Tech News Weekly 339 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.

0:00:00 - Mikah Sargent
Coming up on Tech News Weekly. Will Oremus of the Washington Post stops by to talk about re-bunking, a term that means looking at the potential for election miss and disinformation and trying to get ahead of it by telling people about what could be out there and describing why it's inaccurate. Afterwards I have my first of two stories of the week. It's all about how Google has confirmed a huge trove of data about how its search optimization works and what we can glean from it. Afterwards. Scott Gilbertson of Wired stops by to talk about EOS, an open source operating system. That's all about beating the goo, so let's get Google out of our operating system and keep things clean. Finally, we round things out with OpenAI's reported deal with Apple and how it has Microsoft honestly a little bit worried. All of that coming up on Tech News Weekly.

0:01:13 - VO
Podcasts you love. From people you trust. This. Is TWiT.

0:01:18 - Mikah Sargent
This is Tech News Weekly with me, Mikah Sargent, episode 339, recorded Thursday, may 30th 2024, Tackling Disinformation With Prebunking. Hello and welcome to Tech News Weekly, the show where every week, we talk to and about the people making and breaking the tech news. I am your host, Mikah Sargent, and given that it is a rare fifth Thursday of the month, that means I'm flying solo today, folks, and joining us for the first interview of the day is Will Oremus of the Washington Post, here to talk about prebunking, an interesting word that I can't wait to get into. Welcome back to the show, will. Yeah, thanks for having me. Yeah, so let's start at the top here. Can you explain the concept of prebunking and how it kind of differs from traditional methods of combating election? Well, just misinformation in general, but in particular in this case, election misinformation.

0:02:18 - Will Oremus
Yeah.

So prebunking is obviously a play on debunking, which is when somebody says something that's not true and you come in and show that it's not true and tell people what the truth really is.

The problem with debunking when it comes to misinformation in an election is that, with social media and the speed at which claims can spread virally online, by the time fact checkers get around to debunking something that's false, most of the people who were going to see it already saw it and to some extent the damage is done. I mean, some people might see the debunk and go, oh, that thing wasn't true after all. But I think most of us can probably count on one hand the number of times that we've ever gone back and seen something after the fact that changed our mind. So the idea is to somehow get ahead of misinformation by kind of doing the debunking ahead of time. This requires anticipating what the false claims are going to be and then preparing people and voters specifically in this case, to evaluate those claims and to sort of see through them when they first encounter them in the wild.

0:03:26 - Mikah Sargent
That I have to say. That aspect of it of guessing what people are going to come up with it has to be the most difficult part of prebunking in the first place, and it seems like in many cases folks are tackling the easier kind of more predictable aspects of this rather than you know. Oh, I saw candidate number one catch a rat and feed it to a cat that then he sold to China or something like that. There could be some wild things that people put together and I can't imagine trying to come up with what goes on there. But I was hoping you could provide some specific examples of prebunking campaigns that have actually been implemented globally. I know Google has worked on some in the European Union.

0:04:11 - Will Oremus
Yeah, that's right. So I mean, and especially with generative AI, just to pick up on something you said you can make it look like anything happened now, right, like I don't know something about a rat and a cat, but you could just put that into you know mid journey or whatever, and it would spit it out and you could circulate that on social media. But yeah, in fact, a lot of the claims are somewhat predictable. This is what our sources say. I mean, it's not hard to guess that there will be claims of mail-in voter fraud. It's not hard to guess that, whenever elections come around, there'll be claims that your party's ballots are getting thrown in the garbage and we have footage of it or that you know that the voting machines are rigged or that sort of thing. So some of them really are predictable and those are the ones that some of the state and local election officials are trying to get ahead of.

One way that some states are doing this is they're actually like hosting, like holding videos where they show their election workers like meet the people who count the ballots, see how we do it. We'll show you like this these are real people, these are your neighbors counting the ballots. This is not, you know, these are not like, uh, the Illuminati, like you know, in a closet, you know, cooking up the election results, um, so uh. And then it's also been yeah, you mentioned Google, so Google's Jigsaw division, which is like sort of this. It's sort of this like ideas skunk works thing where they try to come up with, just you know, sort of tech related ideas to improve the world in various ways. I'll say they've had sort of mixed success with that in the past, but they've gotten interested in this prebunking idea. And in the European Union they're putting out videos that are really about they're not so much about anticipating a specific false claim, they're more about sort of educating people how to spot fishy claims in general.

So like there's this cartoon video about scapegoating and about how if you see somebody trying to blame a group of people, particularly if it's like a marginalized group for some big societal problem, they're probably trying to sell you a line. You know that's not, that's not, that's something that you should be suspicious of. They, they have it with. You know a few other just common rhetorical techniques, and so the idea is that if you see the, if you see the rhetorical technique, then no matter what the specific claim is you'll recognize. Oh, these people are. You know this is an example of scapegoating. I should be, I should really, you know, I should check this out before believing that you know the Roma people were being brought in caravans to all vote for the candidate I don't like, or whatever.

