But yeah, it's just the best thing ever and it's becoming more and more valuable to people because tourism is becoming so blown out and over-touristic and people are fed up with tourism and that is sort of like the Airbnb experiences thing is kind of like everybody's been there, done that kind of thing.

01:51:29
What we do is totally the opposite of tourism. So cool. We go into people's homes. Our friends are chefs and milling makers and it's really really, really unique, and so anybody wants it, and I would. By the way, we're talking about media. If anybody is interested in this whole concept, I would subscribe to our newsletter, which is free. On the site gastronomadnet there's newsletter tab and just sign up for a newsletter. We'll keep you posted on everything we do and we'll give you tips and advice and ideas for travel, and so I would recommend that.

01:52:01 - Leo Laporte (Host)
So cool and it was one of the best trips I've ever taken with you in America. And yeah, thank you About three other couples. Four other couples.

01:52:10 - Mike Elgan (Guest)
It was so amazing, so incredible and you can put away the mezcal Leo, I was impressed.

01:52:18 - Leo Laporte (Host)
My favorite moment and you took the picture was that Halloween Was that day of the dead. It was day of the dead, but was it that day? You know what I'm talking about.

01:52:33 - Mike Elgan (Guest)
Oh, yes, yes, it was day of the dead when you got your soul purified by a shower of mezcal coming from the mouth of somebody who purifies souls, a mescalito Somebody it was an amazing and very cold spray of mezcal, it really has a temperature.

01:52:55
So just to explain what was going on, it was really special day. It was day of the dead. The government had banned visits to cemeteries Because of COVID. Because of COVID this was what was it two years ago or something? And so we did. We went to one of the most exquisite mezcal places, where a friend of ours makes really amazing mezcal, and they purified our soul through a ceremony that involved lots of herbs and mezcal, and then we had a great dinner with huge band and then we went to a cemetery. How did we do that when it was illegal? Because it was a Zapotec cemetery. The Zapotec people don't consider themselves part of Mexico. They don't accept the authority of Mexican government over them and they said they were never conquered by the Spanish or the Mexicans. And they didn't close their ceremonies and we were invited to come to the ceremony in the Zapotec town. Incredible experience. That was one of the greatest experiences of my life personally and it was really a phenomenal day. But that was, in fact, day of the dead.

01:53:57 - Leo Laporte (Host)
Yeah, quite amazing. I can't find the picture of me getting spritzed by Ms Cal, but probably just as well. But here's some of the cemetery images. I had just gotten a Pixel 6. So I was trying the dark, the night images. That's the original, that's the Pixel 6 fixing it. Pretty amazing experience.

01:54:19 - Janko Roettgers (Guest)
Yeah.

01:54:19 - Leo Laporte (Host)
Thanks very much to you and Amira. Thank you, yeah, we had an amazing time and it's the last time I'll ever get spritzed down the neck with Ms Cal. I think Never seen, never. He also whipped me with herbs.

01:54:36 - Mike Elgan (Guest)
You didn't say that's the last time you're going to get whipped by somebody.

01:54:38 - Leo Laporte (Host)
Not the last time we're going to whip with herbs, that might happen again. The food was also very good. Thank you, mike, our show today brought to you by Stampscom. Stampscom streamlines all your mailing and shipping to turbocharge your operational efficiencies. It's really like Stampscom is like having a post office in your pocket. You could stay on top of things even on the go. They've been a partner of ours since 2012. We've been using them before then, but they started advertising on the show in 2012. 12 years ago.

01:55:12
You've got to have heard the ads. If you haven't tried them yet, what are you waiting for? We love stampscom. You know, if I ever need a postage stamp, I just go to Debbie and I say print me some stamps. She fires up stampscom and, on our regular printer, prints up some forever stamps and I've got them. It's amazing. You can print right on an envelope. You can print package labels. Postage rates keep going up. They just went up again. But you can save with stampscom the best discounts in the industry up to 89% off the US Postal Service oh and, by the way, they do UPS as well and they have equal discounts there. They automatically tell you your best, cheapest, fastest shipping options. All you need is a computer and a printer, and if you need a package pickup, you just schedule it through your dashboard.

