← Big Technology Podcast
Big Technology Podcast
Did Google Just Fall Behind Again?, iPhone Fold Cometh, Anthropic Files To Go Public
01:11:47
Loading summary
Transcript66 lines
- [00:00]
A
As Anthropic's Claude Code and OpenAI's Codex take off. Where is Google exactly? The iPhone fold is on its way. Will we see it later this month? And do Meta employees still believe that's coming up with MG Siegler right after this. I'm just back from ServiceNow's Knowledge 2026 in Las Vegas and the conversations I had there are ones you're going to want to hear. I sat down with their president and CPO Amit Zaveri on the platform strategy, powering enterprise AI, Chief People and AI Enablement Officer Jackie Canney and Chief Digital Information Officer Kel romac on what AI really means for the workforce, the technical leaders behind ServiceNow's Nvidia partnership on shipping AI at scale and Ulta Beauty on deploying ServiceNow's technology across 1300 stores. If you want to know where enterprise AI is actually headed, not the hype, but the real story, you can find these videos on my YouTube channel. Search Alex Cantroutz on YouTube. Depending on who you ask, between 80 and 95% of enterprise AI projects fail to get AI to work for you. You don't need more tokens, you need better people. A board pairs powerful propriet military tools with senior engineers who've seen it all. That combination means your project doesn't stall, doesn't drift and doesn't fall. It ships. Whether you're a startup that needs to get to market or an enterprise with complex legacy challenges, Aboard delivers exactly what your business needs. Fast Aboard is your partner for AI transformation. Visit aboard.com and let's build something together. Welcome to Big Technology Podcast. It's the first Monday of the month. It's actually June 1st, so we have the first Monday of the month and the first day of the month. And we're joined as always, by M.G. siegler, who is here for his monthly spot to talk with us all about Google's AI efforts, the big WWDC news that's forthcoming from Apple, and what is exactly the deal at Meta. So mg, it's great to see you. We have a lot to talk about. Welcome back to the show.
- [01:51]
B
Thanks, Alex. Yeah, right in the middle of like all these conferences going on IO, wwdc, there's Computex going on like all these different things. So lots going on right now.
- [02:03]
A
Definitely. This is, you know, one of the hottest times of the year for tech. All the big developer conferences. Yeah. Smack in the middle of Google I O and wwdc, which is great because we'll be able to talk a little bit about what we saw from Google and what might be coming from Apple. Let's just talk about Google to start with, because we saw, I think what we would both agree is an underwhelming Google I owe. But the question is, does this really put Google at the back of the pack again? And I think there's a chance that there's going to be some of those bigger issues that we all thought were coming for Google at the beginning of this. And they kind of pushed aside, you know, as they started to build, these great models might be sort of coming back. And so today will be a good time to discuss that. Let's just go quickly to the news, as you put it. They held a giant event to talk up a bunch of products and functionality coming soon. But Gemini 3.5 Pro, their strongest model, just doesn't seem to be ready yet. And while other AI labs do previewing at their developer events, Google has this big Google I O event that you would expect there to be some more meat on the bone. So I think we'll start micro and then go macro here because, yes, the model doesn't seem to be ready, but maybe the bigger issue is they don't have the Codex and the Claude code competitor that you thought they would at this point. But begin if you could just with your reaction to the model news and what you think might be going on there.
- [03:41]
B
Yeah, so they did release Gemini 3.5 flash, right? And they made, they made sort of the whole count, the whole announcement sort of based around that. They, to their credit, they rolled, it seems like pretty much every product over to that new model. But as you're noting, it's not the flagship one, Right. It's not Pro, and they had to sort of awkwardly say, like, Pro is coming, I think next month. So presumably sometime this month in June, now that we're in. And, you know, so they sort of wanted to address, I guess, that elephant in the room because in the lead up to it, right, everyone assumes, like, oh, it's IO, Google's going to make a big splash like they do every year, and they know that there's competition as there is every year. And oh, by the way, Claude just rolled out a new model and OpenAI has rolled out a new model and so therefore it's going to be time for, of course, Gemini 3.5 Pro or something else to be ready. But the backdrop of all this is also with the stuff going on with Mythos on the anthropic side and sort of a killer model above and beyond. Even what's at the flagship, the front end right now, sorry, the top tier of the models right now. And so again they come out and they basically, again, again have to lead with the notion that, yeah, that we don't have 3.5 Pro ready, but we have this great Flash model and it's super fast. And then, you know, once they release it, it's sort of like after the initial sort of momentum, people start digging in. It's like actually it's, you know, it's a little bit expensive and it's, it's good but it's, it's certainly not taking them, you know, steps above what the other, you know, frontier model makers are at right now. And so, you know, why even bottle bother to do this at their flagship? Why not sort of wait. And that was basically the, the angle that I took. You know, I've been going to these as if you. For years and years. I haven't, I didn't go to this one. And you know, obviously I worked at Google for a long, long time, but it's, it's an awkward situation in that this is their big conference every year and again, they want to seemingly make a big splash at those conferences. But the, in the age of AI, it just seems like these announcements don't, don't roll with that cadence. Right? Everyone sort of followed the Apple model of, of, you know, always doing these big yearly events and, and having something, and holding something back for a little bit to be ready to tee it up at these. And it just feels like that doesn't work right now because again, Google looks sort of silly for not having their model ready to go. Maybe they should have done WWDC or another event in June when they're ready to go with the, the Pro model.