0:06:37 - Mikah Sargent
Right, absolutely, and I found it interesting in your piece you talk about how Google, while doing these kind of little videos, talking about these different techniques in the European Union, the company has, at least as it stands, decided not to launch a prebunking campaign here in the United States. Why that's the case and to find out the answer later in your piece was really interesting, and it wasn't something that I had thought of off the you know off the top of my head. But then, also because Google is not doing that, you did kind of talk about some alternative approaches that are being taken. Maybe you could talk a little bit about how local celebrities, so to speak, are also being asked to get involved.

0:07:29 - Will Oremus
Yeah, absolutely. And also I should take a moment to say that this is a story that I co-authored. My colleague, Kat Zekreski, was the lead author on the piece. She's our tech policy reporter and she's fantastic. So Google has not done it in the US. So Google has not done in the US, and they're not saying exactly why, but you can kind of guess it, Right.

I mean, there's been a big concern from the right, particularly Republican leaders in Congress, that this whole idea of content moderation, of debunking misinformation, is really this sort of liberal conspiracy to suppress conservatives or to keep Donald Trump out of office, and they've been very suspicious of it. And anytime a tech company gets involved these days in a controversial content moderation decision, they're likely to get hauled in front of Jim Jordan's Weaponization of Government Committee in the House and grilled by Republicans, especially about why did you do this and are you? Is this a conspiracy, you know? Are you censoring the American voters? So in the United States, the tech companies have become quite shy about doing this kind of thing and I should say I mean, if you're a tech giant, I mean, you know.

I think that there's probably some frustration on their part, but also they don't love being in this business in the first place. They didn't start Facebook or Google in order to be the arbiters of truth. They wanted to make a platform that a ton of people use and make a ton of money at it. They don't like being in the political football, so I think they also were willing you know, I don't know about happy but like they were willing to take a step back and say fine, look, you know, we're going to do less of this here in the United States because it just gets us in trouble. But, yes, the state officials have taken up the mantle. There was an example in Arizona's Maricopa County where election officials were bringing in members of the Phoenix Suns basketball team to help explain how voting works and how the polling stations work. And so, yeah, the idea of getting ahead of election misinformation is mostly in the US, happening at the state and local level, even as it happens at the national level elsewhere in the world.

0:09:38 - Mikah Sargent
Understood. Now, you mentioned this briefly. We kind of both talked about it a little bit, but it's interesting that rebunking is at the forefront right now, given what we've seen in terms of trying to combat misinformation and disinformation. For the past you know 10 years. For the past you know 10 years. What do you think that generative AI has played a role in the sort of arguable acceleration in trying out new things? Or was this destined to happen, given the proven election interference we've had in the past? And how trying to correct afterward election interference we've had in the past and how trying to correct afterward simply showed that it did not work.

0:10:29 - Will Oremus
Yeah, that's a very fair question. A lot of these efforts the prebunking idea predates the current wave of generative AI. But it is true that there are a lot of concerns about how generative AI could influence US elections, how it's already potentially influencing elections abroad. My colleague, pranshu Verma, and I wrote a story last year about how voice fakes where you make it sound like a politician said something they didn't use. You know generative AI audio In the Moldovan election.

They were made to make it seem like a candidate had been conspiring with a reporter to cover up the truth about stealing the election and we don't know how many people that convinced, but it certainly got shared quite a bit.

So that is a real concern.

But when I talk to the people who study this for a living, they say that it's not, that it's a distraction, but you sort of can't take your eye off the ball that social media is still the way that people find out this stuff. Right, social media is still where the claims spread, whether they're created with generative AI or whether that's just somebody making up something false and writing a text post or Photoshop or whatever. The place where the action is is on social media. It's in whether social media moderates it, whether they apply fact-checking labels, how widely their algorithms distribute it and, by the way, I think that's a really underrated factor here. I mean, you hear a lot of controversy about social platforms deciding what posts to take up, to leave up, what posts to take down, but part of the root of the issue is really that these engagement-based social networks were designed to make sort of the most sensational claims go viral, right, and so a lot of the most sensational claims turn out not to be the most true or the most balanced claims, and so I think that's sort of an underrated part of what the underlying issue is here.

0:12:21 - Mikah Sargent
In the piece you all link to a game that maybe some folks have played that shows kind of misinformation, disinformation, gamified for the purpose of getting clicks and kind of gaining sort of overall celebrity and the role that that plays.

And it was a really interesting kind of click through game that I played after I read your piece.

And you know, this is the thing when it comes to those of us who write about this stuff regularly, talk about this stuff regularly, pay a lot of attention to this stuff regularly.

I think many of us do have a built in skepticism when it comes to all of this. I know, you know, for me it's very rare that I ever read something and say, oh, I believe that wholeheartedly and it gets into my brain and then I pass it along as fact to other people because I just kind of automatically assume everything is a lie and don't prove it otherwise. But that's kind of what this prebunking is about, in the sense that it's hopefully going to introduce that curiosity and skepticism and, more importantly, a filter that exists between the stuff that's out there. But it's very clever the way that people use emotion to sway people's opinions or to bypass any filter that they have, and so I was kind of curious what evidence is there as to the effectiveness of prebunking? And then are there critics who say I don't think prebunking is going to do much of anything at all.