01:56:02
I love stampscom. You will too. We've got an amazing offer for you. Keep your mailing and shipping moving at the speed of your business with stampscom, when you use the promo code TWIT to sign up. It's easy. Go to the website stampscom, click in the upper right hand corner there's a microphone. Click that. Enter TWIT. You'll get a four week trial. You'll get free postage, a free digital scale and no long-term commitments or contracts. It's really there's no reason not to do it. Go to stampscom, click the microphone at the top of the page, enter the code TWIT TWIT. I guess it's only fair, since you took the picture, mike.

01:56:43 - Mike Elgan (Guest)
Yeah, it's in a Google Photos album. That from the experience.

01:56:46 - Leo Laporte (Host)
Oh, that's right, that's where it is here, I'll pull it up for you.

01:56:51 - Mike Elgan (Guest)
And this is why your soul is pure.

01:56:55 - Leo Laporte (Host)
You got to write down the neck kids.

01:56:58 - Mike Elgan (Guest)
Yeah.

01:56:58 - Leo Laporte (Host)
By the way, my wife and she will never Lisa will never actually forget this Thinks I'm an idiot for agreeing to that. She said you know I'm good and that's why her soul is so impure.

01:57:13 - Alex Kantrowitz (Guest)
She did. Did you get to drink any of that mezcal?

01:57:16 - Leo Laporte (Host)
Oh, good Lord, Mike, did we get to drink any mezcal?

01:57:21 - Mike Elgan (Guest)
Yeah, pretty much, at least two or three times a day. That was so. We did some mezcal tasting, we did a lot of stuff.

01:57:27 - Leo Laporte (Host)
You couldn't go anywhere, they would. They push it on you because we'll. Haka is the land of mezcal.

01:57:33 - Mike Elgan (Guest)
And you can't. It's basically like tea in the Middle East you sit down, you come over to somebody's house and you're going to get mezcal, mezcal, yeah.

01:57:41 - Leo Laporte (Host)
My favorite thing. That wasn't the mezcal, it was the pulque, the precursor. Okay, yeah To a mezcal, which is a fermented they ferment fruit. I don't know if it's a great for men agave.

01:57:50 - Janko Roettgers (Guest)
It's the same stuff they did. It is the precursor. They prevent it rather than distill.

01:57:54 - Mike Elgan (Guest)
Exactly that's. You know. A pulque is, you know, three, four thousand years old. Mascal is about 300 years old and, to answer your question, Alex, sometimes they do put an agave larva, which we call a worm, in in mezcal, but there's 100 other things they put in it as well, including scorpions and extreme cases. No, but most of the time it doesn't have any of that stuff. Yeah, it doesn't have any of that stuff. The fact of mezcal that makes it so interesting is that they're literally like 50,000 different varieties, based on the different variations of methodology, what you add to it, how you age it, how you, how you cook it, all that kind of stuff. So it's, it's really amazing world. It's an amazing world to explore and we certainly did.

01:58:40 - Leo Laporte (Host)
So much fun. We went to a mezcal several mezcal distilleries, In fact. The place I got squirted also made a mezcal. Yeah, yeah, and they, they. It was the old old way of doing it. It was really fascinating, really interesting. Yeah, we had a great week this week on Twitter, In fact. It was so good we decided to make a highlight reel for those of you who might have missed any of the fun Watch. Oh wait, a minute. You were talking about volumetric lighting.

01:59:05 - Mike Elgan (Guest)
I have the smoker alarms. Here we go From the dead forward shot. It looks like you're really, really angry.

01:59:14 - Leo Laporte (Host)
Previously on Twitter Mac break weekly. What about that thing Mark Zuckerberg said about the meta quest being the best VR helmet out there Period? Trust me, he's really thrilled that Apple is in this market.

01:59:27 - WOT Promo (Announcement)
It lets him sell people on the quest three as a much more affordable alternative to the vision pro. It validates him in so many ways.

01:59:35 - Mike Elgan (Guest)
Coming up on hands on windows. We're going to take a look at AI PCs. These are brand new windows computers that have an MP built in and can hardware accelerate AI tasks Security now.

01:59:47 - WOT Promo (Announcement)
This has got to be the most under reported event that I can remember. The whole Internet was at risk for the last couple months. What? And a group of researchers silently fixed it. What? Windows weekly.

02:00:05 - Mike Elgan (Guest)
Open AI down something called Sora Holy cow.