- [06:15]
A
Yeah. MG this is why I decided to start here today. For us, it's almost like a double whammy for Google, right? Because, all right, you would think that, and I'm actually surprised that there hasn't been more discussion about this maybe because the momentum has been so good around Gemini that people are reticent or hesitant to, you know, say anything bad or anything, you know, potentially skeptical about Google because they've shown the ability to ship. So we'll give them credit there. But this is, to me, you know, potentially, you know, double worrying for them. First of all, they didn't ship the Gemini, the latest edition of Gemini Pro. So if you think about foundational models, they have fallen behind OpenAI and anthropic with mythosin 5.5 as far as we can tell. But to me the bigger issue is they are not playing in this quote unquote super app area. Right. It seems like their bet is make the models efficient, which they have with Flash. Right. That's the big draw for Flash is that you can effectively deep seek your way through AI as you make it super efficient and then you can do more Jevons paradox, et cetera, et cetera. Okay, cool. But for Google, you would really want to see them have a competitor to these Claude Code and Codex apps where, which has driven so much of the growth in AI right now, which is sort of applicable for coders today. But over time, if the AI labs see it, you know, see their way prove out everyone will use these type of apps. So I'm curious to hear your perspective on this. Is this as big of a liability for Google as I'm imagining?
- [07:52]
B
I. I mean I do think it's a problem. You even heard Sundar Pichai was on, you know, with, with Casey Newton and Kevin Ruse on their podcast and he explicitly said like just straight up like that they're a little bit behind in, in coding, you know, with regard to AI. And everyone knows that this is sort of the, the forefront where you need to be not only from a actual coding perspective, but because a lot of people of course think that this is what is the key for sort of the agentic use cases going forward and you know, potentially self, you know, these models that can teach themselves and whatnot and we'll see how that plays out. But still it's clear that everyone recognizes from OpenAI which realized they had to sort of, if you don't want to call it a pivot, they had to reorient the entire business right around. Yeah building towards the super app and bringing in Codex and putting it all together, which they still have not done yet. But presumably we're, we're closing in on that happening. And now yeah, Google recognizing and acknowledging that they're not, they need to be with regard to coding. And there have been reports, you know, that, that everyone from Sergey Brin on down is sort of all hands on deck literally to make make sure that they can sort of catch up in this world. Because of course they're Google, they should be. They have the best engineering talent that money can buy. They should be at, at this sort of forefront of this movement. And yeah, there's, there's just there for whatever reason, they either were focused on many other things you heard at IO, there were a lot of announcements all around search which was interesting, right, because obviously that's their main business and it's long been, you know, their calling card and they've done, in my estimation, a good job of, of sort of making it so that search has been completely reoriented around AI without making it totally destroy the business. Right. Like even they said, like searches are. We've had more searches now than we've ever had before. There's probably some caveats in there of, like, what that actually looks like, like how much of that is back and forth, you know, with, with a mode ve versus, you know, straight up just typing into the search box. I don't know the answer to that. I don't know if that anyone does. But still, again, it has not disrupted their business in the way that everyone was fearful of, say, 18 months ago. And of course the stock and everything has bounced back and they've, they've had this immense ride that they've been on, you know, from a market cap perspective over the past year, which we've talked about before. So again, they've done a good job of that. But that's like they focus so much on that and maybe they should be focused on that because that is the core of the business. But maybe taking their eye a little bit off or just having not even, even a company with that many people, maybe not having, you know, the right resource allocation in order to focus. But again, to me stepping back, it just looks like a lot of these companies, Microsoft, Google, OpenAI even, like they didn't realize what the true value was in what anthropic was going after and no one really realized it until Anthropic just started to explode from a business perspective, right? If they're at maybe closing in on 50 billion in ARR now and all of a sudden valued more highly than OpenAI is, and everyone's like, whoa, maybe we were not focused on, even though we're focused on important things and we think we're doing great stuff in AI, maybe we're not focused on the thing that we need to be focused on right now.
- [11:08]
A
Right. And this is sort of why it's bringing up some of the old questions that I had about Google and many of us had about Google, which is how are they going to play if the entire concept of the web sort of stands on its head because of this AI moment. And so my perspective about this, and I wrote about this in Big Technology on Friday and definitely want to have a conversation with you about it, is that, you know, there was this notion that the idea of a Super app was a misnomer, right, that they, that let's say OpenAI was calling Codex a super app because they were bringing ChatGPT and coding and a browser together. And that's not your traditional super app, which is like what you would have in China and which is like one app that does lottery and you can hail an Uber there, you can do payments there, your bank account is there and it's all in one. And everyone's like, well, that's not exactly a super app, including myself when I first heard it. But you know, after experiencing these products a little bit and listening to people like Greg Brockman and Boris from, from Anthropic, I think super app is the right word. And you know, as you use these tools, they're not just, you know, coding is what they do best right now, but they're not just coding. Right. They're specifically designed to take over your computer and your browser and get things done for you. And so they started with coding, but they're going to do far more than that if these companies have their way. Just one example that I gave in, in my article was that let's say you're trying to hire someone for like an entertainer for a child's birthday. Like right now you would do the research stage in ChatGPT, but you would still go to the website to book it. And to me, you know, I think these labs view that as an accident of history or as an incomplete path. Right. They want everything to happen either within the app, which they've tried to do, or the app actually takes control of your computer, emails the entertainers, see if they meet your specifications, if they're available, maybe runs a background check and then books them and that turns the entire web on its head. So I'm curious what you think about that thesis and if that's true. I think it is then Google sitting this out or not being the leader is actually a potentially company destroying liability.
