TNW 432: How Wrong Are Google's AI Overviews? - Is Google's AI Overview Reliable?

And this week I'll actually be kicking things off. Uh Google's AI overviews, many of you are probably familiar with them at this point, the AI-generated answers that sit at the top of the company's search results look authoritative, they feel definitive, but a new investigation from the New York Times uh takes a hard look at just how reliable those answers actually are. Working with an AI startup called

Maybe it's called OMI. It's O-U-M-I-O-MI. Uh the Times put Google's AI overviews through a widely used benchmark test and found that while the answers are right, about ninety-one percent of the time. That remaining margin of error, when applied to Google's more than 5 trillion annual searches, actually translates to tens of millions of wrong answers.

Every single hour. And frankly, it gets thornier from there. Even when the answers are correct, the sources backing them up often don't actually support what's being said. the big thing here, right? So let's look at kind of the numbers. When we look at an accuracy problem, uh, depending on who you're talking to and what their intentions are, you may hear the term rounding error used. And I could see uh Google trying to do that in this case, but these are big numbers.

Uh Umi tested Google's AI overviews across four thousand three hundred twenty-six searches and did so in two rounds, once in October when the system ran on Gemini two, and then again in February after it was upgraded to Gemini three. Accuracy did improve from 85% to 91%. But again, when you have that many people doing Google searches, 9% is an enormous.

But it can't escape this fundamental limitation that today's AI systems use mathematical probabilities to guess the best response. They do not use a strict set of rules defined by human engineers.

So I want to ask you, Amanda, first and foremost, um, just your general take on Google's AI overview, um, whether you make use of it when you're doing a search, and also I'd love to hear sort of people around you, right? We that that's one of the things that fat that's fascinating to me is like what are people saying about their use or lack thereof of of Google's AI overviews.

And I was trying to Google if there was a way to like do it better. And I Googled like sucking up water, picopeia, because I just like was typing fast and spelled Picopia very wrong. And then Google Overview was like, Picopeia is uh the name of the sensation in which you suck up water and it gets lodged in your throat or like.

If I'm reporting anything and any sort of information comes to me through like a Google AI overview, I'm like, okay, now let me go to the actual source and find what the source is and make sure the source isn't like some random website I've never heard of that Google decided is trustworthy when it's like some guy's blog.

in some ways, but it's more about like what is the website that I need to go to to find whatever it is I'm looking for. I'm doing searches. Um and so it's a little bit different how some people will use it as sort of a question and answer box. Um, so I'm already kind of a person who scrolls past it. But something that ended up being uh I think one of the most heinous uh sort of issues here. They decided to look at

um how Google's AI overview was displaying information. And here's the thing, when you think about it, right, if if the AI has an answer within it. sort of global context, the the information that just exists in its own mind, so to speak. So uh I don't know, the four plus two equaling six, right? It has this information um because of all of the training that it's done, it knows four plus two equals six. And so you could search that.

It would give you the answer. And then I think this is maybe that maybe is a little bit too easy. Uh let's go with something like um Male seahorses carry the children, carry their children in their little seahorse pouch, right? And

And that way that means that it can display this answer in Google's AI overview. But what it doesn't necessarily mean is that when it links to a source. That source actually says that thing because it is trained to provide sources.

But it isn't necessarily trained to provide sources that are accurately representative of the answer. So What that means is it can get the answer right because it knows it, but when it's linking to stuff, it doesn't necessarily mean that those links are actually relevant to the answer that it gave. And that's an issue because people are more likely to believe that it is giving the right answer because it has cited its sources.

Uh so they're calling this ungrounded as as uh the sort of answers, responses that might be technically correct but linked to sources that don't fully support them. In October, 37% of the correct answers. were still ungrounded. But here's what's wild. Even though it got better at answering questions, the questions were more accurate, or the answers, excuse me, were more accurate. After that upgrade to Gemini 3, 56%, more than half of the answers were ungrounded. They pointed to sources.

But those sources did not actually have the answer when you went to it. And so this is a problem. It's not. necessarily going to be the right answer. And even if it is the right answer, then you can't confirm that it is. And so people are just trusting this system.

And that's what I I think that's what's sort of upsetting to me is I don't understand putting a tool that may get things so very wrong

And that sounds like really small, but then you remember that at the time and probably more now, uh, but as of October, OpenAI or ChatGPT had 800 million weekly users. And zero point one five percent of that is more than a million. So that's millions of people that are talking to uh ChatGPT about things that perhaps they should be talking to with a professional or a friend or someone who is not a computer.

And it's just this the scale at which these AI tools are being deployed is just so vast that even if it's inaccurate like point one percent of the time or it hallucinates like such a tiny fraction of a percent. Like the numbers are even bigger with Google than with Chat GPT, because people are Googling more things a day than they are having conversations with ChatGPT.

And the numbers are just so large that like any tiny percent of error is so magnified and impacts so many millions of people in a way that is alarming.

Uh, the article documents several cases where AI overviews got part of the answer right, but fumbled the supporting details. When asked how old the relief pitcher Dick Drago was when he died, the system did get the age correct, 78. but then repeatedly misstated the date of his death in the additional context. In another case, Google was asked about Yo Yo Ma's induction into the classical music hall of fame.

It got the uh link to the organization's website correct. It listed a hundred and sixty five inductees, including Ma, but the AI generated answer said there was no record of his induction. It'll find the right source, but misinterpret the source. Uh it will uh so asked about the river bordering the west side of Goldsboro, North Carolina, and it identified the a different river, which is southwest of the city.

The actual answer though, the little river, which feeds into the bigger river. And it's said that the river ran through the city, but Google system incorrectly inferred that it ran along the western border. So those details are what's wrong about it. So I'm Yeah, right. I'm confused.

And he has not been inducted into the Hall of Fame, which I know he is still an active player. And Google did tell me that he has not been inducted into the Hall of Fame, but then it just gave me a list of other athletes named Bryce.

So there's uh Bryce Drew was inducted into the Indiana Basketball Hall of Fame in twenty twenty-four. Uh Bryce Cotton inducted into the Providence College Hall of Fame. Uh So yeah, that's definitely what people that are Googling Bryce Harper are looking to know is other athletes named Bryce.

It should be like that. And it should say, um, here's our AI overview. It might be wrong. So really is this worth your time at all? Is really what it should say.

Bringing you this episode of Tech News Weekly. Your marketing website sets the tone for your brand and is the one touch point every single one of your customers has. If you still struggle to make small changes and simple updates, well, You're leaving opportunity on the table. That's why so many companies, from early stage startups to Fortune 500s, are turning to Framer, the website builder that works like your team's favorite design tool.

