This Week in Tech 979 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 Twit. This Week in Tech, paris Martineau is here from the information, and this Week in Grandpa's Also, sam Aboul-Samet, our car guy. He'll talk about what's going on at Tesla From Spain. Mike Elgin is visiting. We are going to talk a lot about AI both the announcement tomorrow from OpenAI what will they be announcing? Apple's deal with OpenAI. Google's announcements coming up on Tuesday at Google IO. We'll also talk about Elon Musk and his attack on Signal Is it justified? Tiktok's ban is it justified? And should Apple apologize for its latest ad? All of that and more coming up next on TWIT.

00:46 - Sam Abuelsamid (Guest)
Podcasts you love.

00:48 - Leo Laporte (Host)
From people you trust. This is TWIT. This is TWIT this Week in Tech, episode 979, recorded Sunday, may 12th 2024. Mustache, it's time for Twit this Week in Tech, the show where we cover the week's tech news With a rotating panel of super smart people, and you will know all three names on this panel. Paris Martineau is here normally on this Week in Google. She's been on Twitter many times. Reporter at the Information, where she just published a great article all about BuzzFeed and Jonah Peretti interview and really good stuff, it's true. Nice to see you, paris, on a weekend Nice to see you.

01:39 - Paris Martineau (Guest)
Again I was going to say it was weird listening to the intro. I was like, ah, this week in Google and I was like, nope, this week in tech.

01:45 - Leo Laporte (Host)
I'm glad we can get you on, because you're not going to be on this week in Google this week because we moved it because of Google IOS and you can't do Tuesday.

01:51 - Paris Martineau (Guest)
I know so this way. So I was like I am getting. I was like I need my weekly dose of Leo Laporte, exactly.

01:59 - Leo Laporte (Host)
Well, sounds like there's something wrong with it. But okay, thank you, here's your dose. Anyway, wonderful to have you, paris, also with us. Sam Abul Sam and my car guy, principal researcher. It's so funny because I know it's GuideHouse Insights. But I looked up to the lower third and I said it's not on the lower third, but I know it's GuideHouse Insights. It's in my mind. Also the host of wheel bearings, a great podcast at wheel bearingsmedia. Hi sam, good to have leo again. Uh, second time today he saw the northern lights at lake, did lake? Uh, jealous, pickle, pickle barrel lake, pickerel, pickerel lake. I should remember that first fish.

02:42
I ever caught was a pickerel, a 19-inch pickerel. They're not good to eat, but they are good to catch. Welcome, sam Mike Elgin is also here. He is in Sechis, in the beautiful Spanish Riviera, right just down below the armpit of Barcelona. No, that doesn't put that right. Actually that's the wrong way, something like that.

03:04 - Mike Elgan (Guest)
I love.

03:04 - Leo Laporte (Host)
Barcelona. It's the nipple of Barcelona. No, that doesn't put that. It is Right. Actually, that's the wrong way. Something like that. I love Barcelona.

03:07 - Mike Elgan (Guest)
It's the nipple of Spain, or something like that. Oh, it's beautiful here. Last night we actually rented a sailboat and we're sailing around the harbor at sunset which, by the way, is like 9, 15 pm around here but yeah, it's beautiful here.

03:27 - Leo Laporte (Host)
It's such a beautiful place.

03:28 - Mike Elgan (Guest)
I know you did not see the Northern Lights in Sechis. No, did not, did not. But I saw them on the internet. Yeah, holy cow Paris.

03:35 - Leo Laporte (Host)
did you go out and see if you could see anything in New York?

03:38 - Paris Martineau (Guest)
No, you couldn't really see anything. It was cloudy here in New York. Yeah, it was very frustrating. I saw a lot of great photos, though, and I was very jealous I, you know, I've always thought I'll never get to see.

03:49 - Leo Laporte (Host)
I always wanted to and I'll never get to see the northern lights and then and then you could see him in petaluma, but I still didn't see him you just missed him or you were asleep you have to go outside oh that's true.

04:02 - Sam Abuelsamid (Guest)
That's the problem after dark, uh I true, that's the problem After dark.

04:06 - Leo Laporte (Host)
Here's the problem. This is what it looks like. This is Santa Rosa. It's very nearby. It could just be a normal kind of pinkish night sky. I don't know if I could tell the difference. I think maybe I saw them.

04:19 - Sam Abuelsamid (Guest)
You would be able to tell the difference if you were out there and you could see the waves of light. Yeah, it shimmers in the sky and it's like sheets of light hanging in the sky. It's, it's, it's.

04:30 - Leo Laporte (Host)
There's nothing quite like it well, we have until uh, tuesday there, I'm told uh, because the sunstorm continues. Uh, a lot of I think a lot of people were concerned that this being one of the largest mass coronal ejections in years, like in 100 years, were concerned there would be disruptions of radio TV computing. The only thing I heard about is the GPS on combines and other automated devices might have been impinged to the farmers.

05:08 - Mike Elgan (Guest)
It was Y2K all over again. They were talking about satellites being disrupted and the Internet being killed off.

05:13 - Sam Abuelsamid (Guest)
for a while, In modern farming they've really adopted a lot of precision agriculture. So you know the tractors, the combines, all the equipment you know, and they've got machines that go out and automatically weed and you know it's all guided by GPS as it goes through the fields and goes up and down the rows. You know it's really done a lot to improve efficiency in farming because they can plant everything closer, more precisely, and go in there and then pick out. They've got equipment now that can go out and, using cameras, using machine vision, can actually pick out the weeds in between the seedlings of the plants that you actually want to grow and, using lasers, zap the weeds in between the seedlings of the plants that you actually want to grow and, using lasers, zap the weeds. I've seen the video of that.

06:10
Much more efficiently. It's pretty amazing.

06:13 - Leo Laporte (Host)
Jason Kobler writing into 404 Media the accuracy of some critical GPS navigation systems used in farming. This is the peak planting season. According to John john, a john deere dealership have been quote extremely compromised. So what?

06:28 - Paris Martineau (Guest)
you want wasn't just pretty lights in the sky, really um I mean, obviously that's not good, but I do think, you know, in a zoomed out view, it's kind of cool that we can still have things that are, like you know, magical, fun lights in the sky that disrupt all of our newfangled technology. It makes me feel small in a lovely way. Isn't that great? It's funny.

06:52 - Mike Elgan (Guest)
We go to a lot of farms, but never in the US, and you know, like John Deere's 9RX series will literally wake itself up, drive itself to the field, do whatever all the stuff that a tractor does over a massive area and then go back and then, and you know, the farmer is just like sitting there on Tik TOK all day. But we go to a lot of places, like Oaxaca, for example, that we went to, and we go to El Salvador and some other places where they don't even have a donkey to pull the, you know, the plowing Right. So the world is the divide in agriculture is, you know, Cuba, forget it, they don't even have anything, basically. And so the divide is quite vast and getting vaster. With the advanced tractors that Sam was talking about.

07:38 - Leo Laporte (Host)
We have some US farmers listen to the show. Somebody a couple of years ago sent me a picture of him in his combine which looked like a living room. I mean it really was. He was just there to kind of monitor operations. He could listen to the show, watch the show while the tractor has a wet bar and practically I was. I couldn't believe it.

07:59
So intermittent connection and accuracy problems with quote real-time kinematic systems, or rtk systems, which connect to john deere's starfire receivers which are widely used in modern tractors and agricultural equipment. Rtk systems use gps plus a stream of constantly updating correction data from a fixed point in the ground to achieve centimeter level positional accuracy for planting crops, tillinging fields, spraying fertilizer and herbicide. Apparently the solar storm, according to at least one John Deere dealership landmark implement in Kansas and Nebraska, ruined the accuracy of RTK. One farmer said and maybe this is a stock tip, I don't know that he was just sitting in his cornfield. He couldn't do anything. Kevin Kenney, a farmer in Nebraska, said no GPS, we're right in the middle of corn planting. I bet the commodity markets spike Monday. Just a little tip for you Be a good time to short corn. Or is it buy corn, short it, I don't know. This is why I'm not in the commodity market one of them I go long or short.

09:14
One of them, uh. So, wow, I mean, that's it. You know, uh, I kind of have a recurring nightmare that actually was started by me reading about emps electromagnetic pulses from um, atomic weapons fired high in the atmosphere. The destructive power would simply be the electrical pulses that it would send out that would knock out all computing, all electronics it's. It's a particularly devastating kind of warfare and I think ever since I read about that I've always kind of feared we are so dependent now on electronics. If there were something like an EMP or maybe a solar storm that was particularly aggressive, it wouldn't just be pretty lights in the sky, we would be in deep trouble, we'd have to short everything.

10:13 - Sam Abuelsamid (Guest)
In recent months there's actually been a lot of reports. There's been a lot of issues with GPS for aircraft in Europe, the Russians are jamming.

10:23 - Benito Gonzalez (Other)
They're getting a lot of jamming from the Russians, which is disrupting aviation in Europe. The Russians are jamming.

10:24 - Sam Abuelsamid (Guest)
They're getting a lot of jamming from the Russians, which is disrupting aviation in Europe. We talked about this on Security.

10:30 - Mike Elgan (Guest)
Now Steve. Gibson had a great thing on Security Now with that. Yeah, that was a really nice roundup. That was very eye-opening, that segment.

10:40 - Leo Laporte (Host)
One of the things he said is that many of the jamming devices are on the ground, so an airplane in the air could just say well, I don't take any GPS signals from below, only from above, and apparently you could solve it that way. But it doesn't solve it, of course, for anything on the ground, anything they can't distinguish.

10:59 - Mike Elgan (Guest)
It's part of the largest story of drone warfare and drone information warfare in the Russia-Ukraine conflict. They're pioneering the future of using a little quadrocopter to do everything. I mean, it's just amazing the things they're doing with hacking and then counter hacking, and then they're blocking the signals, and then they have drones that unblock the blocking and it's just really really fascinating case study. When the dust settles on this conflict, it's going to be really interesting to sort through what happened.

11:33 - Leo Laporte (Host)
Somebody made this point on Twitter a few weeks ago that what's really happening is you have a laboratory for alternative warfare in Ukraine. That, in the long run, is going to be a problematic for the entire world.

11:47 - Sam Abuelsamid (Guest)
Yeah, and you know, one of the one of the challenges with GNSS you know it's global navigation satellite systems, that's sort of the generic term that covers GPS, which is the US system, galileo, which is the European system, glonass, the Russian system, and the Chinese system is you've got very weak signals from the satellites that are going to these GNSS receivers and it's very easy to overpower them. I don't know if you remember about 10, 12 years ago there was a company called Lightspeed that wanted to do ground-based long-distance wireless broadband using a frequency that was directly adjacent to the frequencies that are used for GPS and the FCC ended up revoking the license that they had given them because of all the complaints from the testing that they were doing. They were getting all kinds of interference with GPS. It's actually very easy to disrupt these GNSS systems because those satellite signals are so weak.

12:53 - Leo Laporte (Host)
So even if the European Galileo comes online, it's not going to solve the problem that. It's just a problem in general.

13:03 - Sam Abuelsamid (Guest)
Yeah, a, it's a fundamental problem with the technology, the, the basic concept of.

13:10
You know, you've got this constellation of satellites and all all these other systems are online glonass and galileo, and and the china and the chinese have their system this name I've forgotten um, you know all of them work fundamentally the same way, where you've got a constellation of satellites, where the signals are time synchronized and then the receivers, you know, pick up multiple signals from multiple satellites and, based on the time difference between those signals, they can use that to calculate their.

13:40
That's how they calculate the position is the time skew between the signals. They can use that to calculate their. That's how they calculate the position is the time skew between the signals. So you know, if you're, if you've got a signal that is a few milliseconds after a signal you got from another satellite, you know you're closer to that, that other satellite than you were to the one that's further behind. So you can use that, that difference in the time skews from each of the simultaneous satellites to calculate your position. But those signals are inherently very weak. They're not strong signals and so they are easy to overpower with something else running at the same frequency.

14:22 - Leo Laporte (Host)
By the way, it's Baidu B-i-d-o-u, the chinese navigation system, and when I went to the baidu website I see this serve the world and benefit mankind. Wow, okay, maybe we should just all switch to baidu. I think that's really the way to go it's like the serve man.

14:43 - Mike Elgan (Guest)
It's a cookbook, it's a menu.

14:48 - Sam Abuelsamid (Guest)
Most modern devices, like all our smartphones, have multiple GNSF receivers in them. They can pick up GPS, galileo, glonass, so they actually use all of them simultaneously. Oh interesting, oh interesting.

15:02 - Leo Laporte (Host)
That is interesting. Oh, interesting, yeah, it is interesting, is it? Do you know if the John Deere problem is going to be chronic or if they have a quick fix?

15:12 - Sam Abuelsamid (Guest)
You know, for something like what you know what they're doing with tractors, with with agricultural, agricultural equipment. What you could do is you could set up beacons at you know, at the perimeters of a farm you know, or at various locations on the farm and essentially have you know a ground level based system that you know could potentially have stronger signals. You know you'd have to have FCC approval and one of the things that was mentioned in the articles about John Deere you know is they do they actually do have some of these already that are used for correction to give more precise location, like down to centimeter level accuracy, like Topcon apparently is one of these.

15:52
Yeah, and there are, there are systems like that. So you could, you could expand the, those ground-based beacons to, you know, to, to augment what you're getting from the satellites and provide that increased accuracy.

16:09 - Leo Laporte (Host)
Jason quotes another farmer from Reddit who says we didn't shut down, we just carried on and however the wheat went in is how it went in L-O-L. Another farmer said wow, it's really a mess out. There gonna be some stripes in the field today it's back to normal, so I guess it will self-correct. It might have just been interference. It didn't fry the equipment, in other words, but, but, but many farmers, just the larger one yeah, the larger issue is that farmers got 99 problems, and that's only one of them. Yeah, exactly yeah.

16:45 - Mike Elgan (Guest)
One more. They've got a million things going on and that's one of the minor ones.

16:49 - Leo Laporte (Host)
Yeah, farmers Jason spoke to this is again from 404mediaco seem to think the situation is getting better, even though those storms are going to continue. In the Corn Belt, May 15th is a critical date to get corn planted, but you don't want to go out there with your equipment right now. Corn Belt can't get its corn in the ground by May 15th. That's a problem. All right, we'll keep our eye on this one. I have a feeling it's a passing storm, if you will, and they'll be back to planting and there'll be a few farms that might look a little kooky, you know.

17:29 - Sam Abuelsamid (Guest)
I mean, we managed to grow corn for thousands of years before we had GPS. I think they can manage.

17:37 - Leo Laporte (Host)
But that's kind of my point is that we have become so dependent on these systems. Now I don't know if I know how to.

17:44 - Paris Martineau (Guest)
And they're at such a massive scale yeah, it's. It's hard to. It was easier to grow corn by hand without any gps or fancy equipment when it was a small plot, but these farms are giant and they have got to work on very tight deadlines and closer to home.

18:00 - Leo Laporte (Host)
I don't know if I know how to drive anymore without gps. So can we, can we, can we get this fixed? I literally I turn on gps just to get across town. These days I just it's sad. Yeah, find the bathroom. Yeah, please, nearest bathroom.

18:18
It's uh, from now next time you ask siri for the nearest bathroom, you may get a different response. We're going to talk about that in just a little bit. Hold that thought, mike, because I want to take a little break. Great panel, lovely to have. Paris Martineau, the reporter from the information. She has the information, she has the news. Also Sam Abul-Samit, our car guy I've got to ask him in a bit about tesla it's getting wacky over there and mike elgin, who is uh living right now next to one of the largest gay nude beaches in the world. So I hope you're uh yes, I can.

18:58
I have verified that fact we uh way back when my publisher, pearson, had its annual uh conference. There have authors and sellers and I went to see chase and jennifer and I wandered around and we came, came across this beach unawares, but yes, yeah, you're just.

19:20 - Mike Elgan (Guest)
You're walking along. There's a, there's a, there's a seafront. It's very nice. Did you sort of go around the bend? There's another seafront. You're like whoa, okay.

19:27 - Leo Laporte (Host)
That's a front All right Gosh.

19:31 - Mike Elgan (Guest)
There is a seafront.

19:32 - Leo Laporte (Host)
I see the front and the back.

19:34 - Sam Abuelsamid (Guest)
You've got to be ready for it and the back.

19:36 - Leo Laporte (Host)
And all the other sides and in between.

19:46 - Mike Elgan (Guest)
Anyway, wonderful to have you too. Mike Elgin, we're going to debut your brand new newsletter on the show today.

19:48 - Leo Laporte (Host)
Fantastic, I'm very excited about that. Yeah, stay tuned. Yeah, me too. If you have been a long time, as I have, been subscribing to Mike's List, you will want to know about the new newsletter. In fact, the conversation coming up is probably right up your alley, but first a word from our sponsor. Right up your alley, but first a word from our sponsor, bitwarden.