0:14:07 - Will Oremus
Yeah, yeah, there's a lot there. Yeah, that game is kind of fun. I would. I would urge people to try it out if they want. It's called bad news and it actually puts you in the role of the propagandist, the person who's trying to spread misinformation. And you have to, like you know, get. You have to hire bots to spread your falsehoods on social media and you can choose sort of which false claim to push on a given day and how to tap into people's emotions or biases or bigotry.

I do think most people are pretty good at being skeptical of information that goes against their beliefs, right? When we see something that we don't, that we're disinclined to agree with, we tend to treat that very skeptically. When we see something that we're disinclined to agree with, we tend to treat that very skeptically. When we see stuff that reinforces what we already believe, we tend to be very credulous about that, and you'll even hear people who share what's later deemed to be misinformation. They don't actually really care that much if it's true. They're like, I just thought it was funny, right? I just thought, okay, maybe Biden didn't do this dumb thing, but we know he's an idiot, so, like, who cares if this particular thing wasn't true, and so I think understanding the motivations for why people share misinformation is is valuable.

And yes, there are some critics of this idea of pre bunking.

It's not that they say that we shouldn't be trying to educate people about how to spot false claims, but the criticism is more that, like, look, you're trying to teach digital literacy and this is like a whole field of study, right, this is like years of schooling to really become an expert at sifting through truth and falsehood in communications, in marketing, online, on social media, and you're trying to do this with, like a cute 30-second cartoon video.

I mean, yes, educating people to be more literate about what they see online is really valuable, but the skepticism is like, are these little prebunking campaigns actually going to have any effect? And some of the critics in the field say, all right, debunking may have its shortcomings, but at least we know, at least there is solid evidence that when people see a debunking may have its shortcomings, but at least we know, at least there is solid evidence that when people see a debunking, it moves the needle a little bit, like there's evidence for that, and we don't yet have much empirical evidence that these types of sort of short, quick prebunking campaigns have a significant impact on voter behavior, but maybe we'll find out this cycle.

0:16:22 - Mikah Sargent
Let's hope, because honestly it gets a little disheartening at times to hear the even though in that in that case you know folks are not saying don't do this, but they're saying good luck, this takes a. It's like there's got to be something we can do. You know, at least something's happening, at least somebody's trying something, and hopefully we can continue to try something and see if anything possibly works, because by golly it feels a little bleak out there at times. I want to thank you and your co-authors so much for taking the time to put together this piece. Everybody should head over to the Washington Post at WashingtonPostcom to check it out. Of course folks can head there to keep up with what you're doing, but is there anywhere else they should go to follow along with what you've got going on?

0:17:09 - Will Oremus
They can subscribe to our Tech Brief newsletter. It used to be called the Tech 202. We just changed the name of the Tech Brief. It's three days a week and it covers it's analysis of stuff in the news in tech that has implications for policy and society. I co-write that with Cristiano Lima, who's also on our team, and you don't have to be a Washington Post subscriber to sign up for the tech brief. So if you're interested, go for it.

0:17:38 - Mikah Sargent
Awesome. Thank you so much for your time and hopefully we'll see you again soon. Thanks for your great questions, all righty folks. It is break time here and hopefully we'll see you again soon. Thanks for your great questions, all righty folks. It is break time here on Tech News Weekly and I am excited to tell you about Zscaler. We're bringing you this episode of Tech News Weekly.

Zscaler is the leader in cloud security. Honestly, it's no surprise that cyber attackers are using you guessed it AI in creative ways to compromise users and breach organizations from high precision phishing emails to video and voice. Deep fakes of CEOs and celebrities we were just talking about those deep fakes of celebrities and other leaders. In a world where employees are working everywhere, apps are everywhere and data is everywhere, firewalls and VPNs aren't going to cut it. They are failing to protect organizations. They're simply not designed for these distributed environments and, more importantly, arguably the AI-powered attacks. In fact, firewalls and VPNs have become the attack surface in some cases. In a security landscape where you must fight AI with AI, the best AI protection comes from having the best data. Zscaler has extended its zero-trust architecture with powerful AI engines that are trained and tuned by 500 trillion daily signals. Trillion daily signals yeah, daily signals 500 trillion. Zscaler Zero Trust Plus AI helps defeat AI attacks today by enabling you to automatically detect and block advanced threats, discover and classify sensitive data everywhere, generate user-to-app segmentation to limit lateral threat movement, quantify risk, prioritize remediation and generate board-ready reports. You can learn more about Zscaler Zero Trust plus AI to prevent ransomware and other AI attacks while gaining the agility of the cloud. Experience your world secured. Visit zscaler.com/zerotrustai. That's Zscaler Z-S-C-A-L-E-R.com/zerotrustai. TZscaler Z-S-C-A-L-E-R.com/zerotrustai. Thank you so much to Zscaler for sponsoring this week's episode of tech news weekly.