02:00:08 - Leo Laporte (Host)
Yeah, and these are all from text prompts. That's what's amazing People say things and they sound like exaggerations.

02:00:13 - Mike Elgan (Guest)
You know, people will talk.

02:00:14 - Leo Laporte (Host)
This is as big as the Internet, I think this is this one is this one is biggest this week in Google.

02:00:20 - WOT Promo (Announcement)
Did you guys experience the chat? Gbt hallucinates Asian I didn't.

02:00:24 - Leo Laporte (Host)
I must have used it last night at the right times. Here's a couple of examples of notes.

02:00:29 - WOT Promo (Announcement)
K-type requires and bitter lamp are a page Punto. Let me encyclope Me. See my face emoji the biggest to chat Gbt. This makes just as much sense as anything else.

02:00:41 - Leo Laporte (Host)
That's true. Tweet. Bring your brain, we'll do the rest. Good Thorpe, and.

02:00:46 - WOT Promo (Announcement)
Viadon, good Thorpe, and Viadon I want more sign off specifically in that style.

02:00:56 - Leo Laporte (Host)
By the way, that was Molly White, the creator of Web 3 is going just great. Who's our guest on this week in Google last Wednesday? She is wonderful and we will have her back. You might have heard what Steve said on security. Now you should listen to the episode from Tuesday. He was referring to get this a 24 year old. It's been around for 24 years. Security vulnerability, called key trap, which was such a trivial, simple bug. A single malicious packet could literally stall an entire DNS server Enough packets. You could have had bad guys discovered this, taking the entire internet down. Fortunately it's been patched. But it was a fundamental design flaw in the DNS system and this team from Germany, athene, discovered it. They called it key trap, they gave it a CVE and we're able to get it fixed. So but 34% of the DNS servers in North America alone use DNS sec. That means that they one third of all the servers were vulnerable. It has. There was apparently nobody had discovered it until Athene discovered it, or a scene and a bullet dodged, as Steve said. Quite the story.

02:02:23 - Mike Elgan (Guest)
I wonder what else is lurking out there.

02:02:24 - Leo Laporte (Host)
Well, 24 years. Exactly, we're hanging by a thread, folks. All thing could go at any moment. Speaking of bugs, microsoft has fixed the browser and I put this in air quotes bug that edge had that was stealing Chrome tabs and data for months. If you had edge running, it would just, you know, as a convenience to you, import all of your Chrome tabs and data. People have been complaining about this for months on Reddit and elsewhere. Microsoft finally acknowledged oh yeah, sorry, so they've. They've turned that feature off, thank goodness. Yeah, it's yeah.

02:03:12 - Janko Roettgers (Guest)
As they were expanding their network. It was suddenly, was it function that malfunctioned these things happen.

02:03:22 - Leo Laporte (Host)
Mistakes happen, as you know. Speaking of mistakes, I don't know if it's willful, a mistake or or Mike Gallagher, a member of Congress, is wrong, but according to the Wall Street Journal, elon Musk could be withholding satellite internet in Taiwan. The Star Shield service, which is a service that SpaceX has offered the Department of Defense for for DOD use, designed to provide secure satellite communications and other services for national security customers, apparently doesn't work in Taiwan, at least that's according to Mike Gallagher, who is a Republican Congressman from Wisconsin wrote a letter to Elon saying hey, what gives here? It's inactive near and in Taiwan. Now, let's remember that Elon has business with mainland China, which has always said Taiwan is not separate from China. I think something like 25% of 22% of Tesla's revenue comes from China. I guess the implication is that Elon, in the sensitivity to Chinese feelings, has decided not to provide.

02:04:38 - Mike Elgan (Guest)
Or yeah, I mean, we don't know for sure.

02:04:40
I mean, it's an election year and oftentimes members of Congress will come up with something that plays well with the base, that has the right amount of xenophobia and and fear mongering, and turn out later to be something other than what they claimed. Still, it's also this is the big fear that we had with Elon Musk when he was in the process of acquiring Twitter, which is that, you know, in order to sell Tesla's in China, will he censor for the Chinese Communist Party? Will he do this? Will he do that? And that's one of the problems with somebody with as much power as Elon Musk, who has interest, financial interest, in China. To what extent will he? Will he help them, you know, prevent Taiwanese, for example, from getting internet service? We don't know yet, but I would I would take this with a grain of salt until we know for sure what, what's happening exactly, but it, but this is a fear with with Elon Musk. He's got too much power and too many financial interests in China and, by the way, so does Apple.