20:09
I feel like this is almost a public service announcement. If you're still making up passwords or, worse, reusing passwords, you got to have a password manager. Now I have to think that almost all of you by now have gotten the message and have a password manager. But if you don't, can I recommend the only one that I use and recommend and, furthermore, tell your friends and family, who almost certainly are not using a password manager, about Bitwarden? Open source, free forever for individuals. It lets you host your own vault if you don't trust them. I trust them, but some people say no, no, I don't want anybody to have my vault. This will check a lot of boxes for those relatives who have said I'm not going to use a password manager. It is so cost effective and it can dramatically improve your chances of staying safe online. They are expanding all the time. One of the benefits of being open source is the development is very rapid.

21:06
At Bitwarden, they recently added Argon2 as a password derivative hash to replace PBKDF2. If you want, I use it. It gives you immensely more security. They have Passkey support. They've had Passkey support for a long time. They just announced they've added Passkey support to browser extensions and mobile devices as well. This is really great. If you like Passkeys and you've been using it most of the time, it's tied to that device your phone or your computer. With Bitwarden, your Passkeys go with you. Wherever you log into Bitwarden, there's your Passkeys. Passkeys on mobile are available on iOS. They're in open beta on Android. You'll learn more about this soon. World Password Day was May 2nd.

21:49
Bitwarden surveyed 2,400 folks from the US, uk, australia, france and Germany and Japan about their password practices. I don't think this will be much of a surprise. 31% almost a third of US respondents admitted reusing passwords across sites. That's a third who admitted it. I bet it's even more right. 42% incorporate personal information birthdates, last names, birth years raising concerns about password strength and security. 58% of respondents just remember their passwords oh boy. 34% write them down down pen and paper at work, which means there's probably quite a few post-it notes and passwords stuck on the monitors or under the blotters, just like in war games, that the movie came out 25 years ago. They're still doing it.

22:40
Near nearly a quarter of respondents view their workplace security habits as risky. They know it doing it. Nearly a quarter of respondents view their workplace security habits as risky. They know it. 45% know they're storing passwords insecurely. 44% say yeah, I'm using weak credentials. This should be a terrifying thought for business owners.

22:57
It's one of the reasons we use password managers. We use Bitwarden. It's one of the reasons your business and individual users should use Bitwarden. It's one of the reasons your business and individual users should use Bitwarden. It empowers enterprises, developers and individuals to store and share sensitive data safely. With a transparent, open source approach to password management, bitwarden makes it easy for users to extend their robust security practices to everywhere, to everywhere they get online. I've been using it with passkeys everywhere. It's fantastic. Get started with Bitwarden's free trial of a Teams or Enterprise plan, or get started for free, forever across all devices as an individual user. Bitwardencom slash twit Individuals it's free. Unlimited passwords, unlimited devices. It supports hardware keys. It supports passkeys All of that is free. Tell your friends and family say there's no reason why you're not using Bitwarden. It's easy, it's free, it's fast, bitwardencom slash twit and get your boss to use it too. And if you're the boss, you better use Bitwardencom slash twit. I can.

24:05
This one is a public service announcement. Practically everybody should be using this. So, uh, interesting, uh, go round coming up tomorrow, 10 am open. Ai has an event announcing something and it's been fun to watch the rumors. I wasn't sure if we should cover it. You know, rumors extended for everything from ChatGPT5 to maybe I kind of like this a search engine powered by ChatGPT, by AI. But then Sam Altman weighs in, gets on the blower, or as we call it now, the Twitter, and says not GPT-5, not a search engine, but we're working hard on some new stuff we think people will love. It feels like magic to me. Monday, 10 am Pacific. We will cover this stream. Jeff Jarvis and I will cover this stream tomorrow. Are we going to look back on this day and say may 13th 2024 is the day the ai became sentient? Mike elgin, you're. This is part of the stuff you write about now.

25:19 - Mike Elgan (Guest)
Yes, yeah, no, no, no, we're not. Um, I I tend to think that if I had to guess, I would say that it seems to me that perplexity AI is ahead of the curve in terms of providing something super useful. I love perplexity. I use it for search.

25:34 - Leo Laporte (Host)
It's amazing.

25:35 - Mike Elgan (Guest)
Yeah, yes, and I recently became a paid user of it Me too. Yep, I would guess that they're going to do something along those lines and that SAM is not going to call it a search engine. You still shouldn't be calling it a search engine per se, even if it's exactly like Perplexity AI. So, for those who are unfamiliar with Perplexity AI, it basically does a Google search. They use a version of PageRank that they've optimized to even more greatly magnify the difference between reliable sources and unreliable sources so it's better, if you like, authoritative sources and then, after they do this search, they take the results and they sprinkle generative AI, pixie dust, all of that and give you a result, but they give you live links to the original. Let me demonstrate this. So it's like a search engine.

26:29 - Leo Laporte (Host)
I am using Perplexity Pro. You can show my screen what's the best e-bike. That's the kind of thing somebody might search for. Oh, look at this, what are your main criteria for the best e-bike? I'm going to say price and range, okay, and then it's understanding, question reading, response, searching the web. Now it quotes the sources, as you said, but it also gives you a text answer. That is a summary, I guess, of those sources. This is the kind of thing I think people do with search. This is what they want, right?

27:00 - Mike Elgan (Guest)
Right and you get the little number afterwards which is a live link to where they got that particular fact Right. So it's usable like a search engine, but also like ChatGPT, and you also get a little roundup of photos. There's also in the pro version there's a link that says generate image. So from the search, from the prompt that you already gave it, you can just click a button and get a horrible, you know dolly fake thing, a fake image? What engine is perplexity using? Is it using its own?

27:38 - Leo Laporte (Host)
They're using several I have. I'll show you. Actually, I know the answer. I shouldn't have asked you a question. I know the answer to. If you're using pro, you could choose your AI model. This one, I'm using one that they've tuned called Sonar, based on Lama, but you also have ChatGPT-4, turbo, claude 3, which is Anthropix, and they have Mistral, so you can choose your model.

27:56 - Mike Elgan (Guest)
You can also choose your image model yeah, but you can also choose where it goes. So, for example, you say, just look at Wolfram Alpha, just look at YouTube, I like that. And you can say just look at what I gave you and don't look at your training data. You can upload PDFs and other files. You can upload photographs and say what is this? With a mobile app, you can take a picture and say what is this and it'll describe it to the greatest extent possible?

28:28 - Leo Laporte (Host)
You were talking about the MetaGlasses. You finally got me to buy those MetaRay bands. They can now do that too, using Lama, you say what am?

28:34 - Paris Martineau (Guest)
I looking at.

28:36 - Leo Laporte (Host)
And I tried it on Wednesday on Twig. Remember Paris.

28:39 - Paris Martineau (Guest)
Yeah, and it worked sometimes, and other times it took quite a while. It was very funny to me how many times during the recording of Twig, the metaglasses would just randomly activate and start talking to Leo or playing music, I thought that was quite cute.

28:54 - Leo Laporte (Host)
I like that. I had a little friend on my face and I liked having a little buddy. This is. I actually had a Go ahead, go ahead, leo.

29:05 - Mike Elgan (Guest)
Okay, so I did this fun experiment where I was using Pi. Pi has a really Piai, has a really supernatural sounding voice, right, and it just leaves it open to. So does the Perplex AI app as well. When you start talking to it, it just keeps listening for you to say something and will respond to what you say, and so I have been using sometimes the Ray-Ban metaglasses with Pi, because it'll give you audio in both directions with whatever's going on in your phone right, so it's like a.

29:38
Bluetooth headset, and so I like walking around and just chit-chatting with Pi about you know, hey, what do you think about this? What do you think about that?

29:47 - Leo Laporte (Host)
You don't think that's at all pathetic, Mike.

29:51 - Mike Elgan (Guest)
It's very pathetic, it's awfully pathetic, and also I've done it in Mexico and you know people are looking at you, look at you funny he's talking to himself.

30:03 - Paris Martineau (Guest)
They could just think that you're on a phone call.

30:05 - Leo Laporte (Host)
You know, they don't have to know it's a computer or on drugs exactly.

30:08 - Mike Elgan (Guest)
But what I like, what's interesting about it in terms of a glimpse of the future, is that there's an element of agency to it. It sort of decides to bring things up, or it's sort of like that's very rare in any of the tools we have for, uh, for a virtual assistant or a chatbot to say, hey, what about this? And based on your context, based on the previous conversation, so it's kind of I think it's a little glimpse of where we're going. Here's one of the things Not particularly a good one, but so for people to understand.

30:38 - Leo Laporte (Host)
I think it's helpful to understand. I've kind of watched a lot of YouTube videos and I think I kind of understand how all this stuff works. You have there's various ways to train an AI. We've talked about neural networks in the past and expert systems. Most of these right now are what we call large language models. That's where that LLM comes from in LAMA and everything else. A large language model basically trains itself on a vast corpus of knowledge I mean the Internet right of text If it's images, it'll do the same with images and then builds up a probability database of what, basically what word follows this word, and then it kind of uses those weights to, you know, like autocorrect put together a sentence. Now that kind of makes it sound like it's not too bright, but if you get a big enough corpus and you can train it long enough, it actually does a pretty credible job of imitating human writing or or or images, as we know, because it still doesn't actually understand it has no idea what it's.

31:44
No, it's, it's, it's.

31:46
But we're increasingly mimicking it's mimicking what it's seen, right, but we're creasing at the point where that distinction is a distinction without a difference. It's like, well, yeah, but it sounds like it's human, so what do we care if it understands. There's a second stage to all this, which is called tuning. So you LLM which makes a giant model and these models sometimes can be gigabytes, often are gigabytes worth of data which is a compressed down model, and then you tune it. And so that's when you see all these different kinds of AIs based on LAMA or based on OpenAI. They're tuning it. One way of tuning it would be to be able to read x-rays, so you would then add additional information about how to read x-rays, so you have the base model tuned to do something specific. And then there's an additional part, which is the number of tokens it can consider while it's doing all of this and this is what's really interesting Right now, openai's best AI, chatgpt, has a cutoff date.

32:53
Its LLM was trained up to. I don't know what it is currently, if it was a year or so ago in the past, but it's probably more recent. But that's when. It's okay, we got it all. We're going to stop. No more LLM building. We can tune. We can do more, but we're going to basically stop, and that is a little bit limited, because you could ask it, for instance, what's the latest iPad, and it would be wrong. It would not know about the ones that were released this week. So there are ways to get around that, but what I think you're saying, Mike, is what perplexity is doing, which is take all of that and then allow it to consult the Internet and apply new information to its existing LLM.

33:40 - Mike Elgan (Guest)
And, yes, I think that's more than that Go ahead.

33:43
Yeah, well, another thing you can do. You can just grab a link. Let's start with five links. You can just grab five links and drop them into the prompt window of Perplexity AI and it will give you information based on those five links. If you take one link, it literally just references that page. It literally just references that page. It specifically mentions. It says in this page on Ars Technica the writer so-and-so said blah-biddy-blah, make the point.

34:16
So it's much more disciplined than ChatGPT at not using the vast database for facts or information. It'll use it for language, for other things presumably, but it's much more reliable. In that sense, the tuning that you're talking about, I think is much better on Perplexi AI and it's able to be narrow in where it draws the information. I've seen it hallucinate a couple times, but it's not bad, it's pretty good. You don seen it hallucinate a couple of times, but it's like, not bad, it's pretty good. You know you don't want to use it as. One of the things I like about it as a journalist is that if I get a result, I can just fact check it based on what the links they provide and also I can ask it to go get other information.

35:01
I can ask for alternative points of view. It's really great at maintaining the context, so it's pretty good in that respect, but I would absolutely, as a source of information, a thing to bounce things off of. It's better than Google Search and it's better than ChatGPT.

35:17 - Leo Laporte (Host)
I've actually been using Google's Gemini, which is quite impressive and does do both. In fact, I hooked it up to the action button on my iphone so now I can talk to. When I press the action button, I can talk to gemini and, uh, sometimes it'll talk back and sometimes it won't, so actually it's been taught. I was talking to it, so sometimes it just gives it gives a response, search response. So I know it's hooked up, but I can also say how old is Alan Alda 88 years old and for some reason it will answer those things in speech. I'm not sure why, but this is search. Right, this is Gemini AI doing search. So Google's already doing this, and for good reason. Google's terrified Open AI and others are real threats to Google's already doing this, and for good reason.

36:08 - Paris Martineau (Guest)
Google's terrified Open AI and others are real threats to Google's hegemony. What has your guys' experience been so far with Google's new feature, which is, whenever you search something, it'll have an AI-generated kind of response up at the top. And I mean, I guess, to their credit, they say in a disclaimer these are disclaimer, these are ai generated responses. They may or may not be accurate, but I've found, at least over the last week, that they've just been wildly inaccurate every time. I just completely ignore it today I was rewiring.

36:37
I think it's powered by gemini. I was like rewiring a lamp and like had to Google something related to that and it was a hundred percent inaccurate.

36:46 - Sam Abuelsamid (Guest)
It's very hit and miss Um. You know, and frankly I've found all of the AIs that I've played with really to be quite hit and miss Um. You know, when I was at NVIDIA GTC, um, a little more than a month ago, I was in a Q and a session with Jensen Wong, and you know to you know to.

37:07 - Leo Laporte (Host)
by the way, to what the way? One of the big, one of the big benefit beneficiaries of all of this is NVIDIA and Jensen Wong, because they are making the hardware that everybody has to use. They're the, they're the Levi's, they're the picks and shovel guys.

37:18 - Sam Abuelsamid (Guest)
But reinforcing. You know what you're saying about tuning. You know tuning, I think in many ways, is actually the most important part of this. And, and tuning you know tuning, I think in many ways, is actually the most important part of this. And, and something that that Jensen said was really interesting, you know he talked about how you know, as humans, you know we, we take in information and we digest it and process it and kind of link it together with the other things that we've learned before. Um, and you know if, as a human, you know if we take in the information and we go sit in a room and we just keep thinking about it and processing it and and regard and reprocessing it continuously, eventually we'll come out of that room and be completely insane. Um, you know if we, if we don't, you know if we don't have something to ground us.

38:02 - Leo Laporte (Host)
You don't want an insane AI, okay.

38:04 - Sam Abuelsamid (Guest)
I'm just saying, and I think this is the risk that we are increasingly running with AI, especially if it, you know, okay, we've trained these models on all the content that's out there, everything that's been created up to this point, but now, increasingly, we're getting junk that is being published on the internet that is generated by AIs and it is effectively like that, where you, you know these, these systems are processing and creating information and reprocessing it and just regurgitating and retraining themselves on the crap that they have already generated. And this is, I think, what is leading to a lot of the hallucination. And I think you know, I think what we're if we're not very careful. I think you know, if we use these systems selectively like Leo, what you do with your LISP, gpt, you know, or you know, have very specific tasks, have systems that are tuned to a very specific use case, with a specific corpus of data, a specific set of data. I think they can be very useful to us, but if we try, I think, if we try to rely on these things being, you know, just general purpose thing, general purpose information generators, I think we're increasingly going to find ourselves getting bad garbage, you know, because it's going to be retraining itself on the garbage that it has already generated and processing it without having that grounding in truth.

39:45
You know that that's the thing that Jensen talked about is, you know, these systems, they have to be grounded in truth. And if they're not, you know, and if we are feeding them, you know it's like you know, when we had mad cow disease in England in the early 90s, we were feeding cows material that was, you know, reprocessed cows. You know that had had already been sick and you know it led to a very bad result. And I think the same thing is potentially possible and actually probably likely with these systems. If we don't constrain, you know, any of, I think, if we try to generalize them, I think we are going to get into that kind of thing. You know, if we, if we don't have some way to ground the system in truth, Well, yeah, and so there's a.

40:34 - Mike Elgan (Guest)
There's a threat, there's a risk with with data being consumed that's generated by LLMs, but there's also a huge risk of data generated by people, because people are wrong, dumb, misguided, et cetera, sometimes as well, and so I think this is a huge opportunity for further innovation of evaluating both synthetic and human data. There's a whole realm that has to be explored there, and I think we absolutely have to do it. One of the upsides is that we've been afraid, or I've been afraid, for many, many years that we are headed for a world of the what Danny Sullivan called the one true answer, and Google is super into this still, because they know that if you have a virtual assistant, you ask a question, it's like, well, here's the answer. Well, you know, lots of important information isn't just an answer. Is the answer Well, you know, lots of important information isn't just an answer.

41:28
Some of these chat bots are actually getting very good at making it a conversation. So they give you a context, they give you pro, con, whatever, and then they give you three canned questions Would you like to know more about this? Would you like to know more about that? Or you can write in some more. So, when we're wearing AI glasses and we're talking to these things which is, I think, going to be a major use case for using chatbots for just informational retrieval we're going to want to. We don't want the one simple answer that is just selected by Google. What we want is to have a little conversation so we understand the problems, we understand the uncertainty and just have contextual information about what we're doing, and I think that seems to be where we're headed, and I think that sort of gives me a good feeling, much more so than the one true answer future that I've been dreading.