All right back from the break and I want to tell you about my first story of the week. This is all about Google and something that's actually near and dear to my heart as a person who has worked in journalism for quite a while. I have talked before about how I worked at a company called Newsy and at that company I was kind of a standard journalist, meaning I wasn't sort of specific to consumer technology at the time. And working for a startup means many people are going to be familiar with this, means wearing a lot of hats. Everybody kind of needs to have basic knowledge of as many different skills as possible, and it kind of seems like the more skills you have, the more knowledge you have, the more valuable you are to the startup. And in the case of that company, at the time before it was acquired, many of us did have many different skills and I can remember part of that being conversations surrounding how we could better reach our audience, how we could make our content more interesting to our audience, and much time as possible researching how you can better appear in search rankings. And in most cases that means how you best to appear in Google's search rankings rankings. And in many cases that has arguably meant that a lot of sites start to do very similar things in order to appear higher in the ranks.

And I don't know about you, but I have certainly in the past done a search for a question right, like okay, my computer, my computer keeps popping up this prompt that says this error, that says you know, error 9040 cannot access root file, let's say, and you type in error 9040 cannot access root file and you see two or three sites that claim to have the answer to it. But then you go to that site and written in the blog post over and over again is error 9040 cannot access root site. And you've got to scroll and scroll and scroll and you're scrolling past other errors and other write outs of what the different errors are. Because these sites are simply there to try to get you to visit those sites so that the advertisements on those sites are seen by you, or at least registered to have been seen by you, so that they then make money off of using that Google search result to get you to go there. And so there's a lot that's involved here, right, how to work your content in such a way that it appears where it should in Google's search rankings.

And there's a lot of stuff that third parties have had to just figure out on their own. And in many cases Google has kind of come out and said oh no, that's not true, or that's not the case, or oh, it's not part of our guidelines, et cetera, et cetera, et cetera. Well, finally, google has kind of had to answer a little bit for some of those guidelines and, more importantly, the debunking so there's that word again of some of its guidelines, because a source has released about 2,500 documents that claim to be internal documents for how Google search works. This is a website called SparkToro, and SparkToro's Rand Fishkin is the one who received an email, and this is, it says received an email from a person claiming to have access to a massive leak of API documentation from inside Google's search division. Now, whether or not this is the case, whether or not this is true that it belonged to Google, was up for speculation until yesterday as we record this. So we record this on Thursday May 30th. This was on Wednesday May 29th. Google did confirm that the leaked search documents are real.

Now, in these internal documents, there are a number of things that appear to give more insight into how Google ranks stuff that is on the web. Google has for a long time said that the way that people browse in Google Chrome which, of course, is Google's own browser has no impact on how Google search results are shown. However, in the documents and, john, this is on the Google won't comment on a potentially massive leak page. So I think that's the second link on D19, if you want to show it. If you scroll down a little bit on that page, there is a thumbnail from Vogue there. So for folks who are listening to this and not watching, there's a screenshot from a Google search result for Vogue, and in it there's the main website, but below it, many of you might be familiar with the kind of sub links that Google will sometimes show for websites. In this case, it shows Met Gala 2024, fashion Magazine, fashion Shows and Beauty, and, according to the section in the documents that mentions Chrome, these links that are appearing underneath the Vogue website in the search results are allegedly created based on what people visit on the Vogue site when they're using Chrome Chrome.

So Google has said for a while that Chrome data plays no role in whether or not how your search results are ranked, but this seems to show differently. And again, google did confirm that the search documents are real, whereas in the before times it had nothing to say about it. So it at least suggests that Chrome data is part of the results ranking. Now there are some other things, like what's called EEAT, that's, experience, expertise, authoritativeness and trustworthiness, and this is a metric that many an SEO marketer is familiar with because it is used by Google to quote evaluate the quality of results. Now Google says that, even though it evaluates the quality of the results, it does not have a role in the ranking of the search results, and that appears to be true. But something else that is a little bit different is that for the longest time, google has suggested that whether or not a website result is an article or a scientific research paper or anything like that, that that doesn't factor into where Google ranks the piece.

The byline okay, the author byline is something that is shown as a point of importance in these documents, meaning that if a page, if a search result, has a byline, then it appears that that may rank it higher in the results than it would otherwise show. That may rank it higher in the results than it would otherwise show. So that kind of gives us some insight as well into how those, how having some byline on a page is going to make a difference. And so when you come across those pages where the name says, like you know, the, I can't even think of his name. Maybe it's Lucky, lucky the Leprechaun from the Lucky Charms box, like you know, by Lucky the Leprechaun. Whatever they can put in that byline, having a byline seems to be important. So it'll be interesting to see how this massive trove of internal data affects SEO stuff going forward and how we will continue to see how Google's just absolute ownership of the modern web, in particular, when it comes to searching for things on the modern web. Um shapes the entire internet and I I think that it lends even more legitimacy and credence to the regulatory bodies that are currently looking at Google from an antitrust perspective. There's a lot more to check out with that, and I encourage everybody to go read the two Verge articles that we will have linked in the show notes. But it is time, once again, to take a quick break before we come back with our second interview.

I do want to tell you about the awesome IntouchCX. We're bringing you this episode of Tech News Weekly. That CX stands for customer experience. IntouchCX knows that a major goal of your company is to provide your customers with an excellent. There it is customer experience. IntouchCX works with brands around the world to apply ding ding ding AI and smart automation to optimize the customer experience for every touchpoint across the entire customer journey, and this is what's great about IntouchCX. We had an opportunity to talk to them and I specifically brought up how I get annoyed when I go to a page and I'm trying to talk to someone and the bot on the page is very bad. That is something that IntouchCX is fixing. It's making that better.