02:05:40 - Alex Kantrowitz (Guest)
Don't we think that this is a problem for the US government also, that it's relied on such an important piece of infrastructure to be built by the private sector? Not, by the way, it's not the first time we're seeing this Like we also have this in Ukraine, where, like, basically, the United States and the the you know European coalition is effectively beholden to Elon to continue to providing internet access to the Ukrainian army and these turned it off at points. So like this is one of the things that if you're a government you run, you don't walk to build on your own. But yet instead of doing that, we have Mike Gallagher writing letters about Elon Musk. Like we need to get out of way to get this thing to not be dependent on one person.

02:06:23 - Mike Elgan (Guest)
Yeah, and it's not just. It's not just, you know, internet connectivity with, with the, you know. I mean there are many, many governments now that have space programs and the first thing you do if you have a space program is satellites. Literally 50% of the satellites orbiting the earth right now are owned by SpaceX. Yeah, wow.

02:06:43 - Leo Laporte (Host)
Wow. The FTC, you know, is getting a lot of hits for various things they're up to. I think this was one we could all agree is deserved. They have cracked down on H&R Block for deleting tax data when users want to downgrade. It's very easy to upgrade from their free tax service to a paid service. It's seamless. But the FTC says in an administrative complaint it issued yesterday, or sorry Friday, that H&R Block designed its online products to present an obstacle course of tedious challenges to consumers, pressuring them into overpaying for their products.

02:07:25 - Mike Elgan (Guest)
It's called dark patterns. Learn the language. Ftc.

02:07:29 - Leo Laporte (Host)
And you know what? Anybody who's ever used H&R Block, as I have, or I use TurboTax same thing they're. They're full of dark patterns. It's free, it's free. And then you get to oh, but you want to file them? Oh well, that'll cost you $89. It's full of that stuff. And here is we getting close to tax season.

02:07:46 - Alex Kantrowitz (Guest)
I just want to give a standing ovation to the FTC here. Ftc, the critical of you at times, but this one, you deserve a standing ovation Because these tax companies are liars and it just like they're getting away with it without any recourse. And to have the FTC take the stand and say no, you can't promise free tax prep to people if you're not giving them free tax prep. This is exactly what the agency was created for. This is what they're supposed to do. So standing ovation. Lena Khan, you don't get it all right, but this one certainly did. I will defend.

02:08:20 - Leo Laporte (Host)
Lena Khan of the death. I think she's the. She's the best FTC chairman we've had in long, long time. In my memory for sure and this is not something we didn't know about for years ProPublica five years ago did an investigation that found H&R Block was among several other tax prep services that kept Google from showing the free offerings and search results. Well, you can't search for those.

02:08:46 - Alex Kantrowitz (Guest)
I mean Turbo Turbo tax also like they. They have such a sneaky thing where they will add like an extra $40 to your, to your tax price.

02:08:55 - Leo Laporte (Host)
You wanted to fire state returns, oh well.

02:08:58 - Alex Kantrowitz (Guest)
Oh my God, I've done it a few times where, like I finished and I was like, all right, I had one price the whole time I signed up and next thing I know it's like you know almost double. And then you write to them and you're like, oh. So they're like, oh, sorry, I'll refund because it gives you nothing. But I don't even know where I go and I pay attention when I'm on the internet.

02:09:15 - Leo Laporte (Host)
Yeah, I use Turbo tax to do my mom's taxes. She's 91 and they're pretty straightforward and I I get caught every time. I almost always just say, yeah, take my $89. Yeah, it's fine, they get me every single time. I want to end on a happy note, so I will be very happy to report that the, the entire knowledge of human civilization, is now safely backed up on the lunar surface.