42:22 - Leo Laporte (Host)
Well, get ready, because I do think that what OpenAI announces tomorrow will have something to do with the fact that, according to Bloomberg, apple has now closed in on the deal with OpenAI to put JetGPT on an iPhone. Yes, they're talking to Google as well, and you saw, I just used Gemini on the Gemini Gemini. Gemini.

42:49
Mike will know why I call it Gemini. Gemini was what we used to call the NASA program right after Mercury. There was Mercury, then Gemini and then Apollo, but I guess it's Gemini, I guess. So if Apple puts I'm thinking, plus, by the way, okay, the Google, the Apple, the open AI thing is tomorrow. The Google thing is a day after Google IO, where Google is clearly going to announce a bunch of AI stuff, maybe even a deal with Apple. I have to think open AI. You know they had put this announcement off. I think that there's a reason. It's right the day before Google IO. And then in a few weeks we're going to hear Apple talk about AI everywhere. They used the word AI in their iPad announcement this week almost as much as they used 5G a couple of years ago. It really peppered throughout. Apple clearly is feeling the sting of being a little bit behind microsoft and google. Uh on I don't.

43:50 - Sam Abuelsamid (Guest)
I don't think apple's gonna let anybody announce anything that apple's doing until apple is ready to announce it themselves. Good point, so open.

43:56 - Leo Laporte (Host)
Ai is unlikely to say siri use the use the s word, but I I wouldn't be surprised that that's what they're gonna they'll what they will I will suspect they'll talk about is a chat bot, a uh, that you could put on your phone. That could maybe make Google better and and Siri better and and Amazon echo better. I don't know. I think that's what they're. We'll see. Something's going to happen, and the fact that Apple's getting close to a deal with open AI makes me think this is something along those lines.

44:37 - Mike Elgan (Guest)
You think that's nuts is some kind of tool that will help users use ChatGPT or some or all of their offerings in a more in a more constructive or creative way.

44:52
It may have a lot to do with the input side, where you can have where companies, public, you know, authors and so on, have more control over whether their content goes into the, into the, into the borg collective. Um, you know. Back to perplexity, ai, they have a tool that they're that's in beta test right now, where they are basically turning it into a content creation tool. You basically get a thing that's like a word processor and you, you ask, you know, you do a prompt to build an article. The AI takes a stab at it, you edit it so that they encourage human editing of this article and then you can add photos or create photos or pictures or images with AI and then you publish it on their site, publish it on their site, and so it seems to me like it's a cross between the Wikipedia and ChachiBT and TikTok and something it might even be a social type thing. So those kinds of tools, I think, are something that OpenAI probably wants to get into, because they you know everybody's using ChachiBT.

46:02
But it's a little bit of a toy A lot further. Yeah, yeah, they could go a lot further. I also you know the Dark Horse prediction. Something I've been calling for them to do is they should come out with their own glasses, Like all of these LLM chatbot companies, should have their own Ray-Ban Metaglasses type glasses. Do you think that's a?

46:26 - Paris Martineau (Guest)
great I I worry that they I I worry that these companies think that that's risky. Given the way that you know, google glass has gone the way that the the response we've seen to things like the humane pen, it seems like the bar to make a successful wearable device is much higher than the software aspect.

46:50 - Mike Elgan (Guest)
Google Glass didn't have AI and the Rabbit R1 or whatever it is, and the Humane Pen. The problem with them was that they weren't glasses. Glasses are something that are already socially acceptable to wear, are something that are already socially acceptable to wear. That's as close as you can get to the to the sense organs, uh, of a human being, you know, right next to the ears, right close enough to the mouth where you can interact with it, right over the eyes, where you can, you can get visual information, and there's a whole industry in silicon valley that we haven't really seen emerge, where they're building, you know, uh, they want to take on Luxottica.

47:25 - Leo Laporte (Host)
Who, by the way, makes those Ray-Bans? When I got my Ray-Bans, the box said Luxottica on it and I forgot that they own it.

47:31 - Mike Elgan (Guest)
Of course.

47:31 - Leo Laporte (Host)
They own every brand. They own most of us, right, exactly.

47:35 - Mike Elgan (Guest)
But there are companies in Silicon Valley that are working on the technology. Also the designs. They have like 30 designs.

47:44 - Leo Laporte (Host)
I'm waiting for my Brilliant Labs glasses to come. I ordered the Brilliant Labs glasses.

47:50 - Mike Elgan (Guest)
I was going to say, liam, you're going to look like Dexter from Dexter's Lab. I don't care if they work.

47:54 - Leo Laporte (Host)
If they do they are augmented reality.

47:56 - Mike Elgan (Guest)
They look cool. Well, I got to get. I got to. You're the type that would like Dexter.

48:02 - Leo Laporte (Host)
Yeah. I ordered them. I don't know they were supposed to come out last month. Obviously these kinds of things take longer. I just think it's going to be interesting to see what heads up can add to this. But here's Apple's problem. They released Vision Pro. They put out kind of the wrong product. Or do you think that that's a stepping stone to glasses?

48:26 - Mike Elgan (Guest)
That's a very different product. I mean, I think that you know that's a high resolution, I agree, and I think there'll be a real use case for that in the future. But that will be the. You know that'll be the desktop Mac product. The future iPhone is going to be a lot more like Ray-Ban Metaglasses, that with visuals, and, of course, Google itself is going to be almost certainly announcing their consumer AI glasses at Google IO. Oh, you think they will. Oh good, oh, I think they will. Yeah, I think they will. But I think if Ray-Ban Meta has proved anything, it's that this is a super viable category. There's another company who makes glasses that are $160 and they have like 300 different frames that are name brand frames. They use ChatGBT and they somehow do it through siri on their iphone and when you ask them a question, they're almost answering it before you finish the question.

49:30 - Leo Laporte (Host)
They are so fast that it's it's like really mind-blowing that is a little bit of an issue with the meta. As you saw, paris it's.

49:37
Sometimes I'd ask a question and then I'm just sitting there, glazed for, yes, 20 seconds waiting for an answer, and it makes me look like an idiot, frankly, because my eyes start focusing in the distance and I just sit there waiting I. I think that's one problem that we're gonna have to solve, but we'll get faster with these. Here's the story from the new york times. They have their own sources at apple. Apple, they say, will revamp siri to catch up to its chatbot competitors. Apple plans to announce that it'll bring generative AI to iPhones. This won't be till WWDC in June. Story from Trip Mechel, brian X Chen and Cade Metz. All three have been reporting on Apple for as long as I can remember. So clearly Siri has a problem and that's what they're saying. They're saying apple executives were uh. Apple fears if it fails to develop its own ai system, the iphone could become a dumb brick compared with other technology. And I think they realized last year in in some testing that siri was a dope and that they needed to do something better. They have the. They have the uh the scenario.

50:55
Apple's top software executives decided early last year that siri needed a brain transplant. The decision came after the executives craig federighi and john g andrea spent weeks, weeks, testing open ai's new chatbot, uh, and made and. It made siri look antiquated, said two people familiar with the company's work, who didn't have permission to speak publicly but did anyway. So it's been. It's been a year. It often misunderstood questions. Chatgpt, on the other hand, knew that if someone asked for weather in San Francisco and then said, what about New York? They were looking for a forecast for New York. Siri could not do that. So anyway, june 10th, we'll find out what Apple's up to. Monday, tomorrow, may 12th, 13th, we'll find out what OpenAI's up to, and Mayth 13th, we'll find out what open ai is up to. And may 14th, we'll find out what google's up to. So it's gonna be a week, it's gonna be a very big couple of weeks, yeah, to find out what's happening. Are these the glasses you were talking about? The solos mike somebody put this in the discord solo smart glasses.

52:03 - Mike Elgan (Guest)
No, they're, they're, they're called. Uh man, I'm spacing. I actually have a pair. There's so many of them. That's the problem.

52:09 - Leo Laporte (Host)
I can't wait to get my Brilliant Labs glasses. I know they're going to break my heart.

52:13 - Mike Elgan (Guest)
Yeah, and I know I'm going to look like Poindexter, so I don't know if you know this, but that's the company that came out with the AI monocle.

52:19 - Leo Laporte (Host)
I do know that, yes oh yes, we are aware, we are well aware aren't we paris okay, so he's got a dumb monocle around here somewhere the the company that I referenced earlier is called avigent.

52:32 - Mike Elgan (Guest)
Uh, it's a silicon valley based light engine company, so they make the little part that projects um ar data onto your, your, your lens, your glass lenses. They make really nice looking glasses and their whole model is to work with ai companies so that the ai company can sell branded glasses.

52:54 - Leo Laporte (Host)
This, this whole model is, is like so this is kind of like what google glass was, except google glass was projecting it to a effectively a screen over your eyebrow. I know you were a glass hole. Uh, this goes right on the glass lens, which is exactly what it should do the glasses look totally normal.

53:13 - Mike Elgan (Guest)
They look like absolutely regular glass, more so than ray-ban metaglasses do, a little more expensive, because they're trying to do the visual thing. They can do it in both lenses or one of them, and and there's a whole. They'd have some really sophisticated technology for doing that kind of stuff, but to me it's the model. So these are like, these are like the, this is like IBM and Microsoft. Right, they make the hardware. That's kind of kind of a commodity where you just pick the style. You put your prescription lenses in it or you you get sunglasses or both, and then the company is going to brand it. Openai, for example, should brand their own glasses, reflexity should brand it. They all should have their own hardware to go with these things and it's inexpensive. The kind with visuals in it are more expensive.

54:00 - Leo Laporte (Host)
I think these look pretty good. Paris, If you saw this on. Hinge would you?

54:07 - Mike Elgan (Guest)
You need a bow tie.

54:08 - Paris Martineau (Guest)
Yeah, that's definitely the look that all the Brooklyn girlies are going for.

54:13 - Leo Laporte (Host)
Swipe left or swipe right? That's the question. I look like Jerry Lewis. I think is what I look like.

54:19 - Paris Martineau (Guest)
You do look like you're about to do a comedy special. That's a little off kilter.

54:25 - Leo Laporte (Host)
Off kilter with Leo. I can't wait. I really can't wait to get these. They're doing something a little different, mike, because you talked about the Avagens are a projector projecting onto the lens. This has an OLED layer in the glass, so there's an actual screen in the glass and they're able to control it just as you would an oled tv with circuitry on either side of it. I think that's actually maybe a better way of doing it.

54:52 - Paris Martineau (Guest)
We'll see I'm curious how thick they're gonna be, leo well, here and then yeah they look like coke bottle bottoms.

54:59 - Mike Elgan (Guest)
I mean, they look really like.

55:00 - Leo Laporte (Host)
Really well, I'm getting my prescription so, since my prescription is as bad as yours, paris, it's going to clearly be I was going to say it's going to be like double thick.

55:09 - Paris Martineau (Guest)
I can't wait.

55:11 - Mike Elgan (Guest)
And I think the visual component is not as necessary as the glasses look like, not like that. So go to lucidco, which is L-U-C-Y-D dot C-O, is l-u-c-y-d dot c-o. All the this company is the one making 160, 160 smart glasses that work like ray-ban metas not quite as elegant, not quite as high quality, but very inexpensive and much better styles.

55:37 - Leo Laporte (Host)
and they come from big brand names. Yeah, exactly, et cetera. Let me turn off my ad block.

55:44 - Mike Elgan (Guest)
Basically, their model is that they work with these companies and they just make the whatever this part of the glasses is called from, from this to the temple, the temple. They make the temple and all the electronics and batteries and everything is in the temple and they're they're really good. They're really good and these look like regular glasses.

56:07 - Leo Laporte (Host)
They say they'll prevent accidents, but then they also say it uses Siri, which makes me think it's going to cause accidents. Take calls, play music. I was impressed with the MetaGlasses, benito. You heard them too, benito Gonzalez, our producer and musician. And you told me this, mike, the sound is quite good, even though it's in the temples and not in your ear. It was rich, it had good bass. I thought it sounded quite good.

56:31 - Mike Elgan (Guest)
So the lucid glasses are about 80% as good as the Ray-Ban Metas. They're pretty good. Not quite as good, but again they're cheap and so I think people should look at these.

56:44 - Leo Laporte (Host)
Yeah, of course, adding your prescription is going to add some cost to it. Yeah, I, I think we're so. I feel like we're this close to something, but is the AI ready for this? Are we good?

56:59 - Paris Martineau (Guest)
enough with the engines. I think part of the issue right now is all of these products are trying to figure out what the first big popular use case is going to be Like. Right now they're trying to do everything and I think that's really hard at this stage in the AI game. I'm sure there's going to be a point in you know two to five years or more that they'll be able to do everything really well, but right now it seems like maybe focusing on one or two features might make this a little easier. I would be, frankly, I would be incredibly down to spend money on glasses that just did like the Google Translate, like text sort of thing, where if you looked at something and it was in Japanese, it would just automatically translate everything I saw into English.

57:47 - Mike Elgan (Guest)
If they did only that, that would be fantastic. Absolutely so. The real question is why would anybody buy glasses, especially sunglasses, that didn't have AI capability in them?

58:00 - Benito Gonzalez (Other)
In the next couple of years, that's probably the case yeah.

58:03 - Mike Elgan (Guest)
Even if the AI is no better than your normal voice assistant.

58:08 - Leo Laporte (Host)
it's still a useful thing if it's the same price as your regular sunglasses, right? Yeah, I think you're right. I think that's the product. Maybe not this year, Next summer.

58:19 - Paris Martineau (Guest)
I mean, is it though? Is the average person going to shell out on AI glasses? Probably not, Like I try to think a lot of times when I'm talking about like use cases.

58:29 - Leo Laporte (Host)
It's not just AI, it's also like what is my mom it's instead of AirPods.

58:33 - Mike Elgan (Guest)
It replaces Instead of AirPods. Exactly, it replaces your earbuds.

58:36 - Paris Martineau (Guest)
I guess that's.

58:37 - Mike Elgan (Guest)
You have phone calls through them. I mean, it's like it's actually.

58:41 - Leo Laporte (Host)
It's your Bluetooth headset.

58:42 - Mike Elgan (Guest)
It's actually it can already do what you're doing without anything in your ears, so it's just more comfortable to have nothing in your ears, rather than.

58:51 - Leo Laporte (Host)
This is also from the New York Times story about Siri. Rather than compete directly with chat GPT by releasing a chat bot that does things like write poetry Apple has I don't really need Siri to write me poetry Apple is focused on making Siri better at handling tasks it already does, including setting timers, creating calendar appointments, adding items to our grocery list. It will also be able to summarize text messages. So we'll see, I mean this is this is all going to kind of come to a head in the next four weeks. So we'll see, I mean this is this is all going to kind of come to a head in the next four weeks.

59:26 - Mike Elgan (Guest)
And, of course, apple's pitch and the challenge for Apple is Apple's pitch is privacy, and to some degree, getting these things to work usefully and well requires they send data out to the world. Yeah, but I don't think. I really don't think. You know, apple is Apple and they're going to really lock this down. They're not going to, they're not going to give you chat GPT. They're going to use chat GPT behind the scenes, unbranded, to do very, very specific things, like the list that you mentioned about, you know, improving results and summarizing texts and things like that. But it's going to be pretty minimal and probably that minimal. The minimal nature of it is something that's going to be able to live on the phone and it will also, to Apple's benefit, give you a reason to buy the new phone right, cause it's got to be more powerful.

01:00:21 - Leo Laporte (Host)
One thing though that is mentioned in this article. The people who worked in the Siri division say is that Apple's tendency to secrecy, to keeping teams independent of one another, is problematic. They are going to need to restructure because AI will not make sense unless it's threaded throughout all of their products. The Times gives us an example the HomePod smart speaker, which is actually dumber than Siri. It's even less useful than Siri and, as a result, it's not sold. Well, apple has got to change how it does business, according to this Times article and at least one person who used to work on the Siri team. So this will be interesting. Clearly, I think Apple realizes the threat and realizes it has to do something. They killed the car team and assigned everybody that they could who was working on an Apple car to the AI division. I think that signals an adjustment of priorities. Well, we're going to take a break. Oh, we lost somebody. No, we have everybody.

01:01:31
Mike mike's back there's mike's back, I'm here taking a trip down to the beach, but he's back now. All right, we're gonna take a little break. We'll come back with more. There's a lot more to talk about. Mike elgin, samable samad, paris martineau our show today brought to you. Nice job, benito kept up with me. Good job, let me do that again. Mike elgin, samable samad, paris martineau wow, he's so good, so good. Who needs ai? I got benito our show. Let's say it one more time Paris Martineau, mike Elgin and Sam Abulsamed Holy cow, oh he is on the wow, he's fast.

01:02:09
I mixed it up and Mike even got fireworks. Well done, Mike.