Framer turns your dot com from a formality into a tool for growth, helping businesses build better websites faster with real-time collaboration. A robust CMS with everything you need to get great SEO and advanced analytics that include integrated A-B testing. Your designers and marketers are empowered to build and maximize your dot com from day one. I uh was given access to Framer to create a site.

And I have to tell you, one of the first things that I noticed about Framer was just how much control you're given, uh, just how much you can actually access. I have another tool that I have used to make websites. And with that tool, I don't I didn't have much control over everything. Uh the fact that I can like really dig in and get at analytics.

that it's thinking about how I was reading some of the support documents and it was very aware of how you have to set up special files on your site whenever you're doing things like Apple Pay or uh you're trying to get Google AdSense running. That's what I really liked about this. You have as much control as you need. Changes to your frame or site go live in the web in seconds with one click and without help from engineering.

So that your team could actually reduce dependencies and reach escape velocity. Framer is an enterprise solution, it truly is with premium hosting. Enterprise grade security, ninety-nine point ninety nine percent uptime SLAs that give the world's leading brands like Perplexity, Miro, and MixPanel the confidence they need to build their websites and framer.

Whether you want to launch a new site, test a few landing pages, or migrate your full dot com, Framer has programs for startups, for scale-ups, and for large enterprises to make going from idea to live site as easy and fast as possible. So learn how you can get more out of your dot com from a framer specialist or get started building for free today at Framer.com slash TNW for 30% off a Framer Pro annual plan. That's Framer.com slash TNW for 30% off.

Framer.com slash TNW rules and restrictions may apply. Thank you, Framer, for sponsoring this week's episode of Tech News Weekly. All right.

Uh moving right along, as I mentioned, joined this week by the wonderful Amanda Silverling. And now we're talking about, well, journalists. And AI and where things are going with the blurriness. Uh the the the line that's being blurred there. Uh Amanda, take it away.

possibly the identity of Satoshi Nakamoto, who is the uh pseudonymous founder, creator of Bitcoin. And this has been sort of like a mystery in the tech world for decades. And I don't so much want to talk about the story itself. Uh the TLDR is um He thinks that it is this guy, Adam Back, who is like a longtime cryptographer who was like working on digital payments since like the 90s and just

The kind of guy that would know the information you would need to make Bitcoin and who has been involved in these phases. And then the guy said, no, it's not me, which that doesn't mean it's not him, but he denied the report.

Uh anyway, but the reason that I think this is interesting to talk about is the way that they are using AI in this report, which is that there are these like very, very large uh just massive quantities of emails, like emails from various uh cypherpunk crypto uh email lists for decades. And Satoshi, even though he's like a mysterious figure, he still would like comment on these and he was someone that people like talked to. They just didn't know who he was.

And so what he did was he went through those emails and he found certain turns of phrase or misspellings or certain things that came up in Satoshi's writing and then he took the rest of the emails and put it into a large language model to try to figure out like Can the large language model identify what are like someone else that is having like common spelling errors to Satoshi or like similar turns of phrase or um just sort of like what looks like it was written by the same person.

And I think that's just like a really interesting way of using AI and journalism in a way that like On one hand, you can't actually prove anything with that, like you can make an educated guess. Um, but I feel like especially in the last few weeks, there's been a lot of discussion in journalism about how AI should or should not be used in reporting. And in one case, um, Max Zef at Wired wrote a report where he interviewed a bunch of journalists that are using AI.

like prominent tech journalists like Alex Heath, who is independent now. And in the case of these independent journalists, a lot of the time they're like, Yeah, I'm running a one person show and I need to be able to Expedite my work. So if I put my notes into Quad and it writes a draft for me and then I edit the draft, then like what's wrong with that? And um of course a lot of people think there's something wrong with that.

And there's like an interesting divide happening right now in tech journalism where you have a couple of these people that are like, If you're not using AI to draft your stories, then you're falling behind. And then um The Wall Street Journal wrote a story about an editor, I believe, at Fortune who was using AI to like significantly increase his output and just like

Cause like we we get so much news that we can't even cover. And if I tried to cover every single thing that got emailed to me every day, I would have to write like twenty articles a day, if not way more. Um, and that's just not feasible. But technically, if I took the press release and put it into some LLM that is like Fine-tuned with my writing. Could I technically output something that's like vaguely what I would write about it? Yes, but also.

Doesn't that defeat the whole purpose of writing? Um, so these these are basically just the debates that people are talking about in journalism and AI and how we should or shouldn't use it right now. So yeah, I guess on one end of the spectrum you have like it's really interesting what John Carreyroo did of a human would never be able to go through like thirty thousand emails and find like really nitpicky similarities in text.

versus like is there a difference between doing that and like putting uh some press release into Micah GPT.

that perhaps schools don't do a good job of focusing on and perhaps um uh uh commentators don't do a good job at times of remembering is where the dynamics of a work environment Um, the requirements of being alive and everything else that goes into that comes into play. And when you are not the editor in chief of a news publication or someone working in the business side of things and you are simply one of many writers working for a site and you are being

uh strongly encouraged to use these tools. And when you're not using these tools, then you become like there's so much that's involved in this that I think um Sometimes we have to make a little bit of space for or just to have that understanding in that context of Those are the questions that I wanna know. Were you sort of pressured into using these tools and Do you feel like

You know, the difference between you getting a raise versus the other person is that they're faster because they've used uh AI to draft things. And to be clear, I don't think that that is the way that it should be. But I think reality versus idealism when it comes to that are two different things, right? And that's something that I just Am mindful of. What I hate, though, is how often we hear after the fact.

that AI was used in the creation of this or that, and that it is not always explicitly laid out. Um, and made clear. That frustrates me. Um I think that we need to be upfront and honest about that. And in doing so, then you are sort of cleared to to proceed, right? You're using this tool. You're making it clear that you're using this tool. Now I have the choice whether I support this, you know, journalism by reading it or not.

The idea that any of this has to be discovered afterward, um, or that there are stu there need to be studies being done that reveal that AI was used in the making of this or what have you, that Troublesome. I don't think that it should be that way.

the next most probable word. And that's not always the most exciting writing that you can produce when you are a professional writer. But then at the same time, there's times where it's like, If I wanna know how much money did Meta spend on reality labs in 2023, if I Google that, I'm not gonna get an answer. If I ask ChatGPT that,

I will probably get an answer that I can then go back and fact check and make sure that it's accurate. Um I think also sometimes I think about it like Maybe this is not a right way to think about it, but my hypothesis right now is that maybe Chat GPT, like and like the main LLMs in general. Are sort of comparable to Wikipedia when you're in high school, where everybody tells you that Wikipedia is not a reliable source and you're like, Yeah, like I I know that anyone can go in and edit it.