A lot of people are just kind of getting the lowest quality, least expensive, least supported option for customer experience and it's bad for the like. They know it's bad, we know it's bad as customers. This is how you do better. You can provide a great customer experience because companies need to depend on a combination of talented people, efficient processes and innovative technology. But quality service only happens when you've got the right mix of those things working together, and most companies are dealing with outdated and complex systems, which makes it hard to meet customer needs and hard to retain your customer service agents. So customer interactions need to feel smooth on both sides, both for your customers and for your company's agents, and that is why IntouchCX improves the experience for both the agent and your customers at every part of the journey, and they do this through innovative AI and automation.

Intouchcx is an advisor that you can rely on and trust to elevate your customer and agent experience and to add AI and automation every step along the way. So talk to the people at IntouchCX. This was a great thing that we had a conversation about too. They said, look, people who follow up here are going to actually get to talk to us. We'll work together and figure out what is right for you, and I love that.

Whatever your company size and regardless of where your company is on its customer experience journey, its CX journey, IntouchCX can help. Go to intouchcx.com/twit and sign up for a free, no obligation consultation. It will create a tailored customer support strategy for AI and smart automation customized for your business. That's intouchcx.com/twit. Thank you so much to IntouchCX for sponsoring this week's episode of Tech News Weekly. All right, we are back from the break and I'm very excited about our next guest from the break and I'm very excited about our next guest, senior writer at Wired, who is here to talk about an alternative smartphone operating system, gasp Scott Gilbertson. Welcome to the show. Hi, thanks for having me. Yeah, it's great to get you here. So let's get started with kind of a quick high-level description of and I'm hopefully saying this right EOS. What is it? Who makes it?

0:33:57 - Scott Gilbertson
And you've used it, yes, yes, I have, I'm using it still, actually, even after the article. So, yeah, and I think that's how you say it, that's how they said it. So that's what I'm going with. So it's an open-source mobile operating system. It's got a particular focus on privacy, as opposed to some others that are more generic, um, and it is sort of overseen by what's called the e-foundation and that's a non-profit organization that that takes care of the code and community and all that stuff.

0:34:29 - Mikah Sargent
So got it Now in your piece. You kind of kick things off by talking about how Android itself is an open source operating system, so anyone can take that operating system and kind of build on it. Which had me wondering as it did you at least in the piece why is it that we don't see more Android-based operating systems sort of thriving in the wild?

0:34:54 - Scott Gilbertson
yeah, it's a great question. Um, I think there's sort of two answers. One.

The first one is is the not so good one, which is that the core of android that is open source and that's available is only part of what you need to build a mobile operating system. It's sort of the underlying functionality making sure you can send text, make calls, all that of stuff, connect to your radio on your phone but it doesn't really provide the user interface and it doesn't provide the apps and it doesn't provide the network services that we all kind of rely on and use, that are a huge part of our phone experience. And building all of that stuff is complex, it takes time, it takes people, it takes money and there just aren't a lot of projects out there doing it. But the second part I would say is that there are actually quite a few smaller projects. You know I don't know how many users these have, but probably not that many, but there's EOS, there's Graphene OS, divest OS, calix OS and there are a couple others. They're just very niche, very small kind of at this point anyway.

0:36:02 - Mikah Sargent
Understood Now, before we really dig into EOS, I was hoping you could tell us a little bit about your time with Lineage OS and kind of going further with this. You know why there aren't as many trials and tribulations that are necessary that you needed to do to have a fully functioning phone while you were using LineageOS.

0:36:24 - Scott Gilbertson
Yeah, I feel like maybe I threw Lineage under the bus a little too much in that article. I've actually had, you know, a great experience with it and it was the operating system I was using up until six months ago when the EOS people reached out and I tried that. But lineage my experience initially was not great because I didn't pick the right hardware, and that's really the key here is like you've got to make sure the phone you have is well supported by the operating system when you're installing it yourself. I bricked an Amazon device, for example, trying to put LineageOS on it, and, you know, once I moved, I quickly realized that OnePlus and Samsung and Google devices are kind of where you should stick. If you want to try any of these really not just lineage, but any of them unless they have their own hardware, which is becoming more common then I would recommend just going with a phone.

It already supports it out of the box.

0:37:29 - Mikah Sargent
Right, right, yeah, it kind of comes built with it. Now let us kind of get into EOS specifically. So you talk about how it is privacy at the forefront. That that is kind of the whole deal, that's the selling point of it. So tell us about what EOS offers in terms of privacy, what you can expect if you use this operating system.

0:37:52 - Scott Gilbertson
Yeah, so the first thing it does is get rid of all things Google. It probably bills itself as a de-Googled version of Android. So there's no Google apps installed by default, there's no Google services behind the scenes. They're all replaced by something called Micro G, which is it basically provides the APIs for apps to still function, but they aren't connecting to Google.