01:02:14
Our show today brought to you by NetSuite, the less your business spends on operations, multiple systems and delivering your product or service. Well, that means, then, the more margin you have, the more money you get to keep. Well, here's a great way to reduce costs, reduce headaches. Smart businesses are graduating to NetSuite by Oracle, to NetSuite by Oracle. Netsuite is the number one cloud financial system bringing accounting, financial management, inventory, hr all into one platform. One source of truth With NetSuite, you reduce IT costs because NetSuite lives in the cloud with no hardware required. That also means you can access it from anywhere, so convenient. You cut the cost of maintaining multiple systems because you've got one unified business management suite. You improve efficiency by bringing all your major business processes right there, into one platform which slashes manual tasks, slashes the errors, copying data from one to the other. Over 37,000 companies have already made the move from one to the other. Over 37,000 companies have already made the move. Now do the math, see how you'll profit with NetSuite.

01:03:28
By popular demand, netsuite has extended its one-of-a-kind flexible financing program for a few more weeks. But don't delay Go to netsuitecom slash twit right now. N-e-t-s-u-i-t-e netsuitecom slash twit. We thank him so much for supporting this week in tech. We thank you for supporting our advertisers. You're great about that net suitecom slash twit. Make sure you put that twit on the end so that they know you saw it here. Thank you, net suite. Oh man, this is going to be a fun week. I think this is a very exciting week. I think google will have what do you? What are you expecting, mike Mike Elgin, for Google IO, besides this AI stuff?

01:04:05 - Mike Elgan (Guest)
Well, yeah, I think one of the stories is probably going to be it's too many things, Google, too many AI things. I think they're going to say, oh, we have this one.

01:04:15 - Leo Laporte (Host)
They're going to splash AI on everything.

01:04:18 - Mike Elgan (Guest)
That's right, and so that'll be a bit much. One of the strains of complaint against Google these days, especially this week, is that small websites and blogs and you know so on, are complaining that Google search is leaving them behind, just ignoring them, that you don't get the results you used to get with Google search. So it's the most fundamental thing about their business fixing Google search and making it work for content creators, for writers, for bloggers and so on and so there's a there's a strain of hope that they will do something about that, and I'm pretty sure they're not going to even mention it and that they'll do nothing about it because they don't listen to those that sort of thing, and so that's going to be, I think, very disappointing, but it's going to be AI all the way down and it's going to be too much.

01:05:10 - Leo Laporte (Host)
Yeah, actually, paris, we've been talking about this on this week in Google, the small businesses and big businesses who are losing traction as Google changes its search results to perhaps get rid of spammy and SEO driven links. And this is the problem. It's hard to know if this is Google responding to the crapification of the web or the crapification of Google.

01:05:37 - Paris Martineau (Guest)
Or if it's part of this larger trend that I feel like people have been paying more attention to recently, which is the Google search team being increasingly influenced by Google ads and the money makers. I'm sure none of that will be mentioned at Google IO. Opaque when it comes to how search was like the nitty gritty behind how search results work, in part because they don't want people to game the system, but it's quite interesting to have as an undercurrent to this event, to have it be a time where a lot of small, medium businesses, as well as major media companies, have seen a major hit to their uh traffic, their seo traffic, based on just strange behind the scene changes that no one can seem to explain this coincides with the loss of traffic from social.

01:06:31 - Leo Laporte (Host)
So it's yeah, it's pretty tough to to have a website these days yeah, this comes, uh, also from your friend, uh, ed zitron's piece on the man who ruined I say ruined Google search, prabhakar Raghavan, who's now in charge of Google search, who, in fact, before ruining Google search, ruined Yahoo search. I wonder if we're going to see Raghavan at all on the stage Tuesday, or if you're right, mike, that they're just going to avoid the entire issue and and talk about forget. I think google would like us to forget that they, what they really are, is an ad company and if google's smart, you won't see raghavan any anywhere.

01:07:14
Yeah, yeah, mention of them um, seo, um, companies have been kind of playing the game to get better search results, but then what's happened is that companies that have done that are suddenly downranked because Google's trying to get rid of spammy content. Uh, we've seen some of the biggest sites in the world lose as much as half of their traffic after Google's latest changes. But you're right, none of this is going to come up, nope.

01:07:50 - Mike Elgan (Guest)
They have other priorities.

01:07:52 - Leo Laporte (Host)
Is some of this really Google hand-waving trying to get us to? No, don't look at that. Look over here. Ai, we're going to have AI. Here's the story from the Verge. They quote the story we've talked about a couple of weeks ago from House Fresh, which is an air purifier reviews site. They wrote in February about how they were losing traffic because people were creating pages that were bogus review pages. But we're getting all the traffic from Google and now then we see that Google starts to cut the results on these other sites. It's gotten very sketch out there and I guess the real issue is is what does Google owe the web? What is the web? Oh, google, and can you make money on the web these days? Or maybe the whole idea of a website with ads is, is flagging, is not, is futureless.

01:08:59
the other stat I should throw in here is that 52 percent of all americans, according to pew use ad blockers more than half yeah, it's uh tough for everybody out there.

01:09:11 - Mike Elgan (Guest)
Yeah, have you, have you heard about or talked about, the dead internet theory?

01:09:15 - Leo Laporte (Host)
no, it's doesn't sound good conspiracy theory.

01:09:18 - Mike Elgan (Guest)
It's a conspiracy. It it's a conspiracy. It's one of those conspiracy theories that's pretty much mostly true, which is that the rise in synthetic content, algorithmic curation bots and stuff like that is now the vast majority of the stuff on the internet. So the the this problem with, uh, small websites getting traffic uh is you know, as you pointed out from Google's side, there's just so much garbage out there being automatically churned out. I've seen YouTube videos where people are like yeah, you can write a book in three hours with a chat, gpt and people are doing that they're writing like several books a day and just churning them out.

01:10:05
And, as a you know, book authors have to sort of. You know it's not about competing with an AI generated book, which is going to be absolute garbage, but it's just like getting noticed when the number of players in the game is exponentially larger. I think that's a big big thing that's happening. But the dead internet theory is at least fun. There's some true believers who have you know, take a very conspiratorial look at it, but it's a fun thing to search for and look at and talk to AI about.

01:10:37 - Leo Laporte (Host)
There's an excellent article in the uh about all of this saying uh, uh, maybe you missed it, but the internet died five years ago. And the problem is, even if it is, we kind of know that's not true. It's also we kind of know it is true, or it's the trend. Uh, and in particular the notion that the uh I mean she, she dead internet theory suggests that the internet has been almost entirely taken over by artificial intelligence. And the problem is, it's not just the web, it's going to be books on amazon music at the apple itunes store, it's going to be movies. I feel like the Store, it's going to be movies. I feel like our entire culture is going to be taken over by people who are trying to game the system. You saw the story about the guy we did the story on this Week in Google, about the guy who's making hundreds of thousands of dollars a year generating bogus songs on Spotify.

01:11:45 - Paris Martineau (Guest)
Well, I mean, he technically writes the songs, he just spends like five minutes on every song.

01:11:48
There's songs about poop and things right yeah, they're, I mean, the lowest common denominator of content. I guess that isn't ai generated, but I do think that there's some merit to this concept that, as it gets easier and easier to generate content of all types in, just you know, the click of a button, the right of a prompt, yeah, the fact it's going to be harder to find good content amid all the slop the fact that this guy has written 80 000 songs in the last 10 years to generate income.

01:12:20 - Leo Laporte (Host)
even if it's pennies at a time, it adds up. The fact that he wasn't using AI is just a matter of timing. He'll be using AI from now on, right? Yeah, why shouldn't he?

01:12:32 - Mike Elgan (Guest)
But the frustrating thing is, this is why we need Google now more than ever.

01:12:36 - Leo Laporte (Host)
We need a good Google. We need a Google that works.

01:12:39 - Mike Elgan (Guest)
To bring us to the good sites and get us through the chaff. But does it Ignore the chaff? It doesn't.

01:12:46 - Sam Abuelsamid (Guest)
Not yet. Yeah, I don't think there's any system that can yet. And it's going to be tough to break through all that chaff, you know, to break through all of that synthetically generated garbage that's out there. All of that synthetically generated garbage that's out there. And you know, I think, as long as you know, we, I think, as long as we've got ad tech that is paying for this stuff, then I don't know, I don't think that we will be able to get rid of it.

01:13:16 - Leo Laporte (Host)
There is an evil human tendency and and and google suffers this all the time. But people want to game the system. I remember when, uh, kevin rose was doing dig which was a great idea that you would put you know you would read news is that news sites that were voted up by the community so you'd get the best news in theory. But very quickly people figured out how to game it and dig went from the best site on the internet to the worst site on the internet. Uh, very for mr baby man it wasn't the worst site mr baby man loved it.

01:13:50
no, so well, you're talking about twitter, uh, but so see, here's what strikes me. This is a great opportunity for us because we create human content and, presumably unlike Google, we don't have any incentive. Are we really human?

01:14:09 - Sam Abuelsamid (Guest)
though it's true I've never met any of you guys in person, you could be fake.

01:14:16 - Leo Laporte (Host)
Only Paris is human. The rest of us are. Well done, ai fakes. No, I think that that's an opportunity for human scale content. Yes, because if you are a human, you're going to rapidly start turning away from anything like Google that's generated, because it's just going to be garbage.

01:14:36 - Mike Elgan (Guest)
Right, and right now, ai stuff is the shiny new thing. Right, and you know, right now AI stuff is the shiny new thing, so hopefully that sort of idea will. You know, people tolerate AI generated images, for example, some of which is actually quite beautiful and quite amazing, but they tolerate more than I think maybe they will in the future and and this, by the way, you mentioned the change in my newsletter. This is a good opportunity to talk about that, because I call it machine society, because I think that's what we are. We have a new machine society basically, and what I mean by that is it basically started.

01:15:13
This whole the phrase was coined in 1829 by Thomas Carlyle, a Scottish philosopher, who pointed out that mechanization in 1829 was changing human nature, changing human society. All of these machines, all these computerization, digitization, all that kind of stuff that's been intervening in our society or that we use to have a society, has been kind of it's like secondhand smoke, right, whereas all these AI tools is firsthand smoke, and I think people are going to start changing very quickly because of all the ai and everything all the time, and especially when we talk about our communication. So you guys don't use, you know, uh, any of the new ai based um podcasting tools? I don't think. But but podcasting itself we do.

01:16:18 - Leo Laporte (Host)
I mean we actually do we use, we use them. They're actually great.

01:16:23 - Paris Martineau (Guest)
Yes, they are. What tools are you talking?

01:16:25 - Leo Laporte (Host)
about Well. So this is really important. We don't use it to create content in any way. It's always after the fact. We use it to chop up video. You've received these in your mailbox, paris. After you're on the show yeah, show we send you clips that you can put on your social. That's done by an ai with human intervention. The ai proposes and we dispose, but uh, so there I this. By the way, my fundamental belief in ai is it's got to be a partnership with a real human uh, to make anything useful. But that's right. We do use it and we use it for show notes and we're using it for transcriptions there's, but it's always after the fact and I don't I think that's harmless. I hope that's harmless. I think our content is always human driven it's absolutely it's.

01:17:08 - Mike Elgan (Guest)
it's harmless and it's beneficial, and and even, as thomas carla pointed out, all the mechanization is feeding people and clothing people and housing people much better than without the machines, and so there's's enormous benefits. So a lot of podcasters using something called Riverside probably use a tool similar to that, where you can edit the audio and video by cutting the words that it generates, and so it's all tied together.

01:17:32
It's amazing, it's a wonderful thing, but everything is like that People building presentations, people writing emails everything we do is becoming subtly AI, augmented in a good way, and the thing that's interesting about it, I think, is that we have one way to look at it as a partnership between AI, machines and human beings, and the role of AI will never stop growing. It will just grow and grow and grow and grow, most likely, I think, but I think that's just. You know, this machine society concept, I think, is a way to describe what we are becoming as a species, or at least some of us Not everybody is, you know got AI everywhere all the time. Like we are getting to the point. And again, this generative AI revolution is what? Less than two years old. So give it five years and it's going to be quite, it's going to have quite a huge impact on human society.

01:18:36 - Leo Laporte (Host)
I just subscribed to machine society. Oh, thank you. I hope you uh don't mind this, but I subscribed using my special email, laporte and jam 57e at inboxomnivoreapp. You don't mind that, do you? Yeah, not at all what is that word?

01:18:58
so I use. I've been starting to use this open source. I love it. Mailing list newsreader called Omnivore. You see, I subscribe to the information on it and it goes here instead of my mail, but I have to use a special email for it. So when I sign up you'll see. You'll see, mike, you'll see that weird email address. But now my inbox will have in here and I can look at the individual subscriptions.

01:19:24 - Paris Martineau (Guest)
I like the amount of these that are just information journalists who have written the briefing.

01:19:29 - Leo Laporte (Host)
Yeah, I know, I wish there were a better way of categorizing it like that, but I'm still. I like Amir, so I want to see you know when he wrote that, but but for most of the part most part it's like I could see my account. In fact, I have to finish signing in to that address and there's my payment receipt and all that stuff. So I don't want it to go to my email. I want it to go to this, which is basically a newsreader Right, right, that's great.

01:19:54 - Mike Elgan (Guest)
Yeah, plus I read it for you so you don't have to.

01:19:57 - Leo Laporte (Host)
You don't have to. Yeah, oh, that's the other side of it. I don't. I don't read it at all. So, uh, no I. The problem is that my email box is filled with with newsletters. Uh, I'm almost amazed that this ecosystem works, because whose email box is not completely unusable? But this makes it usable. Frankly, I, uh, so I'll, I'll start seeing your stuff this way, which is great.

01:20:21 - Mike Elgan (Guest)
Yeah.

01:20:23 - Leo Laporte (Host)
Great yeah. So I hope you don't mind that I subscribe that way and I you know I'm with you. I think this is going to be the big story for us for the next maybe forever, but then certainly the next few years and this is why I think it's so important that we've got Paris on Twig and Sam, who's the king of cars, and AI and cars are like peanut butter and mayonnaise they just belong together.

01:20:51 - Sam Abuelsamid (Guest)
And Mike Elgin, like peanut butter and jalapenos.

01:20:54 - Leo Laporte (Host)
There you go. There's a delicious mix.

01:20:56
All right, we're going to take a little time out, but we will come back with more, and I do want to talk to you, sam, about what the hell is going on at Tesla. I tell you it's crazy over there. First, though, before we get to the ad for this segment, I would like to remind people that if you want to keep humans creating content that you can understand as a human, that is not motivated by some, you know, covert financial interest, the best way to do that is to support the stuff you like, like Mike's newsletter and like Twit. I hope we're in that camp. I'd like to invite you to join Club Twit Seven bucks a month, ad-free versions of all the shows we do and we do so many great shows video of all those shows, access to a wonderful community of people in the Club Twit Discord. I think that's not to be diminished. I mean, it's not just humans doing shows. It's fantastic people in the community that you can talk with and chat with all the time, 24-7. And, frankly, if we want to keep doing what we're doing without fear or favor, objectively, we're going to need the support of our audience. So if you would twittv slash club twit, we would like to keep doing this because I think you need humans. You really do. You need human created content. Like mike elgin and paris martin know what the information. One of the things I love about the information is. The content is so good and sam and wheel bearings and all of that our show today, brought to you by mint mobile.

01:22:39
Love the mint mobile. This is a good example of a company that you know is the exact opposite of the big cell phone companies. Mint Mobile eliminated the stores, they eliminated the overhead and they've got wireless, premium wireless service down to $15 a month. We just did a big spring cleaning at the LaPorte household and the veterans are coming. We've got 50 bags of stuff that we're putting out front tomorrow morning and it's kind of amazing. When you get all that stuff cleaned out, you go wow, how did I live like this Same thing with Mint Mobile. When you realize how much you're overpaying for wireless when Mint Mobile can offer you plans for $15 a month. When you purchase a three-month plan, you'll go how have I been affording this? Switch to Mint Mobile, you get unlimited talk. Switch to Mint Mobile, you get unlimited talk, nationwide talk and text and data for $15 a month. I tell you what? When I started using Mint Mobile this is about three or four years ago now I signed up immediately, for a year at a time, because it was such a good deal. And what's great about Mint Mobile is, while inflation has driven prices up for everybody else, it's only driven the data up for Mint Mobile. I now get more data for my money than ever before.

01:24:06
All Mint Mobile plans come with high-speed data and unlimited talk and text, delivered on the nation's largest 5G network, t-mobile T-Mobile MVNO. You can bring your own phone. They'll even give you a SIM for free, or you can use an eSIM. Keep your phone number. They'll even give you a SIM for free, or you can use an eSIM. Keep your phone number. Don't worry about that. All your contacts, all your software goes over, or you can buy a phone from them. They've got great deals great deals on new phones.

01:24:31
Ditch overpriced wireless with Mint Mobile's limited time deal. Right now three months of premium wireless service for 15 bucks a month. Now, to get this new customer offer and your new three-month unlimited wireless plan for just $15 a month, go to mintmobilecom slash twit. See that All the plans now 5 gigabytes, 15 gigabytes, 20 gigabytes, unlimited Mintmobilecom slash twit. Cut your wireless bill to $15 a month. Mintmobilecom slash twit. Now you'll have to pay for the first three months up front $45, $15 a month. New customers on the first three month plan only Speed slower when you go above 40 gigabytes. On the unlimited plan I don't think you're going over 40 gigabytes. Additional taxes, fees and restrictions apply. See Mint Mobile for details.