But also if I go into Wikipedia and the answer is right there and then I can see, oh, they're attributing it to uh this article in a trusted journal and then I can click on that and go to the trusted journal. If it is expediting the time you spend looking for the piece of information, I don't know if that is the worst.

thing for writers that are on tight deadlines, but I also am like I think it is like a appropriate to use a slippery slope argument here where it's like, all right, well when you use AI for one thing, like what are you gonna use for it next? Or

He compares AI to a sports car that you can crash if you're not careful, saying you've got to be like a Formula One driver. And that's what I'm thinking is an important aspect here is Using these tools without fully understanding them and understanding

Th there's so much skepticism that I just have by default when I use these tools. And I see people using them without that skepticism. And thinking it's sort of just a, we can solve this quickly situation. And that is, that's what leads to issues. Like, like that's that's where we're making mistakes. That's where this stuff is no longer um of use.

But uh there was a study from the European Broadcasting Union and the B B C that found almost half of all AI generated responses had at least one significant issue with news integrity. Uh, University of Maryland study found that roughly nine percent of newly published newspaper articles were either partially or fully AI generated. Um

And the number of AI-genated articles on the web surpassed human written ones in late 2024, which is also not surprising. That often is just a bunch of the cruft that's out there. Um, but You know, there are people who I think I've seen use these tools in interesting ways, um, being able to I I I think I was talking to a Brar Al Hiti last, I think it was last week, about one of my favorite aspects of AI that is genuinely helpful.

And it's the it depends on which tool you're using, but I think on for Google it's called deep research, um for Claude. I think it's just called research. And I'm not sure what it's called for OpenAI because I don't use um Chat GPT. But um In in both of those cases goes out onto the web. finds a bunch of different sources and then combines them together into a research report.

And so one thing I did was I said, I it was around the holidays, I want to make um Christmas cookies. That's the one that, you know, that happens to be the holiday I celebrate. I want to make Christmas cookies and I want them I they have to be gluten-free because I'm gluten-free. The problem that I've always had is that they don't hold their shape. Even when I chill them, they don't hold their shape. I think it's because of the fact that they're gluten-free.

Um, what I want you to do is go out on the web, not only find the best gluten free. sugar cookie recipe, but also find comments and blogs and other information about what it is that could possibly make the cookies hold their shape better. And then let me know what you find. And I was able to get back a report. That had not only um a great sugar cookie recipe, gluten-free sugar cookie recipe, I mean, but then dozens of quotes and synthesis.

To provide me with the stuff that I needed to do to make the cookie better. Happened to be Xanthungum. And chilling for even longer than you typically would uh to get these cookies to turn out. And boy, howdy did they turn out. That's a cool thing, right? That's a that's a neat way of using it, uh helpful.

while she was working at Facebook, she just took pictures of as many files as possible and was just like indiscriminately like, I don't know if this is anything. I'm just quickly going through and taking pictures so that I can like have this information. And I remember at the time, like I was working at TechCrunch then too. And TechCrunch always has been a bit smaller of a staff than

like the New York Times or like Wall Street Journal. So there were reporters at the New York Times where they had like set teams where it was like your job right now is to go through the Facebook files and find interesting things. And at TechCrunch, we just decided like not to do that because we just didn't have the bandwidth to set up a whole team to do that.

And then I wonder now like how would that be different with AI if like could you put the documents into an AI and be like pull documents that involve discussions about children's safety? Like I don't know how effective that would be because we haven't tried it, but I guess

I don't know. Like, I think that I want to be as skeptical as possible about how AI is used in journalism because I think that this could go horribly wrong very, very, very quickly. But then also as someone who like reports on tech and like knows that This technology often can be good at things. Like, is there a way where we can take advantage of the fact that you can very quickly synthesize a lot of data really quickly?

It's wild seeing. I didn't know that there was going to be during my lifetime something that has is having such a profound impact on. the stuff that I care about, right? Like I knew that That there would be technology that I would find interesting, but it was always sort of hardware based in my mind. I would uh grow up and one day I'd be it'd be really cool that I've got a a wristwatch that I could talk into, like Dick Tracy. Um but the fact that

This is so transformative across all of the industries that are important to me, science, technology, and journalism. Um It's it I it's wild. And I just I never would have guessed that that's where uh you know where I would be, what I would be doing. And I'm glad that we. That many of us are still keeping a level head about things. I think that's important when it comes to this.

So together well we have all the horrors.

Across 180 countries and more than 50,000 businesses, Bitwarden is consistently ranked number one in user satisfaction by G2 and software reviews. With Bitwarden Access Intelligence, organizations can identify weak

reused or exposed credentials and take action immediately, while vault health alerts and password coaching surface risks to individual users in real time and guide them to fix issues on the spot, turning one of the most common causes of breaches into something visible, prioritized, and fixable.

And now Bitwarden is introducing the new Agent Access SDK, a powerful way for developers and teams to securely integrate controlled credential access into applications, automation workflows, and AI agents. This release does not incorporate any AI functionality into the Bitwarden solution and does not grant AI systems persistent or unrestricted access to Vault data. The agent access SDK is a separate.

Open source development toolkit designed to enforce secure, human-approved, and scoped credential access for teams that leverage AI agents in their workflows. Available now in early alpha phase for testing, the Agent Access SDK introduces a secure framework for how agents request. receive and use credentials, helping define a model for safe credential interaction in agent-driven systems. Bitwarden now enables passkey login for Windows 11, by the way.

securely unlocking devices at the OS level, and is working with Microsoft on native passkey support while extending SSO to automatically log users into more apps, making credential management across devices more seamless than ever. For those who want a lightweight option, Bitwarden Lite offers a self-hosted password manager designed for home labs, personal projects, or quick deployments with minimal overhead.

Bitwarden's open source code is regularly audited by third-party experts, and it meets SOC2 Type 2, GDPR, HIPAA, CCPA, and ISO 27001-2002 standards. So get started today with Bitwarden's free trial of a teams or enterprise plan or get started for free across all devices as an individual user at bitwarden.com slash twit. That's bitwarden.com. Slash twit. And we thank Bitwarden for sponsoring this week's episode of Tech News Weekly.