- [13:28]
B
And a big part of the other showcase of IO was obviously the product they're calling Spark, which is weird because there's so many different sparks now within AI. Meta has their spark and I think. Does Microsoft have their spark? There's another spark. Oh, Nvidia has a spark. I think too. There's all these Spark products and they're all different products, but they're all using that name, which is very confusing from, from a branding perspective. But anyway, Gemini Spark, I guess, which is not out. Some people have, I think, early access to it. If you're on the, the ultra tier, which is also confusing because now there's two ultra tiers of, of of culture teams beyond that. So some people are able to test it out on that tier. But still that's sort of speaking to some of what you're talking about. But there's also, I go back to the notion of how long it took them to release a standalone app for Gemini, right? Like Google thought like maybe they would just be fine doing this all in the browser because of course they famously control the most popular browser in the world in Chrome and Gemini is now baked into Chrome. And so I think that they might have thought at one point that that would be good enough and that might be their, you know, quote unquote app for, for AI. And as it turns out, like they needed to eventually roll out the standalone Gemini app, which they now have. But again, to your points, it's like it's hardly a super app. It's super rudimentary. It's just very basic. Like this is a way to access Gemini, you know, natively on a Mac right now. Now obviously that's setting up, you would presume that's setting up to eventually roll out Anti Gravity, which is there also, I think awkwardly branded, you know, coding product and then yeah, eventually roll out Spark or whatever they're going to call sort of their agentic full on computer use tools. But still they're behind the eight ball on all of these things right now. And I do think it's partially because of exactly what you're hitting on, which is that it's, they're sort of ingrained in their nature to be this sort of web centric company. That's how they've been throughout their entire history. And so when you talk about like yeah, making these native apps that are using native tools on your computer and taking over the computer usage and not necessarily doing this all just through a web browser. I think it's a little bit of a, of a mind shift that needs to take and, and a mindset that needs to be altered in order to do some of that. And I think they're trying to work their way through it. But that's probably why you see them now again, a few steps behind and not ready to launch all these things, you know, at the same cadence that the competitors are. And just to go back to the broader point about the super app stuff, I, I am in agreement with you. It feels like there, there's a movement right now for, I've written a little bit about this. There's a movement of all these, all these services to come from Airbnb on down, they're all trying to come up with their quote unquote super app. And obviously, as you're noting, that's coming from sort of the Asian markets where you know, there's players that really do a lot of your life sort of rooting through these things. But in the Western world it seems like it's going to take on a slightly different sort of tilt and there will be these apps that are sort of just like these steroided up apps, if you don't want to call them super apps. And they control a lot of different use cases. And again we go back to the notion of Claude and Claude code and now we obviously are starting to see these things converge into that. Claude code was obviously setting up the idea of computer usage and at what point, you know, you see people already complaining about it. It's like, why are these two separate tabs now in, in Claude itself? Right? Shouldn't they just be one tab? And I think that you're going to see that more and more and we'll see again what, what open AI ends up coming out with, with when they do release the, the quote unquote super app for chat gbt if it's all just like one box or if it's now like gonna be these like multiple stages where you have to pick and it's like, it's sort of an extension of what, what we've had in the earlier days of, of the chatbots where it's like you have to pick a drop down model, you know, model from the drop down and you get super granular and it's like a user shouldn't worry about that. And as we get beyond power users, they're not going to want to worry about that because they're going to worry like especially with computer usage and stuff. It's like, am I, am I selecting the right thing? I don't want it to have access to this, but I don't know which, which, which service I'm supposed to select for it to be able to do that. So yeah, it's sort of opening up a can of worms.
- [17:34]
A
Yeah. I almost feel like once people get comfortable with the AI taking control of their computer until it makes a fatal mistake, they're going to stop checking. I'm almost there. I'm almost there. I don't, I basically. And you know, this will obviously someone will play this when, you know, my computer self destructs or someone shares all my emails, when an agent shares all my emails on, on Twitter. Or something like that. But I'm always allow. I've given access, I spoke about this with Ranjan on Friday. I've given access to my Gmail, plugged it in with ChatGPT after ChatGPT asked very politely. And I'm at the point now where, you know, there's always been this discussion of will computing sort of happen through the messaging layer? And it's been this dream of Facebook and others for a long time that messaging is the new platform. And it sort of was again, the title of my head of my story was the agents and the chatbots are going to merge. And it always felt, and this is why I feel like it's important to gut check this. It always felt to me like it was a nice fantasy of Silicon Valley but was just too hard to execute, was never going to happen. We've already seen the cases of people trying to do like chatbot commerce and that people still want to go to the website. But I will say that it feels closer to me now than perhaps ever before. Like if you, the more you engage with these bots and the more you engage with these agents, it becomes really possible to see how that experience merges and how this does become, I guess, the new operating system or the new interface through which you access, you know, not all computing, but maybe almost all of it. Do you, what do you, do you think? I'm, I'm like getting ahead of myself here.