01:25:15
And look, they got the broadband nutrition label now, which makes it very easy to understand what you're getting. It's always been easy because Mint Mobile is very straightforward. That's those great Ryan Reynolds ads you see on TV Mintmobilecom slash twit. Okay, I'm not Ryan Reynolds, but I'm bringing you the same deal. Thank you, mint Mobile, for supporting this Week in Tech. All right, sam, we got to talk about Tesla. We got to talk what the heck is going on the latest? Well, actually it's not even the latest anymore. Recently, elon Musk took its I would say the crown jewel of Tesla, I would say the crown jewel of Tesla their supercharger system, and fired the woman who was running it and all 500 people running superchargers. Where does this make sense?

01:26:17 - Sam Abuelsamid (Guest)
How did this happen? I think that one possibility, know one possibility is that I'm sorry, Sam.

01:26:23 - Leo Laporte (Host)
It's so mean to say explain Elon Musk? It's impossible. I understand that.

01:26:30 - Sam Abuelsamid (Guest)
One possibility is that, you know, elon is so intent on proving that Tesla is an AI and robotics company that he is almost deliberately trying to kill the car business.

01:26:44 - Leo Laporte (Host)
This, by the way, this story was insane, broken by, I'm proud to say, steve levine at the information. Uh, and we've seen a lot of rumors. Rebecca tanucci was the senior director of the supercharger group. She was in charge of all charging, also destination chargers, right? Yep, the rumor was that Elon was going to go hardcore and fire 10% of Tesla, right, and that she went against him.

01:27:17 - Sam Abuelsamid (Guest)
Yeah, she apparently pushed back on that and said look, you know we're trying to expand the network. We've got new automakers coming in that are getting ready to start using the superchargers. We're trying, you know, we're trying to take advantage. You know, get some of this NEVI funding as the funding from the infrastructure bill to pay for some of the new chargers. You know we've got a lot of projects going on. We can't afford to lose 10% of our staff.

01:27:42 - Leo Laporte (Host)
Furthermore, the brilliant thing that Elon did from day one was nobody's going to buy an electric vehicle if it is not as easy to charge it as it is filling up a gas vehicle.

01:27:53
He built this crown I'm serious, crown jewel.

01:27:57
I didn't buy a Tesla until they put a supercharger in Petaluma.

01:28:00
Now it turned out I didn't really need one, but it just reinforces your confidence that you're going to go on a road trip or whatever, you can drive your car and you can get it filled up fast anywhere. And then to top that off, massive victory in the last few months, as automaker after automaker has thrown up their hands and said yeah're going to switch over to elon's uh, next charger, we're going to support the superchargers and someday soon and I think ford's already done this you'll be able to charge your car at a supercharger. Big win for elon, because he makes money on all those cars that are not absolutely yeah, so uh, so tuesday, according to the information, tuesday 10 am elon musk said he would ask for the resignation of any executive who, quote retains more than three people who don't obviously pass the excellent, necessary and this is the most important one trustworthy test. I have been super clear about this. Steve Levine got this memo and then he singled out Tanuchi and the Supercharger Network and fired them all.

01:29:12 - Sam Abuelsamid (Guest)
Yeah, I mean it's incredibly short-sighted. It's a really dumb thing to do because, consistently, you know, what we've found in doing consumer surveys over the years is that the top barriers for EV adoption are affordability of EVs, because they are expensive, availability of charging and the time it takes to charge. And from the beginning you know from 20, almost from the time they launched the Model S in 2012, not long afterward, by early 2013, they started deploying these superchargers that address those last two the availability and time to charge. They made charging much more convenient. They have done a vastly better job than any other charge point operator at creating a reliable network. You know that's both. You know widespread in a lot of places. But also, when you get there, the chargers almost always work. You know when you go to a gas station, you know it's very rare that you get to a gas station and the gas pumps aren't working.

01:30:15 - Leo Laporte (Host)
You know it's not at all the time at Electrify America and other. Yeah, electrify America has been particularly bad yeah.

01:30:22 - Sam Abuelsamid (Guest)
But you know EA exists only. You know EA is the compliance charge point charging network.

01:30:28 - Leo Laporte (Host)
They clearly don't care.

01:30:30 - Sam Abuelsamid (Guest)
Right. They exist only because Volkswagen got busted for Dieselgate and part of their settlement with the government was that they would invest $2 billion in building out a DC fast charging network, and they have done, for the most part, a really miserable job of maintaining those chargers and making sure that they work reliably.

01:30:50 - Leo Laporte (Host)
Yeah, cause that's not part of the deal, but it's still pretty bad. They only said they'd build them I, so I had a. When I visited my mom a couple of weeks ago in Rhode Island, I was nice. I got to drive a Kia Ioniq 6. Very nice vehicle. But the rented car company Budget says you know, when you buy a gas vehicle you can pay and say just bring it back and we'll fill it up. You just pay, right? Not with an electric vehicle. They said bring it back with 70% or what I don't know they're going to bill me or something. So I had to find a super ea to charge it up, learned a couple of things I didn't know. For instance, the ea chargers have two plugs but you can only use one at a time, did you?

01:31:33
know that and there's no sign on the station down the road from me.

01:31:37 - Sam Abuelsamid (Guest)
Here there's no sign on.

01:31:38 - Leo Laporte (Host)
It says if somebody's using this, you go to a different one. It just has two plugs. So I thinking dumbly, thinking well, that means I can charge. I plugged it in and the poor couple next to me got knocked off it. Their thing stopped, they had. They came running back, said what happened. I said I don't know. I plugged in. They said oh don't you know, only one at a time. So where's the sign that says that? I looked over at another pump and somebody put one of those Hi my name is stickers on there that said one chart, one at a time.

01:32:12 - Sam Abuelsamid (Guest)
Yeah, I didn't even do that Right, so clearly you found one of the one of the older chargers they have. Over the last year they've been replacing a lot of the chargers and the new ones just have one cable why did they have two, if they? Can only have one. I have no idea. I I cannot give you an answer to that question.

01:32:32
It's so crazy and I've never I've never been able to get a good answer from anybody at electrify america about that, uh, but you know, with, with tesla, uh, you know they've done a good job yeah, superchargers always work.

01:32:45 - Leo Laporte (Host)
They're great, they're fast. There's rarely a line they have. Whenever they have a setup, they usually have many, many chargers available. It's very nice and that's one of.

01:32:55 - Sam Abuelsamid (Guest)
That's one of the one of the interesting things. You know, if you look at the, the data about the, the various charging networks, on average, uh, on average, tesla actually has fewer supercharger locations than all the other networks they have. I think currently about 985. It doesn't feel like that, does it? But where they're located they have know they tip. They have on average 12 chargers per site, whereas Electrify America has about um, about 800 and some locations right now 800 sites, but they only have an average of uh five charge five chargers per site, or about four and a half chargers per site. Uh, one near me here has six. A lot of them have four or two, and so you know that's one of the problems. You know when they do go down, you know you end up it's harder to find a site that has enough working chargers, and so there's all kinds of issues with these things.

01:33:58 - Leo Laporte (Host)
And this is bad for. Ev adoption because people see this. They don't. The truth is, most of the time when you drive an EV, you charge at home and it's not an issue.

01:34:08 - Sam Abuelsamid (Guest)
I haven't charged it in years. That's the truth right now because most of the EVs that are on the road have been built in the last five years, so they're new and it's people that buy new cars. The reality is, people that buy new cars are a tiny minority of the overall driver market. We sell, in the best years, about 17 million cars. Last couple of years it's been about 14, 15 million cars and we sell on average, about three to four times as many used cars every year as we do new cars.

01:34:40
Most people don't ever buy new cars. Most people buy used cars and you know, for new car buyers they tend to be a little more affluent because they have to be to afford a new car, so they're much more likely to be. You know they live in a single family home with off street parking and they can charge at home. When you look at the used car market, it's about almost half of those people live in multi-unit dwellings, apartments, condos, my son, lives on a houseboat and he has to charge his Tesla supercharger down the road because there's no other way to fill it, and in Paris you would have to do the same.

01:35:18 - Leo Laporte (Host)
You don't have a way to charge a car. Yeah, but you don't have to have a car in New York.

01:35:22 - Paris Martineau (Guest)
I don't have a car.

01:35:23 - Leo Laporte (Host)
You have a giant electric train that you can ride, which is really cool.

01:35:26 - Paris Martineau (Guest)
I do.

01:35:27 - Leo Laporte (Host)
Yeah, that's a much more efficient way of getting around than a car.

01:35:30 - Sam Abuelsamid (Guest)
It's underground, but other than that, no matter what Elon says, a subway is a vastly more efficient way of moving large numbers of people than putting a bunch of teslas in a hole in the ground if we had a subway here, I'd ride it.

01:35:44 - Leo Laporte (Host)
We have a smart train, but it doesn't go anywhere. I want to go, so it's not that smart. Also, fired daniel ho, who was in charge of all new vehicles, including the model s, the three, the y, the new vehicle being designed right now, or was it? No, it's not being anymore. He was a 12-yearyear Ford executive. Has been with Tesla for 11 years. Fired. Elon writes hopefully. How should I do the Elon voice? Hopefully? These actions are making it clear we need to be absolutely hardcore about headcount and cost reduction. Hardcore about headcount and cost reduction. While some on exec staff are taking this seriously, most are not yet doing so.

01:36:27 - Sam Abuelsamid (Guest)
So you're farred, uh, you know, if you want to be hardcore about cost reductions, maybe don't try and demand your shareholders approve a 50 billion dollar payout for you. He gets paid. You know that that 5050 billion can go a long way to developing new products and building new chargers.

01:36:45 - Paris Martineau (Guest)
There's also been so much tumult going on behind the scenes over the last 10 weeks. On Friday, my colleagues Steve Levine and Becky Peterson were two of the best reporters on Tesla. They dropped this great long read. The headline is At Tesla a wild week that defined the company's future. That kind of goes into one of the key moments that kicked off this period of tumult, which is this week in late February, where Elon Musk simultaneously decided to just tweet out that the Roadster, one of their kind of successor electric cars, was going to be able to accelerate from zero to 60 in less than a second, which was a big surprise to the entire team working on the Roadster and by the way.

01:37:29
You don't want that.

01:37:33 - Leo Laporte (Host)
I had a car that I had the ludicrous mode on my Model X and I stopped doing it because I almost blacked out. It's not safe to accelerate.

01:37:43 - Sam Abuelsamid (Guest)
It's, lud it's not safe to it's ludicrous frankly and 99.9 of drivers on the road do not have the skills to control a car that can accelerate that fast.

01:37:52 - Paris Martineau (Guest)
No, that's dangerous here's the great illustration only wait a minute. I want to hear more. Our illustrations are fantastic.

01:37:59 - Leo Laporte (Host)
I have to show this illustration of Elon. It's really good so what else happened?

01:38:08 - Paris Martineau (Guest)
It was not like the only crazy decision that happened that week. Somehow, in meetings with senior Tesla lieutenants, elon scrapped plans for a $25,000 vehicle known as the Model 2 and accelerated plans to mass produce the Robo Taxi, setting a goal suddenly to do so by the second half of next year, which came as a huge surprise to many of the people working on it. Musk had been promising that the Model 2 was going to come out soon for nearly four years, and just a month before he told investors it was going to be ready at the second half of 2025.

01:38:42 - Leo Laporte (Host)
so we just shook everything up immediately he does say he's going to spend half a billion dollars more on superchargers. There's only one problem, elon you. You fired all the people who know how to build them. But okay, never mind I did.

01:38:54
Okay, I'm gonna say details, details I'm gonna say something that you don't often hear me do. I'm going to defend Elon. He did the same thing at Twitter. He fired, he cut to the bone and we thought you're not going to have a company in a couple of months because the whole network's going to fall apart. And somehow Twitter has managed. I mean, I'm not talking about the content, but technically it's doing just fine.

01:39:19 - Paris Martineau (Guest)
Technically, twitter exists. Are you on Twitter? Most people I know have fled Twitter. I agree the content on it is harder to find. The content is horrific.

01:39:28 - Leo Laporte (Host)
But he did fire most of the engineering team and the site still runs.

01:39:34 - Paris Martineau (Guest)
It's true.

01:39:36 - Leo Laporte (Host)
Right. So maybe this slash and burn, you know, one philosophy is fire everybody, because it's impossible to know who's wasted, you know, energy and who's actually doing the work. Fire everybody. No, it's not impossible, it's not.

01:39:49 - Paris Martineau (Guest)
I mean if you're running a company. You should know that, especially if you've spent how many years as he has at Tesla, the person to blame for you having all that dead weight is ostensibly yourself. Well, that's true chief executive when all those hiring decisions were made but it is.

01:40:06 - Leo Laporte (Host)
I mean, okay, look, there's a, there's a form of agriculture called slash and burn, right where you just you just cut it all down and you burn it and you start over, and I think that's maybe not the best way to do it, and I think that's maybe not the best way to do it, but you can do it, and I think he's doing that. He's taken the sugar cane field and he's just burned it to the ground and now he's going to see what breaks and build it up from scratch. I mean that's. Well, it wasn't the result.

01:40:41 - Mike Elgan (Guest)
I mean, that's well, it's the result. Yes, it's so. These decisions we're talking about are not the result of analysis that the problem with Tesla is that there is way too much waste in salaries, for example. I think that was the calculation for Twitter. Is he basically said OK, we don't need this many people? And this seems like a very emotional decision. And also, Elon Musk, unfortunately, is a megalomaniacal, narcissistic sociopath, and so people like a little easy on him.

01:41:17
I wouldn't, I wouldn't, I won't mention anybody in politics. If, if, if your mind goes there, you can. You can go there. But certain people with this personality, whatever the truth is, whatever sounds good in the moment, so you know, if he gets emotional in a meeting and wants to, you know, in a meeting and wants to make a big deal, he's, he's the dictator of Tesla. He can just say fire everybody. I'm sure it felt really good for like 10 minutes, right, but this is a, this is a mindset, this is a personality, uh, trait, uh, or type, and uh, and we, we, we see more and more of this in the public sphere. Uh, unfortunately, and it it never ends well, it really doesn't. And I don't think Twitter is a shining example of how great it is to have that kind of personality running things.

01:42:07 - Sam Abuelsamid (Guest)
And you know, as a prime example of what you just said, mike, you know, with Rebecca Tanucci, you know she apparently pushed back on the number of people that Elon wanted to cut from the charging group. The charging group felt it was too many people, given what they had to do, the projects that they were working on and supporting all these other automakers coming in to use the superchargers. And she pushed back and so he fired her and then fired everybody that worked for her. That's that kind of you know in the moment, thinking. You know that that is just it's foolhardy and you know it's just a bad way of running a company, can we?

01:42:45 - Leo Laporte (Host)
just tip a little glass to rebecca tanucci. That's a good leader who who said no elon and and got fired with her whole team, but she went down with the ship and just you know what. Good for you, ms tanucci. I hope you land well and you and you and you find a good job. I'm sure you'll find a better employer.

01:43:04 - Sam Abuelsamid (Guest)
That's what a boss should do, I'm sure she has already been recruited by at least three or four companies that I'm aware of, like IANA, which is the new joint venture of seven automakers to build an even larger charging network than what Tesla has.

01:43:21 - Leo Laporte (Host)
Hey, there are 500 people and a great leader who know exactly how to do that, and they're available right now.

01:43:27 - Sam Abuelsamid (Guest)
They're pouring through LinkedIn profiles as we speak.

01:43:31 - Mike Elgan (Guest)
Or they're working on a startup, because that's what built Silicon Valley. That's true Disgruntled employees who go off on their own and start their own thing.

01:43:39 - Paris Martineau (Guest)
I mean, I do think it's notable that you have basically an entire 500-person company out on the market all of a sudden.

01:43:46 - Sam Abuelsamid (Guest)
And that's the beautiful thing about the fact that they can't enforce non-compete agreements in California. The reason why you have Silicon Valley and so many startups in California is because people can leave and go and compete and do something that they think is better or smarter.

01:44:05 - Leo Laporte (Host)
The traitorous ape. I mean eight Traitorous ape would be a movie. I want to watch that movie. But I'm talking about the traitorous eight. This is, I think, what you were probably thinking about. Uh, mike Elgin, a group of eight employees who were working at Shockley's semiconductor lab in 1957. He invented the transistor to found Fairchild Semiconductor. They were young PhD students Shockley hired to build the first microprocessors. They left and the names you probably recognize, like Gordon Moore and Robert Noyce, who went on basically to not only form Fairchild but then go on to found Intel and AMD. They called them the Fairchildren. And yeah, and that's because they had a bad leader and they said we can do better. And they did. And who's ever heard of William Shockley ever since?