All right, we are back from the break. And that means, as promised, we've got an interview planned for you. It's one of those weeks where you look up at the sky and you remember that we are once again a spacefaring species. NASA's Artemis II mission has sent some astronauts around the moon, believe it is the first crewed lunar flyby in quite a while, and along the way shattered some space flight records.

Uh we will talk more about that. Here to break it all down is Rod Pyle, editor in chief of Aud Astra magazine, and of course the host of this week in space. Hello, Rod!

And uh, you know, we've tried this country, we've tried to restart a lunar program a couple of times, once under the first President Bush and then again under the second.

But they never really gained traction and it wasn't until twenty sixteen when Trump came into office. At that point we were working under something that President Obama had implemented to do an asteroid rendezvous. That wasn't really getting a whole lot of positive feedback from the press or the public, or in some circles of science.

So Trump said, Let's go back to the moon. Now in some ways we can thank China for this because we were hearing the earliest rumblings of them planning to land their astronauts on the moon. And I think we all have observed that America seems to to be at its best when it's challenged by something. And so I think that's what it took to get us going again. So this particular flight, the first crewed flight of the Orion space uh spacecraft,

And this is really an engineering mission, uh, as you pointed out in the message you sent me. Um You know, it it's of course getting great acknowledgement from the public and and it's there's a lot of romance to it. You know, we're exploring again, we've moved out of Earth orbit again and all that.

But the core here is we have to test this machine. So this is a whole new spacecraft, looks a lot like Apollo, but isn't at all like Apollo, which was a child of the nineteen fifties in a lot of ways, with early sixties technology. And uh I'm working on a book project with Jerry Griffin, who's a flight director from the Apollo era.

And as we went through our sixty or seventy hours of interviews, we both agreed that we're kind of amazed that Apollo worked as well as it did because we were just out of vacuum tubes then. And transistors are about the size of your thumbnail, you know? So this is a very, very different mission. But what they're looking to test here is the trajectory, which is different from Apollo. It's just a flyby, not an orbit.

Um, not a complete set of orbits anyway. And you got a test life support, which they've been on for nine days now. Which is critical and a lot of those systems came for the space shuttle. Of course, as the press jumped all over, we're testing a new toilet. Which is very cool. It's actually a little compartment in the floor of the spacecraft where you can go in and close the door and kind of hunch in there and do your business. And they finally got the bugs mostly worked out of that.

And um i i it to a very large extent we're testing ground control. You know, the people at NASA have never overseen a crewed mission yet leaving Earth orbit. And it's different. It's really different. I mean, when you're at the space station or in an orbital capsule, you can get back within hours or maybe a day or so. Whereas from the moon, if something bad happens on the way out, like Apollo thirteen, you've got days and days before you can get back.

So this is really a test of the systems and the people on the ground to make sure that they can pull this off for these longer missions coming up. Artemis III and of course Artemis IV, the first landing, which we hope. Will happen in twenty twenty eight. That's the plan.

Uh you know, you have holds, you have scrubs, you have hydrogen leaks, like we saw during some of the testing and all that kind of thing. So so really none of us were prepared for it to leave that day. We thought, Okay, well it'll slide to the third or the fifth or something and they'll have these twenty four to forty eight hour stand downs while they detank it and figure out what's wrong. It didn't happen. They had one ten minute hold, which I think was planned.

And then off they went, and uh cheers erupted all around. So the launch was a big deal. Then once they're in lunar orb uh excuse me, in earth orbit, they have to fire up that upper stage. It's called the ICPS. and get them into a higher orbit, which went perfectly. So they unlike Apollo, which just departed from low Earth orbit, this went to a much higher orbit and uh departed from there. I think it was about 1200 miles at its highest.

And then at that point they had uh already cut loose from that that upper stage of the rocket and they did some maneuvering exercises which were not critical but were very cool. Uh that's to test the handling of the spacecraft with the with the maneuvering thrusters.

And then they fired the service module engine, which is the the little cylinder on the back of the Orion capsule. It's not very powerful, it's about ten thousand pounds, so maybe I don't know, a sixth of what a jet engine would be on a on an airliner.

But it was enough to kick them out of this high orbit. So the upper stage had done the majority of the work getting them up there, and then this just nudged them out of there. So that's the translunar injection burn. I think it was just under six minutes. And again, work perfectly. And what sort of shocked me was usually once you've made that that insertion, you have to do a number of corrective burns.

to make sure you're in exactly the right trajectory because in this case, since they're not going into lunar orbit, they had at that point to try and get as close to the right trajectory to fly past the moon and come home as they could. That's called a free return trajectory. So it's one shot, loop the moon, and as they said in the movie Apollo 13, Isaac Newton's in the driver's seat and he's bringing you home, right?

So from that point, when you're leaving high earth orbit, you have to be more or less prepared to come back to the earth within about two degrees. If you're higher than that, you could bounce off the atmosphere and see you later. If you're too steep, you could come in and actually the heat shield would fail. So that was a critical maneuver. They had six other small rocket firings in the schedule to do if they needed them. But uh as of now, I think they've done three.

All pretty short. Again, not a very powerful engine. One of them was of the maneuvering thrusters. So they really nailed it on this thing. And this is kind of the miracle of the twenty first century, you know. You're not using these banks of IBM three sixty mainframes and all that. You're using modern computer technology. It shows inside the spacecraft with all the glass panels instead of gauges and all that steam powered stuff they had in Apollo. And they just nailed it. So at this point

after this pass behind the moon, which we'll talk about a little bit, they're just coming back. And uh all things being equal, they should hit the atmosphere tomorrow afternoon, and they'll come down about a hundred miles off the coast of San Diego.

Um, w uh, my understanding there were some records broken. Uh, what was that moment like to watch unfold for you? How significant is it? I mean, have we done something truly sort of Did you think we would get there eventually, or was this kind of like a oh wow, this is the the we've finally done it?

Not quite as long as they turned out to be. We were supposed to be doing this a number of years ago, but we got there. Um, there was a lot of stuff to overcome on the way. The heat shield is not behaving as they had expected, so they had to make some accommodations for that. And of course, again, the SLS rocket, you know, we had those hydrogen leaks and helium leaks and so forth.

But you know, once they got going, you're right. It starts stacking up these records. So first time we've left Earth orbit in fifty four years.

So that upper that that rocket propulsion unit behind the capsule isn't as powerful as Apollo's was. So because of that and a couple of other reasons, they couldn't go into a low orbit. So they had to go into this high orbit behind the moon. So they broke the record. We were we were not in Michigan Troll at that moment. We were in the press center.