- [19:23]
B
So my angle on this has long been that I think that there that we're on this steady march now and I think that it seems like OpenAI has been hinting at it for a while that they're working behind the scenes and they've rolled out some things but it seems like they've been setting the stage for at least some of that use case to make the leap to full on voice. Right. And so they're working on these new models, you know, specifically geared around voice. They've been improving voice mode, which I do think is interesting because it's a potential differentiator against Claude. Right now Claude has a voice mode but it's not like their main focus as it has been, you know, dating back to I think the 4O models for, for OpenAI. And so I feel like part of that is probably driven by the work that they're doing on whatever device they're, they're working on, you know, with, with the Johnny I've team and obviously that's going to be voice dependent and what it seems like, and we'll talk about this in A little bit. What Apple's working on is also voice dependent. And okay, like we've talked about this a little bit, but it's important to note like I'm not saying that I think voice is going to be the only interface for computing going forward, but I do think it's going to be a pretty key aspect for a lot of different use cases. I think there will always be the time when you're surrounded by people and you need to sort of privately do messaging via text to be able to do some of this stuff. But in my personal usage now I find a lot of it going through voice and that includes with Gemini which I've been using more of post IO just to try it out. I like their new interface for the new iOS model app at least and I think that their voice mode is pretty good and it stacks up there with ChatGPT right now. But again that's where I find myself naturally gravitating towards and I have not been comfortable enough to do it in these like agentic use cases that you're talking about. And that will be the key to me of when they jump into that space where I can just say to whichever, you know, whatever model I choose, like go do the could search my email and send a response to so and so that I need. You know, I'm going to be busy at 4pm or whatnot when I'm comfortable enough doing that with my voice without needing the prompting and sort of the visual cues. I think that that will be a key sort of next step in my own sort of personal usage of it. But I do think that that's a one, one key element of it to your question. And then just one other thing that while you were talking I remembered from the IO part is that you know, Demis Hasibus came out specifically to talk about their, their breakthrough and what they view as a breakthrough in their Omni models. And this again goes back to sort of 4.0 and what, what chat GPT was, was long ago trying to do with, with beyond just text, right? And doing, doing visuals and, and video and everything else. And now obviously they've pulled back from video, but Google is not that. And what they showed off at IO also seemed a little lackluster with regard to, you know, the Omni model. But it was also clear that it wasn't everything that they wanted to show. But that's all that sort of they had I guess ready to roll right now and that clearly Demis views this as like the quote unquote world model for them and that they're, they're working hard on this sort of behind the scenes, but they don't have enough sort of to showcase yet. But that's sort of the next step of the. Yeah, where this goes, it's like text, voice and then omni, I guess for everything else.
- [22:48]
A
Right. And so whether it's chat or whether it's voice, you know, to see, to see more and more computing happening within like the, let's call it ChatGPT interface, no matter how, which format or form you use to access it. You know, I think what you said is actually really important. You have to get to the point where you trust it.
- [23:09]
B
Right.
- [23:09]
A
And I don't think these things are extremely trustable yet. I think maybe I'm just kind of a psycho that likes to press these products to the limit for the service of the audience, of course. But you know, you could see so much of computing happening there. I mean, you think about it and it's like the chatbot is going to be the place where you're going to want to direct most of your interactions with the web, with the Internet, with information, online with computing. For instance, like when you connect it with your Gmail, instead of searching Gmail for like your flight data, you just go into ChatGPT, give me my flight information, my confirmation number and it spits it right out.
- [23:46]
B
Yeah, I have. So I. My one use case that I use right now all on a daily basis is basically I use Claude to check in on my Gmail. So I don't go to the Gmail interface anymore, but I basically just say give me all this. The situations where I've written about X Topic before and Claude provides a much better interface than Gmail for that. Right. Because it'll give you like a natural language spit out of like oh yeah, in this, in this newsletter on X date you wrote, you know this about X Topic and here's a link to that and versus obviously Gmail right now is full on just, you know, old school search paradigm. Now they have Gemini baked into Gmail, but it's not nearly as seamless and as good as it is on, on these other services that are much more predicated around the agentic use case. So again that's on Google being behind in that, which is weird because they own Gmail. Like you would think that they would have like the single best place, they should have the single best place to do any sort of agentic email workflow, whether it's sending emails or searching emails or doing anything with Email, they should own that. And they don't own that right now. They just are own the layer, you know, behind it that all these other services are latching onto. I don't know a single person who uses sort of AI and agentic workflows that doesn't have it connected to their Gmail. That's like the first thing that they do. So it's wild that it's not Google. That's like owning that right now.