01:44:59 - Mike Elgan (Guest)
Well, I also. I happen to mention earlier. We're talking about um perplexity ai, and I mentioned that it's better than google and chat gbt. It actually is founded by a former. Yeah, google and chat gbt and anthropic executives is former google people.

01:45:15 - Leo Laporte (Host)
I mean, talk about that incestuous industry. Ai is the worst. But do review, read. If you are not uh subscriber to the information subscribe.

01:45:24
I've been mispronouncing his name, it's Levine, steve Levine and Becky Peterson's fabulous long read at Tesla a wild week that defined the company's future. There's just one paragraph here. That's by the way. The pictures are great too. Here's Elon imagining the future. I don't know what he's doing there, but that's by the way. The pictures are great too. Here's elon imagining the future. I don't know what he's doing. That's from getty.

01:45:47
Uh, this uh part about the uh, the team, the model two. The team's initial goal was ambitious to pump out 10 000 model tools a week, half a million cars a year, twice as fast as the rate the austin factory was producing model Ys at the time. Eventually, tesla planned to make millions of Model 2s. This was the low-cost EV that Elon had promised from the very beginning. Right, the whole idea was you make an expensive car to fund a less expensive car.

01:46:11
And on the way down, at the robo-taxi meeting in late February, musk tore up those plans on the spot. In late February, musk tore up those plans on the spot. He told his team to get busy pushing out the robo-taxi on roughly the same timeline it had previously envisioned for the Model 2. It should be in mass production, the robo-taxi, by the second half of next year. By the end of the week, musk's key subordinates were briefing their teams Model 2 on ice for the time being. In the subsequent hours, word of musk's abrupt strategy shake-up filtered through the company ranks. There was wide confusion about how to carry out the orders. The robo taxi was a concept at the moment, not a vehicle.

01:46:55 - Paris Martineau (Guest)
He turns the barge like it's a speedboat, said a former middle-ranking tesla executive, and it's so funny to me that all this happened because some tesla employees were like oh, we're gonna show uh, elon must a nice concept video of the robo taxi that we're working on in the future. And he apparently lost his mind when he saw it and said all right, we're doing that now.

01:47:19 - Leo Laporte (Host)
You have to really be careful when you show people like Elon anything, because it's unpredictable what he's going to do Anyway.

01:47:33 - Sam Abuelsamid (Guest)
But if you don't show him anything, then he thinks you're not doing anything and then he'll decide that you're dispensable, yeah, which is actually probably a good thing for most of those people, because you're probably better off working somewhere else. For a boss who isn't going to, you know, demand your entire life, uh, you know, be dedicated to the company, because that that's just. You know that's. That's a terrible way to live.

01:47:59 - Leo Laporte (Host)
One good thing, though they did stop any supercharger builds, that the ground hadn't been broken. Yet those are stopped, but the Tesla diner and drive-in movie location on Santa Monica Boulevard is already well underway, and so they're going to continue that one.

01:48:20 - Sam Abuelsamid (Guest)
So if you're in Santa Monica, go get a dinner at drive-in and charge your vehicle. It's great. And there were four sites that were being worked on in New York City, the New York City area, that they canceled. You know they already had the sites located, they had everything set up, with utilities to provide the power, and doing a lot of the design and permitting. They hadn everything set up, with utilities to provide the power, and doing a lot of the design and permitting. They hadn't actually broken ground yet.

01:48:45
But Revel, which is a New York based taxi company that uses Model Ys for their taxi fleet, has been looking to, you know, to have more sites for charging those vehicles. They stepped up almost immediately and said, yeah, we want to take over those four sites and develop the chargers, you know, develop the charging stations there for our fleet. And there's other companies that are stepping up to take on some of the other sites that had been targeted for supercharger locations. Bp Pulse, a division of BP that operates charging stations, is trying to get some of the proposed supercharger sites and there will be others that are doing the same over the next several weeks.

01:49:30 - Leo Laporte (Host)
I also love this Kremlin-like picture from the information of the Tesla executives on stage at Tesla's Investor Day in 2023, 16 executives, five of whom are now gone. Head of Investor Relations CFO.

01:49:48 - Mike Elgan (Guest)
And they'll be erased from the picture.

01:49:50 - Leo Laporte (Host)
Yeah.

01:49:51 - Mike Elgan (Guest)
I want to see the.

01:49:51 - Leo Laporte (Host)
Photoshop version of this. It should be pretty. It's nice because they really spread out, so it should be pretty easy to remove those executives.

01:49:59 - Paris Martineau (Guest)
Yeah, they should have made it a gif, where the people who've been fired dissolve into nothing.

01:50:04 - Leo Laporte (Host)
I should also point out that, of these 16 executives, only two are women, and one of them has been fired. Yeah, good job on diversity.

01:50:13 - Paris Martineau (Guest)
Well, I mean famously. It seems a bit difficult to retain your job as a senior exec at elon's companies, generally, I think, especially if you're a woman. It seems to, I guess, give you a leg up if you're willing to have his children, but that's probably not for everyone that was the neural link ceo right, yes, uh, yeah, neural link ceo.

01:50:36 - Sam Abuelsamid (Guest)
Uh had a pair of twins for him it.

01:50:40 - Mike Elgan (Guest)
It helps. If I was that other woman at Tesla, I'd be wearing a mustache.

01:50:46 - Sam Abuelsamid (Guest)
On the plus side.

01:50:47 - Paris Martineau (Guest)
Mustache yeah.

01:50:49 - Sam Abuelsamid (Guest)
On the plus side for those women you know. It does appear that pretty much all of Elon's kids have been produced via IVF, so at least they didn't have to go to bed with them.

01:51:02 - Leo Laporte (Host)
That's a blessing. I'm just saying yeah. The other thing I know and I only mentioned this because Tesla has been accused of a considerable amount of racism on the floors of the assembly lines and so forth that there are no, and convicted, and convicted. There are no people of color in those 16 executives either. All right, let's take a little break. Much more to talk about Including a victory for the people In the case of the people versus the hell divers too.

01:51:36
But first a word from our sponsor. Oh, this is a good one, because we just installed it. Actually, we, micah, just installed the Eufy Video Smart Lock on our engineering room and it was so easy to do. All it took was a Phillips screwdriver. The Video Smart Lock E330, you can install it with a Phillips screwdriver, no drilling was required. This adds fingerprint recognition to our door. One second unlocking. The embedded AI self-learning chip becomes more accurate with every use.

01:52:15
It's got a removable battery, which gives you four months, but it's easy to recharge 10,000 milliamp hours and, by the way, you'll get a low battery notification before it runs out. So it's no problem. There it is. That's the battery swapping that in and out, and then, of course, you control it with your phone. But here's another nice feature You've got passcode unlocking, you've got remote control with 2K clear sight, you've got two-way audio, enhanced night vision and, guess what? You still have a keyhole. There's the fingerprint, though Watch this.

01:52:50
Micah's showing how easy it was to install and how easy it is to use. And if, for some reason, you want to use the key, maybe the battery died easy to unlock too, so you're never left out in the cold. Battery died easy to unlock too, so you're never left out in the cold. So a great door lock with its own camera, which is useful because, let's say, burke is trying to get into the engineering department. We don't really want to let him in. He's ringing the bell and we're saying oh, it's you, okay, unlock it and look at that, he's in. And if we don't want to let him in, it's easy not to let him in. You can let people get in, you can give them temporary codes. This thing is amazing.

01:53:26
But here's the best part. I kind of buried the lead no subscription, no monthly fees. You store your Eufy video lock recordings locally, so you'll never have to pay for storage and your privacy is insured. You'll get a worry-free experience with an 18-month warranty, all backed by Eufy's 24-7 professional customer service team. Now you may say I don't know the name Eufy, but you do know the name Anchor. Eufy is one of the Anchor brands, so you know this is a great company that makes really good products, solid products, and we love this Eufy video lock. It's been fantastic.

01:54:04
Triple security, a camera, a doorbell and a lock all in one. You can get yours today by searching eufy video lock on amazon eufy, eufy video lock or go to eufycom. By the way, there's a whole Eufy security system for your entire home. They all interlock together and you can have a base station that has additional recording time. So it's really a great system, complete system, everything you need. Search Eufy Video Lock on Amazon. Thank you, eufy, for your support this week in tech. So TikTok has decided they're gonna fight it and I'm glad I haven't talked to you. I think paris, you and I have talked many times on twig about tiktok yeah, tiktok's a constant we love tiktok.

01:54:53
Yeah for for the tunnel videos, for the for all of your son's cooking my son's cooking videos, cooking videos For whatever weird stuff Jeff finds.

01:55:03
But also I feel like there's so much creativity happening with TikTok and I know that the Congress is scared about, I don't know, the Chinese government using it for spying or for propagandizing the American people. So they have said you have 270 days to sell, which is, interestingly, the day before inauguration day 2025. Bytedance has gone to court. They've sued, saying it's a violation of the user's First Amendment rights, removing an app that millions of Americans use to share their views and communicate freely. They also argued we can't do it in 270 days, particularly since Beijing will not allow us to sell the algorithm that powers TikTok in the US. In the 67-page petition which they filed in district court, they said for the first time in history, congress has enacted a law that subjects a single named speech platform to a permanent nationwide ban and bars every American from participating in a unique online community with more than a billion people worldwide. They say there's no question. The act will force a shutdown of TikTok by January 19th 2025. A shutdown of TikTok by January 19th 2025.

01:56:23
I have a feeling that whatever happens in the lower court, this will end up in the Supreme Court. Actually, I got it wrong. It wasn't district court, it's the US Court of Appeals for the District of Columbia where they filed the petition Thoughts. Can they win this one, or is the Supreme Court just going to throw it right out?

01:56:46 - Sam Abuelsamid (Guest)
who knows what this court will do?

01:56:48 - Paris Martineau (Guest)
yeah, yeah, I think that there's like a lot of uncertainty, in part because of the makeup of the court, in part because of how politicized this is, part because of the upcoming election. I think it's very interesting that the consensus from a large number of legal experts seems to be that there's no consensus on what will happen, one, I mean, just this week.

01:57:10 - Sam Abuelsamid (Guest)
you know, alito, and I think, as a Gorsuch, you know we're talking about. You know we seem to have lost respect for the First Amendment. Now, you know they were coming at it from the perspective of, you know, trying to defend. You know right wing speech on college campuses, but still, you know there's some very uncertain views on the Supreme Court when it comes to a lot of, you know, what we, until very recently thought were fundamental rights.

01:57:45 - Leo Laporte (Host)
A visiting scholar, Anupam Chander, at the Institute for Rebooting Social Media at Harvard. We laughed at this on Wednesday. Really, An Institute for Rebooting Social Media? Anyway, he said he's among the experts TikTok contacted on Monday. He said TikTok is going to rely on the fact that they created Project Texas, which was a project to move all American data on TikTok to servers in Texas owned by Oracle. And Chander says the real question that remains that I haven't seen an answer to is why or what more would the government have wanted? We've never heard why the Project Texas was insufficient publicly. We've also not heard I mean, I understand the security concerns, but we haven't heard anything concrete about that. Mike, what's your thinking on all of this?

01:58:43 - Mike Elgan (Guest)
Okay. So I don't think the data gathering element of this is. I just don't see how that could be super useful for the Chinese Communist Party. However, the propaganda dimension is massive, so let's just look at a hypothetical. That's actually very likely. It's very likely that the Chinese government will invade Taiwan sometime in between tomorrow and 10 years from now.

01:59:10 - Leo Laporte (Host)
That's a terrifying prospect by the way that is yes, and so.

01:59:13 - Mike Elgan (Guest)
This is a dystopian, um, orwellian, uh, techno state invading a democracy, and that's the, and, and the us will defend the japan, and south korea will defend taiwan, uh, and so we'll it'll be in a war with China. When that happens, china knows that one of the easiest ways to win the war will be to divide American public opinion. Ever since Vietnam, it's clear that that's how you do it, and it would be fairly trivial to throw a switch in their secret, invisible algorithm. Every Chinese company is required by a law passed in 2017 to do whatever Chinese state intelligence services tell them to do. It's a law in China, and so this is a problem, right, and?

02:00:08 - Leo Laporte (Host)
we lost your maybe that's me. Yeah, okay, a problem right. And we lost your maybe that's me. So, yeah, okay so.

02:00:14 - Mike Elgan (Guest)
So this is, um, this is like, sorry, the the problem with this, because I agree with the first amendment stuff. I agree with the the panic about tiktok. I agree with all that. I want tiktok to exist and this taiwan scenario is going to happen.

02:00:29 - Leo Laporte (Host)
That's scary and you're right, that's very scary. And now let me but okay, let me say this mike, doesn't the chinese government have many other ways to do this, including twitter and uh cctv, which is available in the us on most cable systems and face? There are many, many ways that the chinese because we're an open society that the Chinese government can reach American public opinion.

02:00:55 - Mike Elgan (Guest)
Right. The particular is that everybody who works on social algorithms will tell you that TikTok's the best one. I mean, it's the most powerful algorithm for determining what.

02:01:10
But the other issue, the larger issue, is that algorithms in general are problematic. For this reason, a lot of our the problems in our politics the ability for marjorie taylor green, for example, is that it raises more money than any other congressperson for small donors. Did you guys know that? No, she's, she's number one. Why is that? Because she feeds the outrage machine on social, and those algorithms just reward her with people who want to give her money. The divisiveness in our culture, so many of the things we don't like about the modern world, are the result of algorithms that nobody can see, nobody can check that. They tweak all the time. They tweak it. You know Met meta tweaks their algorithms 10 times a day and these algorithms are. I don't know what the answer is, but this is a sort of unspoken problem that we have in our culture, and TikTok is one potential big problem.

02:02:06 - Leo Laporte (Host)
But if they proposed, if Congress proposed or somebody proposed, we do the same thing. Because twitter is used that way and that's, by the way, how marjorie taylor green reaches her constituency it would immediately be shut down on first amendment grounds, right, no, sorry, government can't shut down twitter, so I think it's also there's some aspect of this that feels a little bit like thought crimey, in the sense that, yeah, it does seem likely that the Chinese government could use, and will use, all tools available to it to distribute propaganda.

02:02:54 - Paris Martineau (Guest)
start shutting down systems that exist currently on the idea that it might be used in the future, or should we take a more?

02:02:59 - Mike Elgan (Guest)
reactive approach? I have no idea. I just think it's really interesting that we've got, you know, go back in time to the mid 20th century and democracies had all the advantages in the global conflict between authoritarian societies and democracies, and now it's the reverse. So TikTok is not allowed in China. They have their own version of TikTok which is nothing like the version the rest of the world gets. Twitter is banned, facebook is banned. They're all banned. They control that and so they have. They can change public opinion globally, but nobody can touch public opinion within China, and it's a real problem. I think they, I think the.

02:03:41
The answer is not to ban Twitter, a TikTok, but to figure out this thing we've been dancing around for forever, which is which is the? What is a social network? Is it a publication? Is it a common carrier? What is it? And I think there's got to be a solution in there somewhere where you could imagine a law and I'm not actually proposing this, I just think smart people should figure something out where part of a social network is a common carrier and part of it is like a publication, so the algorithmic sorting thing.

02:04:14
We all throw our hands up and say, well, we have to let them do that part. We have to, but why? Why do we have to let them do that part in the way that they do it? That's a determining. What everybody thinks is the power that, collectively, social networks have, and it's not clear that that's something that people like Elon Musk should be in charge of. It's not necessarily good for society. So you can imagine where there's a law where the default mode is you follow I follow 20 people on Twitter and the default mode is I see every single post that those 20 people post, no algorithmic sorting, filtering or anything. Then there's another channel that I can click over to and it's all. Whatever the social network wants to show me, for whatever reason, they have something like that, but nobody's even talking about it or proposing anything like that anymore. We're just throwing our hands up and just letting Russia decide what everybody thinks about the Ukraine-Russia conflict. Basically, I'm exaggerating, of course, but there's.

02:05:18 - Leo Laporte (Host)
I just feel like this is the risk of being in a free society this is the downside to being in a free society is that we're open. We're open to all sorts of ideas, including ideas that are attempting to undermine our free society.

02:05:36 - Paris Martineau (Guest)
I don't think the way to respond to that is to close down our freedom I also wonder this reminds me in some way of everything that went on with cambridge analytica is. For a period there was this narrative that Cambridge Analytica's understanding of you know, social networks was so powerful it swayed the minds of these humble social media users. But as we started to look into it and as the years went on, we realized that that just wasn't the case. Who saw these sort of posts? Or even with, I guess, the various Russian botnets, people didn't end up being that dramatically or even remotely influenced by these propaganda operations like we'd initially feared. And I do wonder if in five years from now, we'll look back on this TikTok panic in the same light.

02:06:32 - Mike Elgan (Guest)
Is it correct?

02:06:33 - Paris Martineau (Guest)
that TikTok's algorithm is all that powerful.