02:09:47
You may have been following the, the mission of this intuitive machines. I am one lunar lander, the first time we've landed on the moon since Apollo. You may remember that it was actually quite a story about the landing. They, somebody, somebody forgot to flip a switch before takeoff that would turn on the special cameras that were designed to help the lander land on the moon and as a result, it was coming in blind. Let me see if I can find this, because this was such a great software fix. It turned out that there was a package, a scientific package, already on the lander with lidar and some other stuff available to it. So they scrambled at intuitive machines and got that package working and in fact we're able to touch down on the moon. The laser rangefinder on the on the lander we're not working properly because somebody forgot to turn it on. So controllers uploaded a software patch to on the lander to use in their place to use a NASA Doppler lidar payroll. It was originally going to be a technology demonstration to help it land, and it did, except it tripped. It missed a boulder apparently and one of the legs of the lander hit the boulder and the lander landed on its side, which is why they've had a little trouble communicating. But it is communicating and many of the payroll packages will work.

02:11:30
The landing was the first on the moon by a privately developed spacecraft, the first soft landing on the moon by any American spacecraft since 1972. It was a NASA commissioned project though. So it was. It was a NASA payload. The mission carried six NASA payloads through the commercial lunar payload services program, but also on there were some non-NASA payloads. Columbia sportswear had some jackets on there, they say to test as an insulation for a propellant tank, but really I think it was an ad for Columbia sportswear.

02:12:15
The international lunar observatory association put two astronomical cameras up there. Jeff Koontz provided an artwork called moon phases. That's the part that landed face down, by the way. Well, okay, so it's there, but no one was going to be able to see it anyway, right. So now it's face down. But the most interest, by the way, there was another one, that Eagle cam, built by students at Embry Riddle Aeronautical University. Eagle cam was designed to be ejected from the lander during its descent and then reaching the surface ahead of the lander so that they could get pictures of it landing. Did that work, john? Do we know? It didn't eject, it didn't eject, so no pictures. Too bad, because we'd like to see the see the sideways thing.

02:12:59
But the thing that I got my attention and this is written by a Nova speed vac who was with the arch mission foundation they sent up a lunar library made of thin sheets of nickel called nano fish that are practically indestructible and can withstand the harsh conditions of space. What's on those thin sheets of nickel? A backup of Wikipedia. Wikipedia, You're right, a backup of human knowledge that can now endure untouched on the moon for eternity, which might beg the question why, why? How do we edit it when things change? So on there, the Wikipedia, you got it. Six million articles, the English language Wikipedia project Gutenberg, 70,000 public domain ebooks, all the I mean it's all the good stuff. You know, the Dickens stuff, the you know Robert Lewis Stevens is all public domain novels and books, plus David Copperfield's tricks. Well, this is a little strange. David Copperfield's magic secrets, the secrets to all his greatest illusions, including how he will make the moon disappear in the near future. And I'm thinking first of all Don't tell the moon, don't tell the moon yeah, right exactly.

02:14:24
We're going to make you disappear. Please don't tell anybody. He must have paid for this right. He is a billionaire. He's. He's made a lot of money. He's not, I would say, he's arguably not the best musician, magician in the world. I would have much rather known how Penn and Teller did their tricks, yep. But he must have paid them some money for this. It's just publicity, because, except that I was thinking what, if you know a distant future, an intelligent species comes and they land on the moon and they find this thing and they realize there's something on there and they open it up and they see David Copperfield's magic secrets. What are they going to think? Is it like yeah, he must have been the leader of the of the civilization, right?

02:15:08 - Mike Elgan (Guest)
And and why was he sawing people in hell? What?

02:15:11 - Leo Laporte (Host)
the what's going on? What the hell Long now? Foundations 7,000. Record. The Rosetta project of 7,000 human languages, selections from the internet archives collection. The SETI institutes earthling project with 10,000 vocal submissions representing humanity united.

02:15:31 - Alex Kantrowitz (Guest)
Why do they have to send this to the moon? I don't know.

02:15:34 - Leo Laporte (Host)
The arch mission primer, which teaches a million concepts with images and words and five languages. I think that maybe they're thinking you know, if it all goes to hell on the earth, at least it will all be up on there on the moon, but then the only way you'd get to it is to have enough technology to go back and get it. By which time all of this?

02:15:55 - Mike Elgan (Guest)
That's basically that's kind of sort of what happened in space Odyssey, where, where there was a monolith buried on the moon, it was the technology to go find it and dig it up and then expose it to sunlight that triggered much of what happened in the movie Right. So it's kind of sort of like that Radio message to Jupiter.

02:16:16 - Leo Laporte (Host)
And then when you get to Jupiter you find out all these worlds except one are yours, right Like, don't go near us, except Europa except Europa.