- [25:13]
A
Yeah, it's amazing how searching Gmail through ChatGPT is better than searching Gmail with the Gemini baked in within the Gmail window. It doesn't make any sense. And I think that like, as you go through different experiences, like you will see that this is sort of just like talking about what the bull case might be. In many cases it will be a better experience to use ChatGPT to do things on other websites. For instance, like think about booking.com. booking.com has a chatbot in booking.com if you're looking to book a hotel, you can ask it for things, it has all the booking data, but you can ask ChatGPT similar questions about hotels. And it can scan everything, right? So you're going to want to use that. And then of course, like, are you. Even if the booking API is unavailable to ChatGPT because booking wants to keep their own experience theirs, you can just tell, let's say codex go and openbooking.com, log in for me and book the one that you just suggested. And so this is my core question. I mean, of course there's a question for Google, right? What happens to Google? But the bigger question is, does this mean that power, like the power of, how do you put it, like the power of the web or the power of controlling the interface to the web will reside with the OpenAI's and the anthropics if Google doesn't get to that and then what is, what are the implications of that?
- [26:41]
B
So super interesting. I actually was thinking about this earlier today in a totally different context, but it's analogous I think, which is that one of the still very annoying things about trying to use, especially in the current age where cable has been totally taken over by streaming and now we all have five to ten different streaming services and we were all, and we've all been longing for, I think this, this like UI to unify all of them together, right? And Apple has some of it, Amazon has some of it, but no one will can work with it fully because Netflix, the biggest one, refuses to play ball. And Give their, you know, content out to others to sort of be able to surface all of this. And why do they do that? Because they want to control the interface, of course, and they want to control that relationship. And so when you think about that in the context of what you're talking about with the these, with these, you know, AI tools like, you can easily see a world in which a lot of different players think it was like basically a mistake to allow, you know, Google search to become the interface by which, you know, that's the main way you found your way to a lot of sites. Obviously many would already say that. Right, because if you fast forward to where we are now, they're losing search traffic and stuff because of the changes that are happening and in the entertainment world. Right. A lot of people went on to say like, that they, they regretted giving over their interface to itunes. Right. And because again, they were, they were giving up, yielding control to Apple, even though app, you could argue Apple, like saved them from themselves by, you know, what they did with itunes. But still you, you understand the argument of why these companies sort of make that decision. And I worry that we're already in this world certainly with Agentic shopping that you're seeing with Amazon, right? Like not, not, not being able to fully play ball with, with certain other, you know, shopping agents and, and vice versa, others not wanting to play ball necessarily. You know, so Amazon owns the relationship, which obviously a lot of people also regret that, you know, in our, in our web day and age of allowing just everything to be outsourced to Amazon. And so, yeah, I worry that we're gonna not get a seamless move into like, like a one single place where we can do all of this. Like the sort of the pie in the sky version that you're talking about, where there's one tool, but instead we have to use like three or four different AI tools in order to get full coverage of whatever task we want to do, because certain ones just refuse to play ball. And certainly as, as, you know, as the big players, the big tech players are all, you know, doing this right now, we're not going to be able, I do really believe, like, we're going to be able to do Amazon purchasing through Gemini, you know, as seamless as you could through, through Amazon's, you know, own tools and Alexa. It's certainly doesn't seem like it anytime soon. And so that's a world that I think is going to be potentially painful for, you know, this transition.
- [29:31]
A
Right. Okay, so here's where I think it could Be different, which is like, and you can sort of gut check me on this one if I'm, if I'm you know, going too far here. The agentic tools may not need the permission of the sites that they're working with because what they're doing is instead of having a plugin, they are not relying on, let's say the data plugin from booking. They're taking over your browser, they're taking over your computer and they're acting as you going into those sites. And they'll be able to move faster, they'll be able to browse faster. And like, maybe there's some blocking that these companies can try to do on their end to like, sort of like not, not let non human traffic crawl.
- [30:14]
B
Right.
- [30:15]
A
That's very difficult. So I think that they might be able to get around these, these walls.
- [30:20]
B
Yeah. And I mean that battle is sort of playing out right now, right? Like there's certain, certain sites and services that use the different, you know, MCP and all the different sort of layers that you can, that are effectively API layers to be able to call up up certain things within, within the services themselves. And then as you're noting, there is the sort of fallback plan to just have computer usage and yeah, just use the web. And so it would really take shutting down the web or cutting off crawlers, I guess. You know, maybe they could do it in some ways like, so that, yeah, say like Amazon started to say that Gemini's crawler can't do it. That's complicated as we've seen already play out because Google's crawler is also tied to Google Search. And so if you're turning off the crawler for Gemini, you're potentially turning off Google search, which basically no one is willing to do at this point. And gives Google, you know, a huge leverage point in, in any such negotiation there. And Amazon is a similar thing probably with, you know, just with Amazon shopping, like are you gonna cut off access to that? If, if you say like I don't want Alexa to have access to this, therefore, you know, like Amazon might say, okay, then you can't be a part of Amazon shopping. And that's going to be, I'm sure litigated, you know, to know when those types of deals and arrangements. There's one other thing that this brings to mind which we haven't talked about or hasn't been in the news a lot recently, but was starting to play out this way specifically from Amazon because they struck a deal with New York Times to do content right and then New York Times, famously, I think in almost every article has to say that they're being sued. Sorry, they're suing OpenAI for content infringement and whatnot. And so we have this world in the content side, which you see oddly in, in different places. The other day I was trying out Atlas again, which is OpenAI's browser. And if you go to a New York Times story in Atlas and you say within The Atlas Native ChatGPT, that's baked into it. If you say like summarize this page, it refuses to do it it because they don't have an agreement on the content. And that's wild to me because again, it's the web, like I can see it with my own two eyes. And I'm telling the agent that I have access to to do it, but the agent doesn't have access to that content, so they won't do it. Now why can't they just scrape my screen and see what I'm looking at to do it? And I can get around it by basically copying and pasting everything that's in that article that I want to read and putting it into the chatbot. But these are like the silly things and these silly edge cases that I feel that I fear we're going to increasingly run into. So I hear your points about that. Yes, there will be fallback to the web, but I already see, feel like there's, there's these ways that we're seeing that break at the edges.