02:06:35 - Mike Elgan (Guest)
I'm not sure. I don't know that. We have the data to know one way or the other whether say Russia, which invented this whole thing how effective they are at changing American public opinion. Would you know that they actually have scheduled, have organized opposing? You know sock puppet accounts, have organized protests pro and against in the same location to get people fighting together, right, we know that they've done that, like they've actually mobilized people and got them into the streets protesting, and they managed it in a way that would to maximize conflict, and we know that that's their aim.

02:07:12
China's increasingly getting in on that. Iran would love to do that as well, so probably others. So so I don't think it's, you know, I think I think it's kind of more of an urgent question than than people will think, and I think we need a lot better data about it. And I think we tend to think we need a lot better data about it. And I think we tend to think you know, those of us in tech have seen a million moral panics tend to think, well, it's another moral panic.

02:07:37
But I don't think we know whether it's just another moral panic or if, in fact, this is a really the ability of authoritarian regimes to to change public opinion, to maybe change the outcome of elections. We know they're trying, right, they tried, they tried before they try, they're trying again. We know that. For for a fact um, are they just wrong? That you know that that there's no point in doing that because it doesn't have any impact? I don't know. It's, it's, uh, it's something we don't know enough about, and I would love for us to get serious about really figuring out how effective this stuff is.

02:08:12 - Sam Abuelsamid (Guest)
Is part of the answer to this. You know requiring algorithmic transparency so that we can actually take a look at how they're manipulating the content.

02:08:22 - Leo Laporte (Host)
First of all, you can ask it. It's not going to happen, right? So that's like asking Coke for their secret recipe. It's just not going to happen. It's asking Coke for their secret recipe If Coke were run by the Chinese Communist Party. It's just not going to happen. It's asking Coke for their secret recipe If Coke were run by the Chinese Communist Party. It's just not going to happen.

02:08:37 - Mike Elgan (Guest)
But even if we had it, we do regulate ingredients in things like Coke. Sure, there's a list of banned ingredients, and I think that there's an instinct among a lot of people to defend the right of companies like Meta to have their algorithms, and I think we need to revisit that Like why no? I agree.

02:08:59 - Leo Laporte (Host)
Do you want to ban algorithms, though? What do you want to do with it?

02:09:03 - Sam Abuelsamid (Guest)
Not ban them, just let us know what they're doing no-transcript to imagine how that algorithm might work.

02:09:36 - Leo Laporte (Host)
That's not the solution, because the thing you're missing from that ingredient is the Chinese secret service coming in and saying oh, and, by the way, every fourth one should be this video or should be this kind of video. So knowing what the algorithm is doing is not. It doesn't solve the problem. The problem is that the content can be manipulated by an adversary, but, as I'm saying, the content of twitter can also be manipulated by an adversary. The content of everything can be manipulated because we're an open nation and we let people say what they want.

02:10:15 - Mike Elgan (Guest)
I don't think algorithms have First Amendment rights. I don't think. But the algorithm is not the problem.

02:10:21 - Leo Laporte (Host)
That's not the problem. The algorithm isn't. The problem is that these platforms allow manipulation, not through an algorithm. The algorithm is not. The manipulation is the Chinese. The manipulation it's the Chinese secret service saying those videos see that one, that one, that one promote those.

02:10:38 - Sam Abuelsamid (Guest)
That's not the algorithm or you know or or the you know the owner of the company coming in and saying you know promote, you know make sure that all, all, the all, the all, the all the posts from this particular user who happens to be the owner of the company you know, must be seen on everybody's feet.

02:10:57 - Leo Laporte (Host)
That's not an algorithm, so, but you know, I've long, for a long time. I've said algorithms are a problem. I think the YouTube recommendation engine promotes an increasingly extreme video. That's pretty obvious, right? So the I agree with you. Algorithms are potentially problematic. They're, you know, for, I think, all companies. They're tuned for one thing to increase engagement, hence increase profit, and that's potentially problematic. You can outlaw algorithms. I don't know if that's doable, given our constitution.

02:11:27 - Benito Gonzalez (Other)
Hi, this is Benito. I think this is a demand side problem, not a supply side problem.

02:11:32 - Leo Laporte (Host)
I agree what you're saying. It's us, not them. Yeah, it's us, it's that people like it. That's the problem. People like it. They want an algorithm. When you have a choice on Facebook between your chronological feed or your algorithmic feed, people want the algorithmic feed. Same thing on Twitter they want the algorithmic feed.

02:11:55 - Benito Gonzalez (Other)
We say I pay lip service to the idea that I want a chronological and me personally.

02:11:57 - Leo Laporte (Host)
I don't use any of the algorithmic feeds in any of the stuff I use because I hate it. But they're not.

02:11:59 - Mike Elgan (Guest)
It's not as engaging but you're far from but you're far from average. You're not people. You people used to love the beef tallow that mcdonald's used to make their french fries. They were way better I still do.

02:12:08 - Benito Gonzalez (Other)
You know they were because we don't still love it I wish they'd bring that back well, you know I mean one of one of the one of the.

02:12:18 - Sam Abuelsamid (Guest)
You know, the main issues here with tiktok is that we still haven't gotten an explanation, you know, from the, from congress, as to why they have pushed this.

02:12:29 - Leo Laporte (Host)
I understand, but let's stipulate, let's stipulate that mike's uh point, which I think is very well taken, is what they're thinking about. There's going to be an invasion, china's going to use this to manipulate the american electorate.

02:12:41 - Sam Abuelsamid (Guest)
You get 10 of the american electorate to say we don't want this war, you got a problem and and I agree with mike, so even let's just stipulate that that's it and maybe, and maybe you know, anthony posted in the chat a link to an interesting article from Axios Mitt Romney saying that support for the TikTok ban is actually connected to an overwhelming volume of mentions of Palestinians on the platform. Maybe, I don't know, that one seems less plausible, but here's the thing. Even what else has been happening, you know across the country at campuses and you know governors. You know paris, yeah, sending out the.

02:13:23 - Leo Laporte (Host)
You know I don't think that's it. Read that, uh, read that.

02:13:27 - Paris Martineau (Guest)
National institute of health this is somewhat of what I was talking before. This is a study that I linked to in the chat and it specifically goes into trying to identify whether there is trying to address the widespread concern that foreign actors are using social media to interfere in elections worldwide. They analyzed a bunch of stuff relating to uh, russian foreign influence campaigns in the 2016 election, cross referenced this with individual voters. It's obviously a small sample, but the one thing they found that was interesting was that they said we find no evidence of a meaningful relationship between exposure to russian foreign influence campaigns and changes in attitudes, polarization polarization or voting behavior.

02:14:13 - Leo Laporte (Host)
It's a supply side problem, I mean a demand side problem. Benito's right it's us. We've met the enemy and he is us. Pogo was right. Yeah, I think that that's I think I've always thought that I've always for a long time. There was a lot of people said, oh, the reason Trump won in 2016 was Russian propaganda, and I think it was pretty obvious.

02:14:37 - Paris Martineau (Guest)
It's very simplistic, it's a way to, it's a pat explanation for something that's much, much bigger.

02:14:44 - Leo Laporte (Host)
It's a little patronizing, oh, absolutely. It's implying that the electorate, you know, is easily swayed and I think the people voted for trump, sincerely voted for trump. I think that, you know, the hundred thousand people that showed up in new jersey yesterday were sincere, in their interest, in seeing what he had to say. I don't think it's fair to say well, that's a propaganda. I never thought that was it. Uh, that means we have to address those particular people and their issues. It doesn't.

02:15:13 - Paris Martineau (Guest)
I do think it's important to look at this TikTok ban in the context of everything that has been going on with regards to. We are coming off a period of great unrest and general dissatisfaction and anger towards major tech companies. It has become a major sticking point for both parties, a rallying cry for various purposes that they want to crack down on big tech. And it's very difficult to do that when the big tech companies are American, have powerful lobbyists and are also used by everyone.

02:15:49 - Sam Abuelsamid (Guest)
But the Chinese, on the other hand.

02:15:51 - Paris Martineau (Guest)
I think in some ways the Chineseok is an easy scapegoat, because then it also fits the pattern of china and it doesn't solve the problem of privacy it doesn't solve the problem of propaganda, it just.

02:16:02 - Leo Laporte (Host)
It just makes people feel better go ahead, mike.

02:16:05 - Mike Elgan (Guest)
The chinese government also lobbies con us congress and they've lobbied hard against TikTok ban and have done lots of other things to really push against the TikTok ban. Which gives me the creeps, like, why are they so? They claim that, oh, there's no connection whatever, but why would the Chinese government be so adamant that TikTok exist and be legal in the US? Okay, kind of gives me the creeps.

02:16:31 - Leo Laporte (Host)
I agree.

02:16:37 - Mike Elgan (Guest)
But, mike, do you think a TikTok ban will solve any of these things that we're talking about? I think that I think we have a problem with disinformation, and I think we have a problem with foreign state actors who run, you know, sort of authoritarian, dystopian governments, pushing behind the scenes disinformation. The degree to which it's effective, we don't know for sure. There's data on both sides, actually, and I think we really need to figure that out. And if we do decide that that's a risk, if we do decide that it's possible for the Chinese government to just say here's what we want American public opinion to be and just basically tweak it five percentage points, if that's possible, then part of the solution may be doing something about TikTok. And again, like you say, even if TikTok was owned by somebody who was not a Chinese company, it would still be a great place to sort of promote disinformation, to gain the algorithms and all that kind of stuff that they already do on Twitter and elsewhere.

02:17:41 - Leo Laporte (Host)
I don't think it's going to fix it.

02:17:42 - Paris Martineau (Guest)
I think also something that's worth exploring a little bit here is how a TikTok ban would even work. I think someone asked in the chat like, let's say, tiktok is banned tomorrow, how would that work? Would I be arrested for using a VPN to access TikTok? Would ISPs be held liable? I think, from what I've gathered from just doing cursory research on this, it would be more that TikTok would suddenly be inaccessible on the Apple and Google app stores, so new users wouldn't be able to download it. But would that mean that all 170 million Americans who have it on their phones would still be able to use TikTok?

02:18:17 - Leo Laporte (Host)
It'd be enough to eliminate TikTok in the long run. I mean that's. Even this conversation is already harming TikTok because advertisers will move away. Creators like my son will immediately find an American company doing exactly the same thing to put his content on Instagram and YouTube. Yeah, great place to follow. And I think but that's my point is this improves nothing. Privacy is not being protected. The American public is not being protected as if it could be. It's it's theater. It's not a serious attempt. It's a it's a political attempt to score points. It's not a serious attempt to solve the problems.

02:18:54 - Benito Gonzalez (Other)
Like if you want to talk about the real issues here. It's like in America. It's not so bad because it's not homogenous. It's not everybody's using the exact same thing, but there are things like you know. Facebook's effect on Myanmar, let's say.

02:19:06 - Leo Laporte (Host)
Oh, I agree, that was pretty bad.

02:19:08 - Paris Martineau (Guest)
Oh, I agree.

02:19:10 - Benito Gonzalez (Other)
Stuff like that, though, like globally, I think there's a bigger problem In America. I don't think that's as much of a problem. Indonesia, yeah. Philippines, yeah, it is bigger also.

02:19:18 - Mike Elgan (Guest)
And also Telegram and like there's all kinds of problems with gaming, public opinion messing with elections.

02:19:27 - Leo Laporte (Host)
I want to take a little break and when we come back I'll show you how Twitter is being used to promote a Russian agenda in a particularly revolting way, with the cooperation of Elon Musk. But that's coming up in just a moment. Our panel, mike Elgin Great to have you His new newsletter, which I have just subscribed to and you must as well. Machine society dot ai. You're going to focus on on ai, or is it going to be mike's like mike's list? What are we going to see here?

02:19:59 - Mike Elgan (Guest)
it's, it's, uh, it's ai, but also, you know, spatial computing, things like that, next, next generation super high technology. That, yeah, changes how we interact with each other changes. Love it, yeah, love it. So thank you for that, leo and made out of people Machinesocietyai.

02:20:20 - Leo Laporte (Host)
Sam Abul-Samad is a car guy too. He's got I don't know 10W40 flowing in his veins, I don't know. His podcast Wheelbearingsmedia is awesome. If you love cars, you've got to listen to that with Robbie and Nicole and Sam. And of course he's a regular on our network. We love having him on. In fact, he was just on Ask the Tech Guys earlier today and principal researcher at GuideHouse Insights. Very nice to have you, sam. Do you have a newsletter, I guess, at GuideHouse Insights. Very nice to have you, sam. Do you have a newsletter? I guess at GuideHouse you probably have an analyst letter, right?

02:20:56 - Sam Abuelsamid (Guest)
Yeah, we have. There's a blog on the GuideHouse Insights website. You can find that there. I think it's under the latest insights tab on the top of the page, where all of the analysts contribute their thoughts on there and the reports that we publish and white papers that we do. So there's lots of stuff to find there.

02:21:20 - Leo Laporte (Host)
I'm always learning more about you too, which I really enjoy, because earlier today I found out that you lived in North Suite. How long were you in the North? I didn't live there, you worked there.

02:21:32 - Sam Abuelsamid (Guest)
You lived in North Sweden. How long were you in the North? I didn't live there.

02:21:34 - Paris Martineau (Guest)
You worked there.

02:21:36 - Sam Abuelsamid (Guest)
When I was working as an engineer in the 1990s, I spent several winters up there up in northern Sweden, northern Sweden, the area around Arvidsjar and Aryaplug, which is near the Arctic Circle, is a common area where a lot of automakers and suppliers do cold weather testing throughout the winter, because it gets cold fairly early, you know, you get snow and the lakes freeze over and it lasts until well into April typically. So it's a good place where you can spend several months doing winter development, and so I've spent a fair bit of time there during my engineering days.

02:22:15 - Leo Laporte (Host)
I've learned every time it's beautiful. It's gorgeous up there. It's cold, though, yes.

02:22:21 - Sam Abuelsamid (Guest)
Actually not that crazy cold. You know, I found it actually to be much colder when I spent time in northern Minnesota, but it stays more consistently cold when I spent time in northern Minnesota. Oh, all right, but it stays more consistently cold. So, instead of getting to 40 below zero, it might be in the teens or upper 20s, but it'll just stay that way for weeks on end.

02:22:44 - Leo Laporte (Host)
That's cold. I'm sorry but that's cold. Thank you, sam, it's great to have you. And, of course, paris Martineau great piece on jonah peretti interview with the man who founded buzzfeed, from that before then, huffington post, and his vision for news and where he's going now that buzzfeed is struggling was very interesting.

02:23:04 - Paris Martineau (Guest)
I mean kind of germane to our conversation. The future of BuzzFeed is allegedly generative AI.

02:23:13 - Leo Laporte (Host)
How depressing, how depressing, but this is where the good work. Now you've got to find a new subject to work on.

02:23:22 - Paris Martineau (Guest)
Listen, you're telling me. Any people out there. You work for a tech company. You think there's an interesting story going on or something I should look into? My signal is in the lower third here. It's martineau01 on signal.

02:23:37 - Leo Laporte (Host)
Reach out to me give her the tip that will create the next long read on the informationcom. Is that how it works, or do you? Do you go out looking for problems?

02:23:49 - Paris Martineau (Guest)
I will say I've gotten. I mean, I've a very unique period in my career over the last year or two, because I had previously not been that tips based. I would generate almost all of my own story ideas fresh from the dome, but over the last year I've gotten some fantastic tips, often from twit listeners see resulted in so many great stories blow the lid off your company signal.

02:24:14 - Leo Laporte (Host)
martin 0-1, that's all I gotta say. Thank you, paris. Our show today brought to you by zip recruiter. Have you ever set your phone alarm? I have to buy tickets the instant they go on sale, you know, because if you want those great seats, you better buy them the minute they go on sale. Buy them months in advance, sure. When you want the great seats, you better buy them the minute they go on sale. Buy them months in advance, sure.

02:24:34
When you want the best, you got to act quickly or somebody else will get it instead. Same thing when it comes to hiring for your business. You want the best people for your open roles. You want to get them before the competition swoops in, and there is real competition these days for great employees. What's the best way to do it? Zip Recruiter. Zip Recruiter finds qualified candidates fast Right now. You could try it for free, by the way, at ziprecruitercom slash twit. Ziprecruitercom slash twit.

02:25:03
How do they help? Well, they have a lot of different ways to bring your listing to the top, but here's one I love they use their powerful matching technology to identify top talent for your open roles the minute you post your job. Ziprecruiter's smart technology goes through all the. You know they have million, literally more than a million, current resumes because people come to ZipRecruiter looking for work. It sorts through them, looks at your qualification requirements, looks at their qualifications, then starts giving you a list of people who look like they'd be right for the job. Now you can go through that list, you can screen them, you can sort them and then you can invite the people you want the most. You want to beat those other guys out for you, invite them directly and I have to tell you that brings you right to the top of the heap. People love it when you say, hey, you look like you'd be great for this job. Can we talk? That's going to help you a lot. Get to the front of the line.