And then EOS really goes beyond that with its privacy tracking system, where it will let you block, on a per app basis, every single tracker that any app has, which was really eye opening for me, because I did not realize, you know and I'm a technology reporter, I do this for a living and I didn't realize exactly how many trackers are in the apps I use. I just looked at it before we started and in the last 30 days, my phone has blocked 30,000 attempts to phone home. Yeah, so it's. You know there's a lot happening that we don't realize is happening, and it kind of points that out to you, which I found extremely enlightening, in a bad way, I guess.

Yeah, Um, and then there's there's two other things that I really like. One is the ability to spoof your uh geolocation and IP address, though I will you kind of have to use those sparingly, because if you're spoofing your geolocation, you know your maps aren't going to work and none of that stuff is going to function. And then, within the app store, when you're thinking about downloading some new app, they actually have a rating one to 10 rating of how the privacy is, how many trackers it will have installed and kind of attempt to educate you ahead of time, which is nice.

0:39:35 - Mikah Sargent
Got it Now, of course, one of the main reasons we all have smartphones or that many, I should say many of us have smartphones instead of just a phone that's got some buttons on it for dialing is because it means we can have apps on our phone. Given that this is a de-Googled device, that is, you know, built on an open, an open source operating system, are there apps, and you know what was that experience like for you in getting apps on the device.

0:40:07 - Scott Gilbertson
Yeah, so there is. There are apps and there is an app store. It's interesting because it's kind of a blend of, I guess, three different app stores its own homegrown app store. It also can download anything from the Play Store, both through your account if it's something that you've paid for and you need to be signed in to get the app, that still works, or you can download things anonymously from the Play store, so you're not associated with it. And then it also connects to fdroid, which is an open source repository of android apps, and it kind of puts them all in one single interface, gives you a nice search tool and, like I said, it has the privacy rating at the bottom of every app. So, yeah, it's very well done.

The criticism I think that some people have, which is occasionally valid, is that maybe some versions of the apps are slightly out of date. The other issue is that and I did not run into this, but I have heard from European readers that some of the banking apps don't function. But I've heard from European readers that some of the banking apps don't function as soon as they're able to tell that they are not on a Google signed operating system. They just won't connect. So that may be an issue for some people. Like I said, I used several US banking apps and did not run into that, but I have heard from some wired readers who did.

0:41:35 - Mikah Sargent
We are getting some feedback in, not not audio feedback. Uh, some folks are mentioning in our live chat, uh, something that I had not originally planned to ask about, but since there's some interest, uh, can you tell us a little bit about the, the troubles of locking the bootloader, but how eos maybe does allow for locking the bootloader and for folks who, this guy, don't necessarily know what the heck it means to lock or unlock the bootloader, why would someone want to do that or not do that with EOS?

0:42:06 - Scott Gilbertson
So the bootloader is what loads the operating system and runs it on your phone, and you want it to stay locked because you only want to run the operating system you intended to run, not something an attacker might have placed there. There's also some persistent malware that is able to get around your attempts to delete it if your bootloader is unlocked. I don't know how common these are in the wild. I've been running an unlocked bootloader for five years and haven't had a problem. So it, I guess, depends a little bit on who you are and what you're doing, but that is necessary because you are loading an operating system that wasn't signed to the hardware. So relocking the bootloader can make it impossible to restart your phone.

Relocking the bootloader can make it impossible to restart your phone and some alternative OSs Graphene and a couple others that are downstream from Graphene are able to do that and that is probably more secure. In the end. Eos is able to do that on a select number of phones, but having already bricked a device once, I have not tried it, so there is a certain element of risk involved, especially um, when you're not using, like the Marina phone, which is probably their best supported hardware since they make it, but yeah it. There is a security risk there. Um, personally I'm I'm okay with it, but some people might not be.

0:43:33 - Mikah Sargent
The risk averse Got it. Okay, let's move along here with the next question I had, because, in theory, the reason why somebody would want to use this is to de-Google yes, and for a more private experience. You do mention in your piece that EOS has features that are like other smartphone operating systems that they sync data through their servers, they provide online backups, and so it kind of had me wondering is everything encrypted? Is it encrypted at rest, and do you have concerns about choosing to sync your data with the company at the root of this versus a big tech company like Google, because sometimes people will make the argument that, oh, they have more trust in the company that has to deal with these attacks regularly versus these small companies that maybe aren't maintaining security as much. I'm curious where you stand on that, particularly as someone who has made the choice to go with a de-Google experience.

0:44:36 - Scott Gilbertson
Yeah, I mean I should say I guess that I have not tested Marina Cloud, which is what their whole suite of offerings is, very extensively. I didn't need it as much as I want to live a Google-free life. You know, like Wired's, email runs through Gmail. I still have to have Gmail and some other apps on my phone. What I don't use is any of the sync and backup from Google, and for that I've actually relied on my own sync and backup rather than Marina Cloud.

But they do offer a suite of services, everything from email to, like you said, backups of your data and stuff like that. It's all open source. You can go look at it on GitLab. I could not find any data on whether it's been independently audited to see if how its security is. So that is moderately concerning. But I mean it's all hosted in on european servers, like the email, everything, um. So it's governed by the gdpr and all that kind of stuff and behind the scenes a lot of it is using next cloud, which is a pretty well respected open source project that has a good track record with security. So I mean that's again one of those things that people probably have to make their own choice on that If you're very risk adverse, you probably want to do your own syncing of some kind or backups and you can actually self-host Marina Cloud. But that's quite a project. I have not undertaken that.