- [33:04]
A
Right, right. Yeah, it'll be a very, I mean, there's going to be some, some serious fights here. So. Okay, let me ask two questions, then we'll move on. Yeah, two big picture questions. We've talked about these, you know, anthropic and OpenAI valuations and you know, their addressable market largely in terms of coding and enterprise. Right. That's sort of been the thing that people think is going to take them, you know, to the promised land. Not necessarily consumer. If this works though, you know, maybe because again, because they would ingest so much of the experience online, those have to be massive, massive businesses. And so maybe this sort of super app style way to access the web, if it works, becomes the core business over time. Like it's almost the analog to the SAS apocalypse just in consumer.
- [33:57]
B
Yeah, and again, when I hear you talk about that, my mind naturally just goes to, I think that they need to do, they need to get that part in order because if and when they try to do their own devices, like, and if those devices aren't screen Predicated how you're going to be able to do that is mostly through voice and, or I guess glasses in some capacity. But they need to basically have everything sort of teed up to perfection in order to do these workflows. Because the second that any of that stuff starts to fail or you have to say like, oh, just. You're gonna have to wait till you get back to your computer. Like, it just feels like it's going to be too big of a hassle for a consumer for an end user to do it, and they're just going to stop sort of doing those types of use cases. And I think this is just like, you know, basically what we saw play out with the first instantiations of Alexa and HomePod to some degrees and, and Google Home and everything else where Google Assistant, I guess, where, because you couldn't do everything that you wanted to do, you know, they basically just fell back to these really rudimentary use cases. And the fear with these new devices coming out would be like, like if they're just like really expensive newfangled music players or weather weathermen and, you know, these other sort of very rudimentary simple use cases versus all this agentic stuff that we're talking about, which is obviously what they aspire for them to be. You know, you, you're out and out on a walk and you just say like, yeah, send, send an email to Rick and tell him that I'm going to be 10 minutes late. And it knows to follow up on the email that you had going back and forth to set up the appointment and whatnot. And if you can't trust that agent to do that without a screen, I feel like, yeah, that's where this all breaks down.
- [35:38]
A
Yeah. No, I love the way that you're thinking about it. We have Greg Brockman Round three coming up at the big Technology AI Summit later this month. And I'm just kind of, I don't know if I'm gearing up to be like, where, where do we want to take this conversation? I think it has to go in this area especially, you know, what the plan is. I love what you're saying that if you can't get everybody to participate, you know, do you end up have. You need a network effect basically for this stuff to work?
- [36:03]
B
Yeah.
- [36:03]
A
Otherwise, you know, it's just not. You're right, it might be just another fancy, fancy alarm clock. All right, here's, here's my second big picture question. Going back to the Google example. We started with Google. Let's end the Segment with Google.
- [36:18]
B
Yeah.
- [36:18]
A
Don't they need to get, they need to do this right? Don't they need to get involved here? Like, to me, sitting this out and letting the other companies try their hand at this experience and not putting all the effort into it just is like a recipe for disaster. Or am I being too inflammatory?
- [36:35]
B
No, I mean, I think that they do and I think that they probably recognize that. Again, I view this, this IO was super awkward because everything as we kicked off, talking about was just like, like coming this summer or coming later this year, coming in a few weeks, coming to ultra users, coming to. And it's like nothing is just like ready to go. And I feel like the whole reason they did the timing of it is because that historically this was when they've done it. And you know, now in the age of AI, they didn't have the time, they didn't have the timing lined up for all these different things that they want to do and that they want to talk about. And yeah, to your big question, like, shouldn't they be doing exactly sort of what, what everyone else is going for? I think that they are. I think that they know that they have to do that. I just don't. I think that because of the way that they have sort of, yeah. Aligned around doing, getting the search stuff out of the way, they sort of focus on their strengths and figured out like, we're the only ones who do this and this is our core business, so we have to get this nailed down. And then I do think that the Demis has, his team is probably behind the scenes working on all these crazy models and crazy health and life science ideas that they want to work on. And so it's not to say that those things are a distraction, but it's just a matter of focus. It's what you want to focus on. And, and Google famously, you know, goes through these cycles where they're doing too much and then they have to pull back. Right. And, and fewer, fewer wood behind more, fewer arrows and all that, that sort of stuff. And so I feel like they know that they need to do exactly what you're talking about. And I do think that they will get there at some point, but it's a matter of like. Like, is there. Yeah. Is there like a first mover advantage, much like there has been for quad code, right, where it's like all of a sudden because even though by all accounts like Codex is good, but it's just not like cloud code was there and just hoovering up all of these users and so like if one of these players gets to that quote unquote super app stage and yeah, hoovers up all the users, like is anyone else going to be willing to switch? Especially with memory and everything else, you know that that's integrated into this?