'Cause you know, they don't want a bunch of looky loos standing in the observer's gallery, you know, pressing against the glass and cheering and all that. But um there was a message read up from Bishop Control and then there was a message read back from the crew. And you know, it's funny, in the Apollo years When you heard this chatter from the ground loop and from the astronauts back and all that

You know, those guys were test pilots, so they were talking like this. We're very proud to do this for our nation and so forth. It was all ki it was like listening to a sports guy. Teams pulling together, go team, you know. So it it was kinda like that. And then the guys on the ground were mo mainly just speaking tech speak.

This time there's a scripted message going up to the crew, a scripted message coming back, but you could tell that the crew was having a blast. Oh, you know, it sounded again like the moonwalkers did in the 1960s, early 70s. They were really enjoying themselves. So they read back a message about what this this record meant. But, you know, it was really one of those things where uh irrationally, if you're an observer, you're sitting there waiting for okay

I I what am I gonna feel here? And it just kind of went by. You know, it was a scheduled event. But when they started actually doing activities behind the moon as they were looping behind it. And they were only out of touch for about forty minutes. And and in Apollo, when they went behind the moon, they had to either fire a rocket engine to slow them into lunar orbit or critically fire the rocket engine to get them out of lunar or lunar orbit to come home.

There's only one engine on Apollo eight, so if that failed they'd still be there. They were just cruising. So you know, there wasn't a lot of pencil chewing going on while they were behind the moon. But there were some some really beautiful moments. One of them was um naming a crater after Reed Weissman, the commander's wife, who's a fellow astronaut who had died in twenty twenty two, leaving him to raise his two teenage kids.

And uh i I guess the crew had come up with this idea while they were in quarantine the week before.

So that was so cool. That was really one that y there was not a dry eye on the house and even a Mich Control you saw, you know, those stalwart people kind of

And he started with Welcome to my own neighborhood and then he closed it with Don't Forget to Enjoy the View and that man, that got everybody and and you can see the crew up there just kinda going w we didn't expect'cause this was a big deal. So that was that was really, really wonderful.

Was this a uh was this a mission that involved any new learning as far as what we know about the moon or other m because I know obviously as as you talked about engineering here, we're we're making sure that that life support works, et cetera. But there were images sent back. You mentioned the naming and everything. Did we learn anything new as far as celestial bodies go, or was it all just our own science?

So it's pretty well understood. So the viewing and science program they had for that backside turn. NASA made quite a deal about i you know, they were looking for, okay, what's new about this that the press can latch onto, right?'Cause'cause that's like a a a struggle they have all the time. And really the the message they kept putting out was it'll be the first time that the upper and lower attitudes of the far side will be seen by human eyes. So they've been photographed.

But the point was made by the science team that the human eye can discern things that photographs can't. So because they were so high above the moon, about forty-two hundred miles, as opposed to the Apollo orbits, which were between sixty and a hundred miles. they were able to actually see from the from the top to the bottom. So we didn't really know what to expect. One thing that did come out of it was they saw six white flashes while they were there.

And these have been been seen from the near side of the moon telescopically before, but they were apparently meteorite impacts on the moon, small meteorites hitting, and then this big burst of energy as they as they melt when they strike'cause of the high temperatures. So that was notable. And then of course there was a solar eclipse as they came around the far side where the uh

the the the sun was blocked from their view. Now normally on Earth, if you've ever seen a solar eclipse, they're three to seven minutes. They're pretty short. Um but the Earth or the excuse me, the moon's diameter exactly blocks the sun, so you see all of the corona and all that that activity going on there. In this case the uh

the sun was much smaller than what was blocking it. So it took almost an hour, I think it was fifty three minutes to come from one end to the other. But what they were able to see was the outer corona, which is a very different view than seeing the entire corona. And also one thing they were looking for on the science program was to see if they could s get an idea of how much dust is in the lunar exosphere. So the moon has a very, very slight atmosphere. Almost nothing.

But because it's a very elec electrostatically rich environment When the Apollo astronauts were there, they were covered in lunar dust all the time'cause their their spacesuits were acting like magnets just sucking all this dust up. So one thing they want to understand better is how much of this stuff is in the atmosphere. And a way you could do that is by rear illuminating it.

So they got that. But mainly they would just take a picture, picture, picture, picture. You saw the video, they darkened the capsule so that they could see out the windows better. I think they had four Nikon F fives, which is interesting or uh D fives. which is their top end camera as of about twenty sixteen. But apparently that chip was determined to be the best thing they could they could use. And then a couple of others.

And weirdly, this is a complete aside, but I kept waiting for Nikon to make a big deal out of this. We're going to the moon again, just like we did with Apollo.

They have three sets of parachutes that have to work perfectly. They have uh drogue chutes, which are the little ones that pop out first, do a little bit of slowing at the capsule, and then serve to pull out the pilot chutes, and then finally the three main chutes come out.

And the reason they do it that way is because if you just pop the main chutes out going the speeds they are, they tear and get destroyed. And parachutes if you talk to engineers, parachutes are like wild animals. You never know quite how they're gonna behave exactly, even though we've done this a bunch of times. So there's a lot of things that have to happen properly.

There's been a fair amount of concern about the heat shield, which you may have read about in the media. Um, on the first flight of the Orion capsule in 2014, which went up on a Delta IV rocket. There was a little more uh ab ablation than they had planned for. So the the the idea behind these shields is they char and as they burn that wicks the heat away and you survive reentry. So it's different from the shuttle's tiles which just resisted the heat.

Um, there was a little more charring and a little more removal of material than they were comfortable with than in the Artemis one flight, which took place a couple of years ago. They had the same problem. Now they're using the same stuff to make the heat shield they did in the nineteen sixties, but I talked to the chief of that program in twenty fourteen about it.

The chemistry had to be altered because of EPA rules and if if you're my age you remember spraying models with cryon spray paint and they dried like that hard as a diamond. Now it's a little slower and takes longer to cure, and it's the same basic issue with this heat shield. So what they figured out was um that the as the spacecraft was re-entering and the heat shield heated up.

There was outgassing inside. There was actual gas pockets being created in these blocks of material, heat resistant material. And it was causing them to pop off in in bits, not the whole tile, but but pieces of them before they would have liked them to. So for this particular mission, they're going into a steeper reentry, which means it'll be hotter but for a shorter period of time.

and kind of counterintuitively the extra heat and this is all on a on a graph somewhere that I can barely understand. That's why I'm a journalist, not an astronaut, right? But um It heats it faster and that allows for the gas to escape in a timely fashion and they think they're gonna be just fine. And in either of the test flights did the shield burn through, I should note. It just eroded more than they wanted it to. So everybody's pretty pretty confident this is gonna work out just fine.