0:46:11 - Mikah Sargent
And then my last question for you before we let you go. I'm curious if you think EOS comes close to being a viable alternative to the mostly locked down experiences you get from iOS and Android and, in other words, do you feel like this will always be a niche offering. Or is the Google free experience compelling enough and easy enough to use to draw in more people, or is it kind of just one of those? I wish more people would hop on board this, this, this Google free lifestyle?

0:46:44 - Scott Gilbertson
Yeah, that's a great question. I mean it's hard to answer. I kind of feel like it's going to be very niche if you have to install the operating system yourself, because you know you've got to at least figure out how to boot up the command line and connect and run a lot of software to do that. But if more companies, more phone manufacturers, start using these operating systems by default shipping with them, I think it becomes a much more viable option for mainstream users, especially, you know, because the hardware can be so tightly integrated with the software and there's nothing stopping say, samsung from building out their operating system on top of EOS rather than stock Android. So that might be a way in which some of these projects become a little more mainstream. But as long as you're talking about installing alternate operating systems, I do think that remains the province of, you know, nerds like us who want to tinker with stuff.

0:47:48 - Mikah Sargent
Absolutely Well. Scott Gilbertson, I want to thank you so much for taking some time to join us today. Of course, folks can head over to wiredcom to check out your work, and we'll, of course, havea link to your review so that they can read everything else you talked about. Um, if they do want to follow you online, where's a good place for them to go to do that?

0:48:08 - Scott Gilbertson
yeah, I don't use a lot of social media. I have my own website. It's luxographnet, l-u-x-a-g-r-a-fnet, and, uh, you can see what I'm up to there. Thank you very much awesome.

0:48:19 - Mikah Sargent
Thank you so much and I hope to see you again soon. Thanks, bye, all righty folks. Another story of the week coming up, but I do want to take a quick break to tell you about Club TWiT at twit.tv/clubtwit For $7 a month. You out there can join the club and we'd love to have you do so. When you join the club, you gain access to every single one of TWiT shows ad free. It's just the content, none of the ads. You'll gain access to the TWiT+ bonus feed that has extra content you won't find anywhere else Behind the scenes before the show, after, after the show, special club twit events get published there. It's a lot of fun. Um, we have you know. Anytime you join, you gain access to this huge back catalog of content, as well as access to the members only discord server, a fun place to go to chat with your fellow club members and also those of us here at twit.

I'm excited to say that every what was it second? Now I'm forgetting, is it second or third Wednesday every, uh, third Wednesday of the month starting in June, you can join me, uh, for Mikah's crafting corner, which is just a time to get together and we'll all be working on crafts that we have. So you bring your craft, I bring my, my craft. We just kind of chat, hang out and show what we're doing. Maybe you know somebody needs some knitting advice or some crocheting advice or drawing advice, whatever, so we can chat about that. It's going to be a cozy, kind of chill time to just work on crafts with some other people, and you can only be part of Micah's Crafting Corner, along with all the other great stuff, by joining the club. I should also mention that, along with all of that, you also get the video version of our Club TWiT exclusive shows, including iOS Today, hands on Mac, hands on Windows, home Theater Geeks and the Untitled Linux Show. Those are only available in video versions to those of you who are subscribers. So please, if you have not joined, consider joining to hang out, chat and have a good time with all of us. Okay, to round out the show, my second story of the week.

We have heard now from good old Mark Bloomberg, aka Mark Gurman, that Apple is, as is to be expected, working on its own generative AI offerings Including. Now we've heard rumored that the company might offer up custom emoji based on the conversations that you're having. So you can imagine I, as I always do, have conversations about chihuahuas and coffee, and so if I was talking to someone about chihuahuas and coffee, maybe a custom emoji of a chihuahua holding a cup of coffee would appear. But also more important and practical things like providing summaries of recent emails that you've received, providing summaries of that on and pop and group chat that you have. That is just going on and on and on and making sure that you're kind of up to date with the information that you are provided. Those are some of the things that Apple is reportedly working on on device. So in this future version of its various operating systems, your device itself will be able to run instances of Apple's AI technology to locally process that stuff, provide that offering and give you some information about whatever it happens to be, but not all of it is able to happen on device reportedly been in talks with both Google and OpenAI about partnering with those companies in order to provide a server-side experience, and you can imagine that, while things like summaries and little emoji and probably a few other things can happen locally, some stuff really requires a large language model in the cloud and Apple, as you know again, this is all reported reportedly using its own Apple Silicon in servers to do a lot of the offsite generation.

But it's also thinking about partnering with companies and, according to the information, apple has finalized, or all but finalized, a discussion with OpenAI to use its GPT technology in the next version of iOS, ipados etc. Of iOS, ipados et cetera. According to information, apple and OpenAI have been in discussions since the middle of 2023, so the middle of last year and the developers at Apple have worked on kind of pairing OpenAI's APIs to Siri S-I-R-I, the Apple virtual assistant, showing that it's possible for Siri to better understand what somebody wants. By the way, in the Discord chat there is now a latte or a cappuccino, probably with a chihuahua standing on top of the cappuccino. Somehow it is very light chihuahua, anyway, distracted by that chihuahua coffee.