- [38:36]
A
Yeah. Now I'm not 100% on this, but this is one of the reasons why I think this is difficult for Google is that. So Google famously, when they built this great Gemini model, they sort of centralized it in something they called the engine room, right. And they put some of their best people on it and what they did was they then farmed it out in a way or they worked, they collaborated shall we say, with the many product areas within the company to integrate it. Right. So Google has these famously sort of walled off product areas. Not walled off but like, you know, they're, they're fiefdoms in a way, not in like the same nature as Apple, but they have their, you know, politics.
- [39:14]
B
Sure.
- [39:14]
A
And they built this engine room with AI and they sort of work together this, with the centralized group, work together with these product areas to integrate the AI in all the products. So that was, you know, sort of push from the core, pushing out. The tough part here I think is if you're going to get this right, you need to almost reverse that, right. You need the product areas working to centralize in a way which will, you know, require some ego subsetting and some prioritization that some might not like. And that is sort of where you might end up hitting some sort of build out wall where you're not able to work as fast as, as an OpenAI and anthropic because they just don't have that legacy stuff to work through.
- [39:58]
B
I think that's a very good point. I had not thought of it from that perspective, but you're right, it's basically Google's immense size and capabilities become a hindrance in, in the building out of this. Right? Because like, imagine that they feel like they have to answer for OpenAI's Super App, Chat GPT Super App coming out. And it has, you know, from what's being been reported, it has chat GBT baked in, it has Codex built in and it is Atlas built in. So if Google feels like they have to answer that with something with Gemini Chrome and Anti Gravity and those are three totally disparate huge teams working on these things. Certainly Chrome, which has not been, you know, historically like LinkedIn because LinkedIn to the AI group, I know again we talked about it earlier, they had, have they have Gemini baked into it, but it, even now it sort of feels tacked on. It feels like it's not, you know, like it's not an AI native product and if they have to sort of make it an AI native product in order to meet the moment with regard to these, the super app question, like can they really do that? And it's going to be, it's, it's way harder than it will be for any of those much smaller, more nimble teams to do it or who are more. Yeah. Oriented in the way that sort of Apple is or some other companies. I do think that it's going to be very, very hard for Google to do that. And that's the, that's a good reason for hope for. Yeah. How you compete because again like when you take a step back and you, you're open AI and you're looking at that landscape and you see Google and Google is a 4 trillion plus dollar plus company and they have hundreds of thousands of employees and they have all of the cloud engineering, all the cloud capacity and CPUs and everything that they need. Like you look at that and say like, how can I possibly compete with this? This, I'm a startup, I've raised more money than anyone's ever raised in history. But how am I going to compete with a $4 trillion company that already has all these inherent advantages? And that's how you do it. You use their strengths against them as a weakness. Whereas like they can't integrate as well and as quickly as you can because they have all this legacy stuff. And that's how, you know, Microsoft's of the world get disrupted and yada yada.
- [42:02]
A
Just to go back to the Gmail example, if you are, let's say you're a super app, right. And so with ChatGPT, Codex and, and Atlas together and let's say you had, let's say OpenAI had Gmail. Let's say Gmail was one of their products. For this to work the best, you want Gmail to almost fall by the wayside, to disappear into the background and for these messages to be surfaced in ChatGPT. And if you are, let's say you are a, you know, responsible for the Gmail product area, that is for you. Yeah. You want me to disappear myself for the greater good? You know, it's very difficult to. In a corporation that is tough.
- [42:48]
B
And I think you're exactly right. Because when I think of like the future of email, you know, to use your example exactly, it's like I believe that the success state of the future of email in AI is that we're not going to have to use it all the time obviously. And instead it becomes like our agents basically emailing each other and us getting these like to do list app. Right. Things like where it's like yeah, MG and Alex emailed their agents, emailed one another and we're setting up a time to coordinate, you know, the next, the next podcast and you know, do you want X, Y or Z time? And just hit a button and, and select that and then it does the emails behind the scenes and you and I never opened Gmail again. And again to your point, the PM for Gmail, like what are they doing in that world? Yeah, that doesn't seem to do great. And so you know, and they're rolling out this new. One of the things that's not, that's supposedly coming but again not quite ready is this you know, agentic version of Gmail where it's like a. From the screenshots that they've shared and whatnot, it looks like, you know, yeah, the high level like to dos and things like that. And so they're trying to kind of for that world that we're talking about. But again, the real ultimate answer is again not to open Gmail itself. And there's a tension there for sure.