This is supposed to be right the foundation for the lunar presence and even more after that. Um, where does Artemis specifically go from here? And what do we think sort of success on this mission unlocks for what comes next?

So Artemis three was supposed to be the first landing within a year and a half or two years. Now Artemis three is going to be an Earth orbital mission with an Orion capsule. And whichever one of the two lunar landers is ready to go, they're gonna do rendezvous and docking tests. So one's the SpaceX Starship, lunar version, which hasn't been progressing nearly as well as everyone had hoped.

And uh my money's on the Blue Origin version, which is a more traditional design, kind of like uh an Apollo lunar lander on steroids, but a lot bigger and more capable and it can stay on the surface longer. So Blue Origins flying one robotically later this year as a test, and then they hope to have one ready for the Artemis III mission. Then Artemis four is intended to land on the moon in twenty twenty eight. At that point

For five s five and six they'll still be using the hardware they have now. After that, at least according to what Isaacman said, they want to start using commercial hardware, including commercial rockets for launch because they're less expensive than than the um SLS, which is about four billion a shot. Which uh you know, if you look at it through the eyes of the Apollo years, that's not terribly expensive because the Apollo missions

I think all then, you know, these numbers are kind of flimsy because it depends on what you include and so forth. But about three hundred two hundred and fifty to three hundred billion dollars in today's money so far This program, Artemis's cost under half that. So they're trying to do a lot with less. But the the overall idea is look, we've got to get the pace of these launches up. We want to launch

m more than one per year, maybe even uh as often as six months once they get going. And the big idea is to build a base on the moon. They were working on this thing called Lunar Gateway, which is going to be a moon orbiting space station. That's been on again and off again for years. They finally officially mothballed it now. And so that's gonna be put on the back shelf. They'll either

try to figure out how to land uh'cause they build hardware for it. They may land some of that hardware as a moon base, but probably they'll try sending it off to Mars robotically as a test for an ultimate Mars mission. Um, and at this point they wanna put money, they've got I think two hundred billion year mark now.

Um, of course that's future appropriations depends on the NASA budget, which changes every year, to uh build a lunar base. And again, why are we doing this? We've been talking about it for an awful long time, just like we've been talking to go about going to Mars for a long time.

But China said, Hey, not only are we gonna land people on the moon, we're gonna build a lunar base with the Russians and other aligned nations. So it's this competition thing. So we've always wanted to do it, but I think that's what lit the fire under this idea.

No, it's more than you think. Your name, your contact information, social security number, home address, even information about your family members all being compiled by data brokers and sold online. On the web can buy your private details. And this can lead to identity theft, to phishing attempts, to doxing, to harassment. But now you can protect your privacy with DeleteMe. You know, as a a host of of tech shows. Obviously, I have an understanding of how little privacy we have online.

And the unfortunate reality of having so much of my personal information out there, how it can compromise my safety and my security. And that's why when Delete Me offered uh the chance to use the tool. I was blown away by how much of it it was able to remove. And I continue to recommend and use Delete Me to solve the problem. Uh it helps keep me protected online. Delete Me is an it's a

You need to be able to say what needs to go and then they know from there to to proceed with that. Delete me will send you regular personalized privacy reports showing what information they found, where they found it, and what they removed.

And then Delete Me, of course, is not just a one-time service. That's incredibly important because those data brokers continue to gobble up your data. Delete Me is always working for you, constantly monitoring and removing that personal information you don't want on the internet.

To put it simply, Delete Me does all the hard work of wiping your and your family's personal information from data broker websites. So take control of your data and keep your private life private by signing up for Delete Me. at a special discount for our listeners. Today, get twenty percent off your delete me plan when you go to joindeleteme.com slash twit and use the promo code TWIT at checkout. The only way to get twenty percent off is to go to joindeleteme.com slash twit.

And enter the code TWIT at checkout. That's joindeleteme.com/slash twit. Code TWIT. Thank you, DeleteMe, for sponsoring this week's episode of Tech News Weekly. Alrighty, we are back from the break, and this final interview actually took place earlier today. Uh, but let's head into it and give it a listen.

Apple just recently turned 50, and while the company has spent the last five decades reshaping how we interact with technology, from the Mac to the iPhone to the Vision Pro, the next 50 years might be even more unpredictable. With AI disrupting the landscape, and of course, having uh new technologies, spatial computing still finding its footing, economic headwinds reshaping who can even afford this premium tech. The road ahead is anything but certain CNET.

Took a big swing at mapping out where Apple might go from here, pulling in futurists, expert opinions, and a whole lot of informed speculation. Here to walk us through it is CNET's Scott Stein. How you doing, Scott?

Oh man, scary stuff because it's just, yeah, it's a lot to try to to process and think about. What made you want to take on the challenge of looking ahead at where Apple might be in another 50 years? And then frankly, how did you approach a question this big?

So the thing I realized is that while fifty years does seem like a long time, I kind of had this perspective, honestly, my own age, where I go like, okay How many decades have I been around and how long have I been at CNET? And even when I started at CNET, I thought that laptops would be gone and they're still not gone. And for over a decade, I thought the iPad and the Mac would merge.

And they still haven't really fully done that. So maybe, maybe maybe that's the right time span to look at for some of these things where I'll say, oh, it's coming. And it could be taking a lot longer than that to actually get here.

And I kept it mostly just that conversation. And we talked for for a while, riffing about all these things. One thing, not to speak on her behalf, but to say one thing she really impressed on me was not being a a futurist myself.

that futurists look at possibilities and possibility landscapes. It made me think of Doctor Strange and, you know, the multiversal paths and that it's not a it's not an absolute set of predictions, but but you look at these you look at these different directional thoughts, and that's part of the game. Of futurism. So I don't even know if I really fully represented that in the story either, but we tried to kind of approach it as we chatted. What you thought about with

With that was one of the first things we we brought up was the idea of, you know, who will it be serving? Who will Apple be serving? And she brought up this thought about which I've often thought about is that who can afford the tech? And Would this be a company that I mean, we already talked about Apple going towards higher end design and appealing to higher end uh prices and sensibilities? And then there are products like the Vision Pro that are that literally are that. And

um, you know, eight hundred dollar wheels for computers and and whatever. But Apple is also a company that is in the mainstream and has a lot of I mean iPads are very affordable. Um the MacBook Neo

is extremely affordable. Um iPhones can be affordable with trade in plans. You know, it's not like that it's clearly not nobody can afford these. Everyone affords these. Um so I I wanted to put that in perspective too. But I think the thing that was interesting was was that You know, if there's uh one of the possibility sets is like if there's an increasingly bifurcating like divide between the rich.