02:16:26 - Alex Kantrowitz (Guest)
The aliens will get to the moon and they'll be like wow, they messed that up real good. Yeah, no wonder.

02:16:32 - Leo Laporte (Host)
No wonder they left us, no wonder there's just a smoldering field where the earth used to be.

02:16:37 - Alex Kantrowitz (Guest)
They couldn't even land this rover right. What happened?

02:16:39 - Leo Laporte (Host)
I guess they were drinking way too much Before they sent it over lying there. Oh my God, the entire civilizations records. This is the third time they've tried to do this, by the way. They failed to, but times before Right. Here's what Nova Spivak writes we have safeguarded our cultural heritage far beyond earth, ensuring that, no matter what the future holds, our history, knowledge and accomplishments and our best magic tricks will persist untarnished on the moon for eons to come.

02:17:19 - Alex Kantrowitz (Guest)
Because people don't take themselves too seriously at all.

02:17:23 - Mike Elgan (Guest)
And it's not that far removed from the earth. It's kind of right there.

02:17:28 - Leo Laporte (Host)
So, in keeping with your team of 10 and a half, Wow, you say it's right there, but it took us a long time to get there. I mean, it wasn't the easiest.

02:17:38 - Mike Elgan (Guest)
In the universe scale. The earth and the moon are essentially touching, but the good news is that we're actually at the cost of a new era of space exploration, which is really, really exciting. There was a great article in the Atlantic from Marina Koran called Apollo's sequel will be a gold rush, and I don't know if anybody's seen the TV series for all mankind, but it's starting to look like a for all mankind kind of a thing. So just briefly, for all mankind is a alternative history where the Soviets beat the Americans to the moon.

02:18:13
The space race didn't end with moon exploration, but actually heated up and a lot more money was plowed into it. And then they go into the future and to Mars exploration and everything else we're about to, and it was all public, private, a lot of public, private stuff. Some of the most advanced technology was private companies. There's an Elon Musk like character at some point who shows up with his own space program, and so is a great series. But but over the next, like five years, 10 years, it's going to look a lot more like this series where the, where there's private companies, lots of countries are going to be having space programs of their own, a lot of moon exploration, probably a permanent presence by the United States on the moon and probably a couple of other countries. Serious, you know, the US is getting serious about Mars, and so it's. It's really an exciting. I think it's going to be an exciting decade for for space enthusiasts.

02:19:05 - Janko Roettgers (Guest)
And also there's been this argument made that the moon is sort of a pit stop on the way to the Mars, right where you could like maybe stop and recharge or whatever. So you could also see the other way around. If anybody ever comes to visit us, maybe they're going to stop on the moon, first they're going to look at this rover and they're going to read all this stuff, and then they're going to turn around and maybe change that, travel a little bit and find a way around this. Who knows?

02:19:27 - Leo Laporte (Host)
We figured out how to play the nickel micro fish. And here is what it says Never going to give you up, never going to let you down, never going to turn around.

02:19:40 - Mike Elgan (Guest)
That's all it is. It's just one video.

02:19:42 - Alex Kantrowitz (Guest)
It's just it's just Rick Astley over. You see, this guy never saw them in half.

02:19:50 - Leo Laporte (Host)
Well, that's interesting. I mean, look, and it's one of the reasons we launched a show called this week in space, because I felt the same way, like that, we are on the precipice of something very exciting, it's. I mean, you and I grew up watching, you know space exploration. Um, you know, the last time we landed on the moon I was 16 years old. It's been, it's been a while, and uh and so, uh, it's fun to see it happen again. I don't know, the same arguments are going to pop up that popped up in the sixties, which is well, so what? Why we got real problems here at home? Why are we spending billions to do this? And I guess it's just. It's human nature that we want to explore and it's not. It's a small percentage of our overall expenditure.

02:20:35 - Mike Elgan (Guest)
Well, that that was actually one of the uh things that um was interesting about the show for all mankind and which I think is true, and I think it's actually been demonstrated by the Apollo program, which is that you know, for every dollar you spend in space exploration, you tend to get, you know, $30 back in in economic value the invention stuff like that. In the series. Actually, the Soviet Union never falls because they're they're spaced out through exploration funds, their country, to the point where they don't have an economic collapse. So I don't know.

02:21:12
Yeah, so it's, you know, I, I, I tend to think that if you, if you want to justify it in terms of dollars, a sense you can, it's pretty easy to do that.