These are the things that are possible with generative AI, but Apple employees have worked to connect Siri again to chat GPT, to open AI's GPT technology, and have been able to do things like have it do more complex queries so it could better kind of understand what is being asked of it and, more importantly, kind of take what's being said, break it down into its component parts and do all of the things. Because right now, if I want to turn off lights in a room or I should say, turn on lights in a room and change their color and change their brightness I can either make a scene not a scene as in, like I'm getting upset, but make a scene in the home app, or I can Siri do this, siri change the color, siri change the brightness. And that's not what I want to do. I want to be able to say the name once and then say turn on the lights, change the color to orange and set the brightness to 75.

This is something that could be possible if it's able to be, you know, combined. But the big problem with this and it's able to be combined but the big problem with this and it's something that we talked about on MacBreak Weekly and something that I think is going to be fascinating is we have seen Microsoft, google, openai, anthropic to a certain extent, although I would say far less all of the big companies that are working in generative AI, both in text form, but also photo form and, in some cases, video form, run into some issues. Google right now is being memed to death because of it using Reddit as a source of information for its search results and many of those search results that you're seeing out there are faked. But the fact that Reddit is being used as a kind of a source of information means that you know you may get some odd results when you're just trying to look something up. So, whether it's inaccuracy, or it's issues with photos that are being generated that are historically inaccurate, or photos being used or what we talked about earlier disinformation and campaigns, apple is not going to want to be involved with any of that. Apple is not going to want to have anything to do to. It's not going to have anything to do with that, is not going to want to have anything to do with that. It's not going to want to be responsible for any of that. So, trying to figure out how to connect Siri and OpenAI's chat GPT together and not have it appear as if it's Apple that's doing this, I am very curious about. Is it going to result in every time I do one of these queries it's going to say brought to you by OpenAI or brought to you by ChatGPT, or it's going to have a little badge of ChatGPT in the corner? I can't wait to see what Apple does at WWDC when it comes to that, because I don't think that the company is going to at WWDC when it comes to that, because I don't think that the company is going to feel comfortable having any responsibility over what is generated by a chat bot that has been shown to have a bunch of misinformation or, in some cases, disinformation misinformation or, in some cases, disinformation.

Secondarily, when it comes to this deal, we know that OpenAI and Microsoft have a very deep partnership and, in fact, microsoft did a huge investment in OpenAI of $13 billion B billion. And given that Microsoft has spent so much money on OpenAI and has introduced and included OpenAI's technologies in its PCs and has gone so far as to launch a whole new line of computers called Copilot Plus PCs all of this stuff, that Microsoft has gone so far as to launch a whole new line of computers called Copilot plus PCs All of this stuff that Microsoft has gone all in, for them to suddenly hear or I should say for it to suddenly hear that OpenAI is also partnering with Apple, who is a competitor in many cases. That's a little concerning. And, reportedly, satya Nadella is a little bit CEO of I should say CEO of Microsoft, satya Nadella is a little bit CEO of I should say CEO of Microsoft, satya Nadella, is a little bit concerned about that partnership. So I don't think that that's going to have any bearing in the short term, given that WWDC is just around the corner and we're hearing reports that there's been a locked in kind of agreement. But perhaps it will have bearing in the long term and we could see Apple continuing to reach be doing because that's what Apple has always touted as the thing that sets it apart paired with how Apple will try to sell us on the fact that it's maintaining privacy when it has to do stuff that happens in the cloud, and the bearing of Microsoft's partnership with OpenAI on Apple's reported upcoming partnership with OpenAI. All of this is really fascinating and I am keeping an eye on all of it and of course we'll, when the opportunity provides itself, have folks on the show to talk about every aspect of this and more. So keep your eyes out for that and, of course, go subscribe to the information. Who has some great reporting folks. That is going to bring us to the end of this episode of tech news.

Weekly. The show publishes every Thursday at twit.tv/tnw. That's where you go to subscribe to the show in an audio and video formats. I mentioned Club TWiT at the top of the show, so I'll just say briefly or not at the top of the show, but just a little bit ago so I'll mention briefly twit.tv/clubtwit $7 a month. Consider joining.

We'd love to have you and, of course, help support this stuff that we do, help us keep doing what we do. And if you'd like to follow me online, I'm at Micah Sargent on many a social media network where you can add to chihuahua.coffee that's C-H-I-H-U-A-H-U-A.coffee. That's where you can go to get links to this stuff that I do online. Check out Hands on Mac and iOS Today, which will publish later today, and you can watch Ask the Tech Guys, which I typically co-host with Leo Laporte on Sundays. I will be out this coming Sunday, though, so it will be Leo solo, but you should still check out the show. We take your tech questions live on air and do our best to answer them. Thank you so much for tuning in for another episode of Tech News Weekly, and there will be a special guest host joining you next week. Bye-bye.