and and and people who are poor, which is already happening now. Like, you know, science fiction books think about this. You know, w you know, we're are we in this point where there's a an oligarchy of, you know, people that are increasingly at the high end. I are there companies that cater to that? Who are those companies? This feels like Cyberpunk territory, but, you know, if we're thinking about this,

Where does Apple land in that space is something that we were you know, we were bringing up, um, for sure. I think they feel like they're straddling it, you know. Um where it's it's still pretty mainstream, but but it's also, you know, she brought up a lot of people

um in i who who can in in studies that can't necessarily afford things uh th w won't go with iPhones. You know, they'll go with something that's more affordable, an Android phone or something like that. So Uh that's something we want I put at the forefront.

that could actually change how we capture and revisit memories.

phones and and watches. So I'm saying that as a preface to I was wondering about this myself, how much spatial really be a factor. But I think what's interesting to me is the more I look at the spatial landscape and think about it.

I remember from the very beginning learning from companies that it had tendrils in other areas. I I remember going to a Microsoft HoloLens 2 event where uh Alex Kidman was talking about like the edge compute uh picture of things and that it that's where it began to get in my head the idea that headsets And cars and robots and NVIDIA gets into this too when they talk about things like Omniverse.

a lot of these things are are kind of solving for each other. And that's what got interesting to me is that yes, there are VR headsets. Yes, there are smart glasses. But when things start increasingly Recognizing the world, scanning the world, and being able to understand that contextually with you. with visuals and audio and other sensors, then that can be done

for other things too. And self driving cars use LiDAR, headsets use LIDAR. Robots will use SLAM navigation headsets will will use SLAM. You know, it's like so there's There are similar sets in in computer vision, all these things. And so

I think now as AI has been the ultra hype territory, I I was hesitant about like, well, how much is AI part of this? AI feels almost like the internet where it's like, yes, it will be there, but what are the things within it that will be interesting? Um that a lot of stuff now in the AI landscape is talking about world models and

contextual AI, things that like, oh, we understand the world and we understand what to do with it. And I kept thinking, this sounds like the stuff I was thinking about with VR AR headsets. The it was the same conversation. And I it's no accident. I think it's pivoting to stuff that probably Might have more applications or there might be more investments for that. But it's like once you solve for one, you might also be solving for the other. So a lot of that spatial stuff.

pretty big and I think that's why you see people like Tim Cook and others talking about spatial is that I don't think it's just about what's on your face. But but some of the tech I see too that really blows my mind. Uh one of them is is Gaussian splats. And this is the sort of stuff where you don't really see it a lot in the everyday world or may not be aware of it. But it's it's three D scanning, it's been a been around for a little while, but it's becoming more mainstream.

the true sense of capturing stuff that feels like you're in the space, like a real Real 3D capture. A lot of things now from people who do it well. You can see it in Google Maps. You can see it in um Apple's Vision personas use it. Um Meta had it in this app for the headset where you can look around your room and then capture it in 3D.

And it feels like you're stepping into a a a space that gets increasingly more detailed. And there are ones that are doing this mixing in video too. There's some video splats, so it starts to feel like a like a holographic capture. Where I think that's really interesting is that it goes to something that we didn't quite talk about yet, but Apple is a company that's really into camera tech. Like I didn't even get into this to the degree that I probably could have, but like

Computational photography and cameras, I mean there's the Artemis mission with the iPhone, like everything they're doing, even on the Vision Pro is shooting things in immersive, they're very camera focused. And I think that their evolution of the camera tech is is gonna be a huge part of their the future, especially since their idea of memory capture right now, like spatial three D

spatial stuff is really just like stereoscopic right now. Yes. But if they can get to the point where those cameras are doing consumer level like Gaussian things where they can, you know, really scan your room and have a a memory that you can revisit.

That's stuff that who knows what you'd have. Something you could live on a headset or pull out pieces of it like Blade Runner, where you know it's like z in Zoom and in ants and or have things that um you know, could could have like value for training or recognizing your environment or robots could use it or again. There's a lot of interesting zone for I talked forever about this, but that's why I got so excited about it is I see a lot of a lot of tendrils.

You've got all this new tech, right? But Apple really is the iPhone company. And so looking at the future of the iPhone itself. i i do phones disappear? Do they evolve? Do they just keep being the center of everything? Where that's that seems to be Arguably one of the hardest crystal ball questions to ask. And I don't know if there is a crystal ball out there that any of us can know for sure. But yeah, what's your take on that?

Yeah, they change, but it's like um laptops, which again, like I thought those would be gone and and nope, yeah, they're they're still really useful. Um phones Are so essential right now, I just don't see how they will disappear. Um, and the reason I think that is that that. You know, people talk about stuff like glasses maybe replacing phones.

Potentially, but we're not um the thing that's interesting about phones is that it's our bridge to the online and the digital. It's our like little It's our it's our little pack. to interface with that. And'cause we're not cyborgs yet. I mean

Not for a while. And maybe it's something I thought I got much more interested in the idea of like Does the phone just kind of sit in your pocket and then you have a lot of these peripherals that are interfacing with it and it's like your power pack, um, which is your glasses and your rings and your whatever you might have, but you know, AI pens.

Yeah, or you know, it's like it it it runs it, which is what Meta was showing with with Project Orion a couple of years ago, and Google has with their Project Aura, um, is kind of a similar model that's coming out this year. So I thought, yeah, but then you have all these folding phones. Like you've then some people are like, oh no, the phone unfolds and becomes like in the Westworld movies it's transparent or whatever, it's it grows pieces.

I I feel like it could be both, you know, but this uh that suggests to me that like The phone. I saw Neil Stevenson commentary recently, which I think was very I got kind of triggered me. It w it was very specific about kind of like the m headsets are over and, you know, phones are what everybody has. And I think he was going on a very particular point.

But I felt it was kinda silly because it's a yes and versus a a a binary thing to me. Um in that the phones as they get more advanced will power a lot of things you may optionally wear. But yes, the phone will still exist, but I don't think the phone in twenty years will look like a phone we have now. I don't even think most people want that. Most people are like, I'm sick of

That's the funny thing, is that people talk about, oh, everyone is so used to looking at their rectangles. But most people are currently sick of looking at their rectangles and feel kind of disgusted at social media or they feel like, oh, I need this. But so I don't think it's a perfect device. And that's why I think it's going to change because I I think if it was perfect, sure, but I don't think anyone has acknowledged that it's perfect.