02:21:20 - Leo Laporte (Host)
Well, I don't know, if that's, it's also cool.

02:21:23 - Alex Kantrowitz (Guest)
Like we should. We definitely should have something in our budget for like cool stuff. It's cool Like municipalities do fireworks. Federal government should do space exploration.

02:21:31 - Leo Laporte (Host)
It's like fireworks, it's like you're right, cool budget.

02:21:35 - Alex Kantrowitz (Guest)
It just makes you feel good.

02:21:37 - Leo Laporte (Host)
It's less than the entire amount of money we spend on pizzas every year in this country. So it's, you know, it's like that.

02:21:44 - Alex Kantrowitz (Guest)
And that pizza money is money well spent.

02:21:46 - Leo Laporte (Host)
So certainly nobody's complained about that.

02:21:49 - Alex Kantrowitz (Guest)
I'm contributing to a large percentage of that budget.

02:21:52 - Leo Laporte (Host)
Well, if your pizza's here, it's good, because we're about to wrap this sucker up. Thank you, alex, for being here. Alex Cantowitz, host of the big technology podcast, the big technology news, get letter at big technologycom, the author of always day one, and boy, the guests you have on your show spectacular. I will definitely be listening to the NVIDIA show, thank you. Yeah, that's coming up Wednesday.

02:22:15 - Alex Kantrowitz (Guest)
We, we, definitely. I'm very stoked about our lineup. We have, yeah, nvidia show coming up and then Ryan Peterson from flex ports coming on. We have, yeah, lots, lots of good stuff, nice.

02:22:27 - Leo Laporte (Host)
Nice, all at big technologycom Always a pleasure to have you on. Thank you.

02:22:32 - Alex Kantrowitz (Guest)
Great being on Leo, great talking with you.

02:22:34 - Leo Laporte (Host)
I wish I'm going to have to listen to that Den's interview. As a big fan of force square, he has this. I'll be curious what he's up to.

02:22:41 - Alex Kantrowitz (Guest)
He has this very cool new app that he's working on called BeBot, and basically it's a. He thinks audio is the next audio augmented reality platform.

02:22:52 - Leo Laporte (Host)
So the co-founder and former. Sorry, I started playing the podcast. Yeah, I'll wait till I get home. I'm sorry, it's location.

02:22:58 - Alex Kantrowitz (Guest)
I love that, leo. By the way, that's what we want. Thank you for that, I can't wait to hear it. Bebot is the location aware audio notification so you can be walking through a city listening to a podcast.

02:23:11 - Leo Laporte (Host)
I love that.

02:23:11 - Alex Kantrowitz (Guest)
And then, for like 10 seconds, BeBot will come in and be like hey, by the way, there's an event going on, you know, to your right, like tomorrow night.

02:23:18 - Leo Laporte (Host)
Lisa and I did that in Rome. We downloaded our guest was on. Our funnel was an app that when you walked by something would say, oh, you see that, and it would tell you what it was. And then you kept walking. It gets silent, and then you keep walking and everyone's like, oh, you know that thing on your right that's a thousand years old or whatever. It was better than having a guide, it was wonderful.

02:23:40 - Alex Kantrowitz (Guest)
So I'm with you, it's so cool, yeah. And imagine it tells you like oh, the restaurant you walk by, the chef's special is, you know, abc or something like that. So just have this contextually aware buddy with you as you go through cities. I think it's awesome.

02:23:54 - Leo Laporte (Host)
I think that's what my new space glasses are going to do.

02:23:59 - Alex Kantrowitz (Guest)
Absolutely.

02:24:00 - Leo Laporte (Host)
I can't wait. It's your fault, Mike. I can't believe I bought those. What do you welcome? I held up on the. Ray Vans and I ended up getting the brilliant labs frames, and I got the geekiest ones too. What's the thing on the back? Is that a speaker?

02:24:17 - Mike Elgan (Guest)
I think those are batteries or something. I don't know what those are. Who?

02:24:20 - Leo Laporte (Host)
knows, I'm wearing them to work the minute they come. Mike's list now for 22 years Amazing, oh 24 years. And yeah.