02:26:01
Amp up your hiring performance with ZipRecruiter. Find the best people fast. We know because we use it it works. See why four out of five employers who post on ZipRecruiter get a quality candidate within the first day. For us it's usually within the first hour or two. Just go to our special address right now ZipRecruiter. You can try it for free. Ziprecruitercom slash twit. Ziprecruitercom slash twit. Ziprecruiter really is the smartest way to hire. We thank them so much for many years of supporting our shows. We're very grateful.

02:26:37 - Paris Martineau (Guest)
thank you, zip recruiter also a brief note that I just realized as I was reading, that is, the periods. Unlike in google or gmail addresses, the periods matter. So my signal is martineau dot zero. Oh, they don't let you. They don't let you, not include the dot in there, which I think is put a dot in there. Benito, we've got guys, we got it, no wonder you haven't been getting any tips, let's get those tips.

02:27:03 - Leo Laporte (Host)
Kids, martineau dot zero one there we go guys benito's gonna fix that. If you were a follower of Elon Musk on Twitter, you might say well, signal is just broken right, we should all be going to the secure telegram.

02:27:21 - Paris Martineau (Guest)
Yes, mr Musk.

02:27:22 - Leo Laporte (Host)
Well, you know. Of course, as soon as Elon said well you know, there's people don't people don't know, but there's security flaws in Signal. No, there's people don't people don't know, but there's security flaws Signal. Susie said that Meredith Green is what's her name?

02:27:35 - Paris Martineau (Guest)
Meredith Whitaker.

02:27:36 - Leo Laporte (Host)
Whitaker at Signal. Yeah, the CEO responded saying what? No, there are no security flaws. It's open source. There are no CVEs active. We fix everything. What are you talking about?

02:27:48
Elon Matthew Green, who I fully respect as an independent cryptographer he's at Johns Hopkins very smart guy on Twitter, by the way, at Macedon as well says Telegram has launched a pretty intense campaign to malign Signal as insecure, with assistance from Elon Musk. The goal seems to be to get activists to switch away from encrypted signal, which is secure and reliable, to mostly unencrypted telegram. Now we should mention telegram was created by Pavel Dourov. They used to call him the Mark Zuckerberg of Russia. He's Russian. Now I don't know what his relationship is with Putin because he left. He fled Russia after the Russian government forced him to sell his company to them then founded Telegram. He's in Dubai right now, but we should be very clear A, as Matthew is, telegram does not end-to-end encrypt conversations by default, unless you manually start an encrypted secret chat.

02:28:58
All of your data is visible on the Telegram server. Given who uses Telegram, matthew writes, this server is probably a magnet for intelligence services. We should also point out Telegram uses a roll-your-own encryption which has never been fully vetted, whereas Signal's code is completely open source. You can download it, examine the code, examine the crypto libraries, matthew writes, even if you don't want to do that, many experts have. That doesn't mean there's never going to be a bug, but it does mean lots of eyes going to be a bug, but it does mean lots of eyes.

02:29:36
Pavel dura, of the ceo of telegram, matthew, goes on to say, has recently been making a big conspiracy push to promote telegram is more secure than signal. This is like promoting ketchup reference. Yes, this is like promoting ketchup is better for your car than synthetic motor oil. Is that true? I don't think so. Telegram is not a secure messenger, says matthew. Full stop, okay, so uh, yep. This is an example of how any social network, even an american own and run social network, can be used to promote uh, anti-american or anti-activist, uh bs there's a lot of anti stuff on twitter.

02:30:20
Oh, it's just one tip of the iceberg, obviously. Yeah, signal is safe. Use signal. Do not use telegram. I like telegram. I wish everybody could use telegram. I wish all my friends would, because I would use it with everybody, because it's fun, it's got stickers. I love it, but it's not secure. Don't kid yourself that it's secure.

02:30:42 - Paris Martineau (Guest)
And the one caveat I've got to give to people about Signal, which I agree entirely secure. The one thing that will make it a little less secure is if you're trying to do something like maybe leak sensitive information about your company and use signal on your work phone. Surprise, surprise, your company can still see that a lot, as a lot of apple employees over the last couple of years have realized this is good.

02:31:05
You do have to a little bit of opsec, basically just basically you gotta have a phone that isn't owned by your employer it's not secure until it's sent out over signal, but when you're typing it in your company can see it.

02:31:19 - Leo Laporte (Host)
Okay, that's very good, I like it. Uh, let's see what else. Oh hell, divers 2 we talked about this last week. Going on with this. This is a great story story what is Helldivers 2?

02:31:32
It is the game of the year. Of course, we're only in May, but the game of the year. An independent company created it. It was on Steam. Sony, a couple of weeks ago, announced that in order to play it, you have to have a PlayStation Network account and you have to log into it, which meant that people in many countries could not play it because psn was not available in those countries, which meant that the reviews on steam for hell divers 2 went from extremely positive to extremely negative. We we showed them last week. I can't show them now because it's back. It's better because Sony reversed its decision. The people won. In this case, it was something like 500,000 negative reviews.

02:32:25 - Mike Elgan (Guest)
And then the people came back and brought the reputation back up, right back With positive reviews. They responded, which is amazing.

02:32:32 - Leo Laporte (Host)
So cool. Yeah, this is, I guess, in a way, a positive story, isn't it? Let me go to the, because I was going to show you I showed you last week the bad reviews and how bad it was. Now, on Steam, mostly positive, 627,000 reviews and a lot of the newer reviews say things like for the love of liberty, democracy has prevailed For democracy. Sony, really, as often is the case with Sony.

02:33:08
Consumer activism let's screw it up, and the consumer activism has won. See, maybe Steam is the social, with Sony. Consumer activism let's screw it up, and the consumer activism has won. See See, maybe Steam is the social network of the future. It's going to save us and democracy. Please no, no.

02:33:22 - Sam Abuelsamid (Guest)
What could go wrong?

02:33:23 - Benito Gonzalez (Other)
Well, I mean, the thing about this story is like yes, it all sounds good like the people won, but this is because we gave up 1,000 rights 20 years ago.

02:33:30 - Leo Laporte (Host)
It's true that they were able to do this, even right. We lost a long time ago, guys.

02:33:35
You know what's sad, benito, to me? This is Benito Gonzalez, our technical director, and producer and editor produces the show. That's really the important job. What's sad to me is that video games in the early days were highly copy protected. They would do all sorts of really. They'd punch a hole in the disc and things like that to make sure you couldn't copy it. But then they realized this is really bad for business. We are making our legit customers into pirates and not slowing down piracy at all. And so they stopped.

02:34:09 - Paris Martineau (Guest)
I'm curious have they stopped though? Well, no, then they turned there's often and now it's back.

02:34:15
It's back. There's a lot of games where you have to have online connection for anti-piracy bs. I just saw anthony posted in the chat that sony is kind of taking a similar playbook to other games where, like people who'd pre-ordered the pc port of a one of, I guess, an upcoming game in countries they don't have access to the playstation network are being notified that their order has been canceled and they're getting auto refunded. It seems like the privacy and piracy controversies just keep getting worse, which is sad for gamers like me.

02:34:53 - Leo Laporte (Host)
Oh, well, are you going to get the new nintendo switch when it comes out later this year?

02:34:58 - Paris Martineau (Guest)
I'm not, I'm steam deck pilled oh and what do you play on steam?

02:35:02 - Leo Laporte (Host)
you're not playing animal animal crossing anymore.

02:35:04 - Paris Martineau (Guest)
Huh no, I'm playing. Actually I just finished a replay of a game I first chatted out on twig in november called pentament, uh oh I want to play that so bad okay, you should, jeff should, everybody should.

02:35:19
It's a really cool game where you're like a, a small artist, in a bavarian alpine town in the 15th century. Uh, you're there to work on your masterpiece and as like a little illustrator in an abbey, and then a murder happens. It it's been described as uh the name of the rose meets uh night in the woods, which is like a kind of indie game recently I can't wait to play this murder.

02:35:45 - Leo Laporte (Host)
She wrote yeah, and look, I love the graphics. It's yeah it's beautiful.

02:35:49 - Paris Martineau (Guest)
Yeah, I literally before I got the email from benito asking for me to join this emergency twit. I wept today because I finished the game. It was quite sad I mean, it's just a very emotionally moving game how long does it take you to finish? It.

02:36:07 - Leo Laporte (Host)
It's pretty short, it's like 20 hours well, it's 20 bucks, so 20 hours, 20 bucks, that's still a deal.

02:36:12 - Paris Martineau (Guest)
And that was my like second uh playthrough of it too, because there's a lot of choice. It's one of those like choices matter games, um. So I went back and did a couple different things this time, which was fun, but I know that you have to get out of here in five minutes so we're gonna wrap this up pretty quick.

02:36:31 - Leo Laporte (Host)
I have a couple of things I want to ask you about. Number one were you offended by the apple ipad crush ad? So when apple announced its new extra thin ipad, they showed a video that took all the things we love a piano paintings, all the human creations, synthesizers and crushed them with a hydraulic press. And then, when the press lifted up, all there was was this really thin iPad, apparently this upset people a lot, did you see it?

02:37:13 - Sam Abuelsamid (Guest)
It's stupid and tone deaf.

02:37:15 - Leo Laporte (Host)
It was pretty tone deaf. I saw it. It didn't bother me because we were watching the event, but I understand what people are saying.

02:37:24 - Mike Elgan (Guest)
The problem is that that's what the iPad is.

02:37:27 - Leo Laporte (Host)
It's kind of too on the nose no-transcript.

02:37:58 - Paris Martineau (Guest)
You have one of the most valuable tech companies in the world putting out an ad where all of those creative fields are symbolically and physically crushed.

02:38:07 - Leo Laporte (Host)
Benito, you're a DJ. It starts with a turntable. It's got many of the tools you use. Did this?

02:38:12 - Benito Gonzalez (Other)
bother you, an iPad is not replacing a trumpet. It starts with a turntable. It's got many of the tools you use. Did this bother you? An iPad is not replacing a trumpet. I'm sorry, that is never, ever happening. Or a piano is never happening.

02:38:19 - Paris Martineau (Guest)
What about that trumpet game the week ago? Doot, doot, doot. Somebody fixed it on Twitter.

02:38:24 - Leo Laporte (Host)
Miles. Davis ain't playing iPad.

02:38:26
Ran it in reverse, Ran it in reverse and so it started with the iPad and then it lifted up and all of these things were there. That made a lot more sense. Apple did apologize, but notice it's still on YouTube. I'm watching it right now on YouTube. Apple said creativity is in our DNA at Apple and it's incredibly important to us to design products that empower creatives all over the world. That empower creatives all over the world. Our goal is to always celebrate the myriad of ways users express themselves and bring their ideas to life through iPad. We missed the mark with this video and we're sorry they will never run it on TV. I'm a little turtle. I don't know who it was, but yeah, that's close.

02:39:10 - Paris Martineau (Guest)
Okay so, but it's a tempest in a teapot, but I understand and upset yeah it does feel kind of quaint to have a news cycle of people getting mad about an ad, though it's true, it's true.

02:39:23 - Leo Laporte (Host)
Uh, paris martineau, you are the best ever. Thank you for being here. I'm sorry we're gonna miss you this week, on this week Week in Google, because we are moving this Week in Google to Tuesday so we can cover Google IO and apparently that's when your skee-ball club practices.

02:39:39 - Paris Martineau (Guest)
Yeah, you know, my skee-ball practice is at 1 pm Eastern on a Tuesday. It's definitely not because I have a job. Oh, you're going to work. Yeah, I have this whole thing called a salary job.

02:39:52 - Leo Laporte (Host)
Oh well, okay, All right, but we will be covering Google IO on Tuesday Without you. Sad to say, jeff and I we're doing this Week in Google at 10 am. We also got Alex Kantrowitz and Alex Kantrowitz of the Big. Technology Podcast. That will be a lot of fun. Thank you, benito, for scoring that, so he's going to replace you, but you'll be back on this Week in Google, a week from.

02:40:13
Wednesday yeah, you can't keep me away and if you want to join, you could. If you want to join, our Macintosh show, which is going to be airing at MacBreak Weekly, be airing in the this Week in Google time slot. If you want to join and talk about the new iPad, you're more than welcome. I'll let you know. Are you getting the new iPad?

02:40:33 - Paris Martineau (Guest)
No, no, I have an old iPad and it's fine.

02:40:38 - Mike Elgan (Guest)
She has a trumpet.

02:40:39 - Paris Martineau (Guest)
I've got a trumpet, so I don't really need an iPad. I don't need an iPad. It already does everything you crushed it.

02:40:45 - Leo Laporte (Host)
Thank you, paris. Thank you to Mr Sam Abul Samad Wheelbearingsmedia twit social. You've spent a lot of time today with us on the twit. Thank you for your many hours.

02:40:58 - Sam Abuelsamid (Guest)
My pleasure, always fun chatting with you guys. Always great to get your insight, glad to get to chat with Paris as well, always a pleasure.

02:41:08 - Leo Laporte (Host)
She's been a great addition to twit. I agree, I agree. Thank you, sam, and thanks to Mike Elgin. Congratulations on the new newsletter, machinesocietyai, and let's not forget your son's Hello Chatterbox, that's right.

02:41:24 - Mike Elgan (Guest)
Teaching AI to children? Yes, and it's Hello. Chatterbox is the only smart speaker that's legally allowed in schools because it's completely private. He has no way of knowing even who the user is, nor do any of the services they use, so it's really great for education. So if you're an educator, I would get on that. And another thing, leo the gastronomic experiences, which are wonderful, ai-free, technology-free yes free.

02:41:53 - Leo Laporte (Host)
It's made of people and food additional. It's the best traditions and and I very old school stuff. Lisa and I went to oaxaca with mike and amira on their gastronomat adventure and it was amazing. I've made such great friends from it and I can't wait to go on another one.

02:42:09 - Mike Elgan (Guest)
You got one coming up, yes we, uh, so this is the first time I've talked about this publicly we just put this up. So we're doing, uh, we're launching a new one in sicily okay, okay, sicily gastronomic experience.

02:42:22
I want to go. Uh, we love sicily and it's just a beautiful food culture there. That's arabic food, spanish food, greek, phoenician. You go, you name it. It's a great place. So really excited about that, and I also want to tell people who are interested in doing these experiences that they're feeling up earlier and earlier, so it's like eight, ten months before the experience they are filling up.

02:42:45 - Leo Laporte (Host)
You know why? Because people go on. These ones keep coming back. It's such an amazing experience. Yes, yes, and so your real problem is it's only you know a few couples and basically you know, as soon as you go on one, you go okay, I'm booking the next 10. Thank you, I want to do this one. I really want to do this one.

02:43:04 - Mike Elgan (Guest)
We had a guy who, um, who did, who did one, and then he did another one and now he's got 10 couples that are friends of his that are all doing it and it's just, it just keeps snowballing like that. And it's so small because it's just.

02:43:17 - Leo Laporte (Host)
Mike and Amira and the chefs and the restaurants and the culture that they know so well. Uh, all over, the winemakers Love it Gastronomadnet. I guess Sicily will be a lot of good wine. Can't wait. A lot. Thank you, Mike. Thank you, Sam Paris, get out of here. You got to go Go play that ski ball thing.

02:43:41 - Paris Martineau (Guest)
I got to go watch, get to a 945 showing of the movie in Bruges, so thank you. Oh, I love that movie. It's fantastic.

02:43:49 - Leo Laporte (Host)
I'm excited to see it in the Cuba Theater. Oh, Bruges is so beautiful. Well, have a great show. That sounds like fun.

02:43:56 - Paris Martineau (Guest)
I shall, Mike. When are you going to do?

02:43:57 - Leo Laporte (Host)
Bruges for Gastronome Ad Adventure that would be awesome.

02:44:00 - Mike Elgan (Guest)
That's a great idea. Maybe next month? No, I don't, it's the most beautiful, beautiful town, adorable In Belgium.

02:44:10 - Leo Laporte (Host)
Thank you all three of you. Have a wonderful evening and thanks to all of you for joining us. We do this Week in Tech every Sunday at 2 pm Pacific, 5 pm Eastern, that's 2100 UTC. You can watch us do it live. We turn on the live streams right when the show begins on YouTube YouTubecom slash twit slash live on YouTube. Youtubecom slash twit slash live and I think all the shows we do that with, including tomorrow's 10 am broadcast of the OpenAI keynote. Jeff Jarvis and I can't wait to see what OpenAI is going to announce. And then the following day, same time, 10 am Pacific, for Google IO to see what they are announcing.

02:44:49
We live in interesting times and we are very happy to be here to cover it for you your support, of course, much appreciated. Thank you to all the Club Twit members. If you're not a member yet, may I invite you personally to join. We'd love to have you Twittv slash Club Twit for more information After the fact. On-demand versions of the show available at Twittv On YouTube. Youtube, there's a video channel dedicated this week in tech and of course, you can subscribe. That's probably the best way to do it. That way you'll get it immediately, as soon as it's all, as soon as benito cleans it up and makes us sound smart, that's when we'll put it out. Thank you, benito, thank you, john jammer b, thanks to all of you for joining us. We'll see you next time, as we have said now in our 20th year. Wow, another twit is in the game. Bye-bye.