Spot on there. Um, there's a in your piece, there's a section on wearables, on health monitoring, uh, beyond what the Apple Watch does today. And I think it's one of the other places where there's a lot of exploration. Assistive tech, neural interfaces, smart fabric. Uh, when you were doing this and and sort of tr conceptualizing the future there, what are some of the most exciting or unexpected possibilities that you came across?

into a landscape that could be like the, you know, is a singularity gonna happen or is it not? Or is it or is this already kind of sort of the singularity? Like are we already is is the acceleration of tech going to keep getting weirder?

The re and what I was thinking is that there's a lot of comfort in tech. Like when I talk to my mom about using the phone, she wants to be able to understand it. She wants she she she likes what she likes. She likes the MacBook Air for what it does. And we're having generations now that are now gr have grown up natively with the products.

So the the what I this is all getting around what you were saying is that I think we're gonna have people that have lived in this comfort of the tech for a long time and Apple will have an audience of people who are like real legacy people that they're catering to.

And so part of the caring and the digital tech is also like how much data and how much stuff you may already have now going back like I've been weighing myself since twenty thirteen on a smart scale, but like you will probab you'll get to the point where you have decades of of Apple Watch data and

What do you do with that? And and so part of it, I think, about smarter systems that can know your whole like life arc, which is kind of creepy. But also like Apple's already into stuff like um using the Apple Watch to monitor remotely and A lot of safety and awareness things, f fall detection, um, you know, AFib or or hypertension. But Annie was bringing this up too, like, you know, the ability for them to get into

intelligently designed uh smart care that might keep an eye on you or help you out or be your companion. Um, and there's a lot of companion AI tech out there. And I think a lot of what she was bringing up and I was thinking about too is that Apple often gets into categories late versus early. And then they will aim to do a kind of thoughtful, intelligent and and trustable design for it. And so I think this is one of those areas that's certainly like that where how do you feel comfortable

putting faith in something to let's say there are cameras that, you know, monitor you or or things that are increasingly sensor based where it feels like you might be under surveillance. And is that creepy or is that useful? It depends. Like if it means you didn't have to have

If you could live more independently and feel like something was was monitoring you and helping you. I don't know, it's interesting. I think we're already in that s in that space and in the next fifty years, that seems like where a lot of health tech and preventative health tech You know, might be going and they have the w they have the watch already, they have all that health research things they're doing and there there's already a lot of stuff in that zone.

And then I even talked to my um it was just I put it in the story. I love it. There was like a a moment at the end where I was talking to my thirteen year old and I was like, Yeah, where do you think Apple's gonna be? And they're like, They'll be dead. And I was like I think you said dead and I was like

What do you mean? He's like, I don't know. There would be around, you know, and and for him it's just like an iPad. You know, it's so it's not like when I talk about Apple it feels like a huge company. And for him, he he's like, Okay. I mean, I don't know. I yeah, I don't I don't That I don't know if that really matters. And generationally, it's an interesting question. You know, at one point

I think about that with social media. You know, at what point will social media possibly just not matter to a new generation and they just don't really care and it's passed out. Um I think there's also questions like uh I remember hearing this and this is like something where I'm anecdotally remembering this. about when things stop becoming a

a fad and and have long term hooks. I think it's like you have to survive fifteen years. Like when you when something like the Transformers or you know, Super Mario, like when when things eclipse and become something that becomes timeless, which Apple obviously has crossed over that at fifty years, like well, well, many times over I I would think they're still gonna be around for the sheer financial size and

the number of of footprints they have in tech, but Annie brought up too about creatively destroying yourself in order to move forward. And I thought that was interesting. I think it's something that Apple's already done in some ways. I think about the iPod and I think about I mean, I don't know if that's the right definition of it, but to me it's like, you know, they've had moments they've

s they've gotten rid of their own tech to transform other pieces of tech. And so it wouldn't surprise me that Apple has several surprising transformations in those decades where it's not like I kept thinking about like are they gonna be making furniture? You know, are they like are they just gonna be making like shoes or that aren't even tech? And I don't know, but I think um They're already involved in film and TV.

uh very advanced fabric design and material design for things like the watches and the vision pro bands. I mean, like you pay attention to it and they they take great pride. They you know they'll always like go into like kind of a it's like a kind of a joke where they go into all that detail

the m the the one of those proof points of like when I looked at AirPods in 2016 and they were like a meme joke of uh hell no, I'm not wearing those. It's ridiculous. And then everybody's wearing

moments. Yes. Um'cause those become things that la like the iPod that last like decades or decade plus and um That's where I think about their survival. I I don't know how long they're gonna do TV and film, but their presence and services. is interesting. Like that's stuff that how how dependent do we get on that and how much do they keep building that out? How much do they interconnect with other things becomes also how essential they are. Um

And where they get involved with AI, although I still feel like with AI, it's I don't know. A AI's tough for me. I feel like a lot of it doesn't interest me. And then um Yeah. Yeah. And then it's like honestly, and then some of it's intriguing, but it feels like glue. It feels like maybe it's the infrastructure for stuff that people will come up with applications for. To me, it's like why smart glasses are interesting. That's why

Or if Apple's trying to come up with like, oh, this is what or even Soro is like, oh, this is what you can do with AI. It may seem stupid, but like this is a thing. I feel like it's hard to find those things. Claude code is like a thing, but um I think that'll be interesting too, is like when do companies find those things that and computational photography is like that too. Like AI knitted together with visual capabilities is gonna go wait in all sorts of

totally wild directions that I don't think are about generative AI. I think they're about like other things too. Huh.

for helping us. So be sure to check that out. Along with that, you'll get some access to special feeds. We've got a before the show, after the show feed that has special bits and clips. We've got a feed That has our live coverage of tech news events like the upcoming uh Apple WWDC. And of course, we have our feed that has our special club twitch shows like My Crafting Corner.

plus so much more. Stacy's book club, uh, coffee time, etc. If you would also like to join our Discord, that's something else you get. Access to our Discord where we have uh folks who are part of the club and also those of us here at Twits chatting with one another.

So be sure to head to twit.tv slash club twit to check it out. If you'd like to follow me online, I'm at Micah Sargent on many a social media network where you can head to chihuahua.coffee. That's C-H-I-H-O-H-O-A.coffee where I've got links to the places I'm most active online. Uh be sure to also check out my shows that'll publish later today or may have already published, including iOS Today, Hands on Apple, and also check out Hands on Tech, which publishes every Sunday.

Thank you so much for being here. I'll catch you next week for another.