02:24:32 - Mike Elgan (Guest)
Yes, and thank you all for subscribing to my newsletter. If you want to check it out. And also one quick plug, check out hellochatterboxcom. I don't want to belabor the point, but it's an awesome AI education. We talk a lot about AI and now it's they're teaching students prompt engineering as well as everything else.

02:24:51 - Leo Laporte (Host)
This is Mike's son, kevin, who is so brilliant and came up with a wonderful thing for schools or even for parents at home who want to help their kids understand technology have some fun building a talking computer and programming it so that you understand. You know, as we get into this AI world, having a kid know the difference between an AI and a, you know, a human in a box is very important, and this is a great way to give them that kind of literacy not just literacy and coding, but AI literacy. So hellochatterbox.

02:25:22 - Mike Elgan (Guest)
You're the only smart speaker that's legal in school, that's copic compliant. So if you're an educator, yeah, this is the only one you can use and I would recommend it highly.

02:25:33 - Leo Laporte (Host)
Very nice. If you have a Raspberry Pi, you can. You can get this, the everything else, and if you don't, they also sell a Raspberry Pi in a version as well. Hellochatterboxcom. I always like to give Kevin that plug. I think that's a really great thing, thank you. And don't forget gastronomadnet, and go on one of these wonderful trips that Mike and Amira put together. They are incredible. What's the next one?

02:25:56 - Mike Elgan (Guest)
The next one is Mexico City in about three weeks or so, and that's going to be super fun, fun, fun fun.

02:26:03 - Leo Laporte (Host)
Yanko Rettgers is new newsletter lowpasscc and you were very kind to put a little special deal in our club Twitter discord. Thank you for doing that, yanko.

02:26:17 - Janko Roettgers (Guest)
Sure thing, we want to check it out. The newsletter is all about AR VR streaming. Every week I have a free and a paid tier, and if people want to check out the paid tier for free for a month, the coupon is in the top.

02:26:28 - Leo Laporte (Host)
Very nice Filtering the future. We also can see you on Fast Company and elsewhere. Always a pleasure to have you on.

02:26:36 - Janko Roettgers (Guest)
I was happy to be here.

02:26:37 - Leo Laporte (Host)
Thank you, Yanko. Thank you to all of you who joined us, especially our club Twitter members, who make this show possible. If you're not a member, Twittertv slash club Twitter, please join the fun, join the community, support what we're doing. The show is available still for free. Ad supported. Yeah, we still got some advertisers. Thank you to all of our advertisers. You can get a copy at twittv after the fact. You can also go to YouTube.

02:27:03
There's a full time YouTube channel dedicated to this week in tech. The video of it you can subscribe in your favorite podcast client. Just search for twit. All of our shows are available. That way, you can even watch us do it live. We stream the show while we're doing it, like right now at YouTube, youtubecom slash twit. We do that for all of our shows and if you were in the club, of course, you can watch the live stream before and after the show as well. Thanks for being here. Thanks to Benito Gonzalez, our producer, who does such a great job, chimes in from time to time. You're glad you're not working at Twitch anymore, right. Sure, yeah, okay.

02:27:42
Thanks to John Jamerby Slanina, our studio manager, who got the smoke machine working and then it broke. That was it. That was his last act. He's got a Burke, a Burke's fixing it. Well, not yet We'll see. He's, he's got. He's a. That's Burke McQuinn. He's, he's our repair monkey. What is he? What do we call it? What's his official title? Studio engineer. Thank you, much nicer repair monkey, what. He is very good at fixing things.

02:28:13
Thanks to all of you for being here. Oh, he says now it may not be fixable. Well, how are we going to have fog in the studio if you don't fix the smoke machine? It was fun while it lasted, I guess. Yeah, we got it for the New Year's Eve shows, but I think a little volumetric lighting is good on every podcast.

02:28:32
We do Twitter every Sunday from 2pm to 5pm Pacific time, that's five to eight Eastern. God knows what it is in Australia, something like 9am Australia time, 2200 UTC. You could do the math. So watch us live if you want on YouTube, or get the show after the fact. Either way, we're glad you were here. We hope we'll see you next time and, as I have been saying now for 19 years, be 19 years in April, holy cow, another twit is in the can. Bye-bye. What, not 19 years? I was really on the first. Viet on oh good, okay, we're going to do it one more time. And as we've been saying for 19 years hard to believe, since 2005 when we started this thing, good sort and Viet on.