- [44:05]
A
That's right. Yeah. When you were saying they're trying to do these to dos, it's like I'm thinking, but it just doesn't belong there. It belongs in Gemini. All right, let's take a break and come back and talk about the iPhone fold and what's going on at Meta right after this. This summer always changes how I get dressed. I want pieces that feel lighter and more comfortable but still put together. That's where Kintz comes in. They focus on well made essentials in breathable linen and soft organic cotton. The kind of basics you will keep coming back to. Everything at Kintz is priced 50 to 80% less than similar brands because they work directly with ethical factories, cutting out the middlemen. So you're paying for quality, not brand markup. Personally, while I'm moving between air conditioning and summer heat, a lightweight zip down sweatshirt from Kintz has been a life changer. Elevate your summer wardrobe. Go to kintz.com bigtech for free shipping on your order and 365 day returns now available in Canada too. That's q u I n c-e.com bigtime tech for free shipping and 365 day returns. Kins.com bigtech most leaders know how work is supposed to happen. But when it comes to how it actually gets done day to day across tools, teams and handoffs, they're mostly guessing. That's exactly the problem Scribe Optimize was built to solve. Trusted by over 80,000 enterprises, including nearly half of the Fortune 500, it gives leaders a live view into how work is really happening across approved business apps without interviews, manual process mapping, or extra effort from the team. And because it's continuously analyzing real workflow activity, the insights stay current instead of going stale the moment a process changes. You can see which workflows are happening, where time is going, and which tools are involved. It automatically surfaces top issues, explains why they're happening, and even recommends ways to fix them with estimated time savings. And importantly, it's built built with privacy in mind. So activities only captured in admin approved business apps and user level data is anonymized by default. The kind of visibility that used to take months. Now it's just always on. If you're ready to stop guessing and start seeing, visit scribe how BigTech that's S C R I B E how slash BigTech and we're back here on Big Technology Podcast with MG Siegler of Spyglass. You can go to spyglass.org sign up for the newsletter. Highly recommend it. One of my favorite reads in the tech world, and I'm sure we're going to all be glued to Spyglass for the WWDC coming up this month. You it's kind of funny the way that you're, you know, obviously I wonder the big news will be Siri they may. They may preview the iPhone fold. They may preview the glasses. It's sort of up in the air. We're going to have a new CEO that's incoming, right? That we're going to probably hear from a lot or a little. Maybe it's Tim Cook's last hurrah. Maybe it's the Turner show. So many variables around wwdc. But anyway, let's talk about the iPhone fold, because that to me is the thing that I'm most excited for. You call it the iPhone's BlackBerry moment in that it's kind of for folks, if you're thinking about this and if you're on audio. The way that we need to describe this fold is it's not a phone that sort of folds out from a vertical axis, right? It's a phone that folds horizontally, kind of like an old flip phone. And that might make it easier to type because you won't have to hold it. So you know, so uniquely to make sure it doesn't fall over, which is what you pointed out in your story. MG if that's the way that it looks like, I'll be honest, I'm a little bit stunned. The folding out on the vertical axis to me is always the benefit here. So you can watch things with a bigger screen, but Apple doesn't seem to be going that way. Talk through the fold.
- [47:58]
B
So there's a few things in there. First, I wouldn't be shocked if it's still sort of marketed that it folds out horizontally, not. Not vertically. So it's not like the. A flip phone. It's more like a fold. The weirdness that you're alluding to is basically because it's so short relative to the current, like foldable phones. So I have a Pixel fold, which is, yeah, just when it's folded together, it looks like sort of the regular size of a regular, you know, smartphone that you would normally see. Like the iPhone looks right now, this new fold, by all the mocks that we've seen coming out of Asia and, and everything else and all the leaks, it seems like it will be this short and squat, you know, little thing that looks like it should flip open like an old razor and whatnot. And obviously there is a razor that does that now and I think Samsung makes a version that does that. But I do think Apple will probably still align it around folding out more book style, even though it's short and fat, even though it might be awkward. But that's like, that's the biggest thing about this, right? It seems like they have a good moment to sort of try to own a new, these new sort of use cases for this new type of device that's not simply making a big, you know, a smaller screen and making it a bigger screen if there's actual, like, ways to use the device when it's folded. And this is what I talk about in my piece where it feels like it's actually a great typing device because it's a little bit fatter than what the current phone is, which is maybe a little bit too thin for people who are used to sort of thumb typing, you know. And again, I allude back to the old BlackBerry days where those were a little bit fatter, for lack of a better phrase, a little bit wider. And so they were a little bit more conducive to that thumb typing, which obviously became a thing and crackberry thumb and all that sort of stuff. And so like, there's a world in which we maybe go back to that because this device is a little bit better for typing when it's folded up like that and then you unfold it to either do like you're talking about, watch videos and, and do other sort of gaming and other content that way. And maybe it's for quote unquote, real work if you want to have like two apps side by side and things of that nature. But it's a, it's a very big question of like how they get this working. I think some of the reports are, you know, talking about that when it's unfolded it will look a little bit more like maybe an iPad mini interface, but it won't be exactly iPad OS but it will be a souped up version of iOS in terms of yeah what it, what it can showcase when it's like that. And I would imagine that sort of like widgets and other things that are, that are more than just the typical, you know, laid out apps on a screen because that would be sort of boring. And then that just brings me back to the notion of like I think if Apple is going to do this and obviously they're doing it, I think that they would only do it or at least frame it, that they would only do it if they can really do something different with it. That's not just another form fact, the same iPhone, but another form factor of it. I think that they want it to be sort of a newer type of device, a different style of iPhone where it brings different use cases with it. Now if they can pull that off, who knows? I mean again, no one's actually seen this. We've seen dummy units of it. We've seen what it looks like. I do not think we're going to see it at wwdc. I think that they will wait for the fall to do it. I think that yeah, wwdc they'll have their hands full with making sure Sirius is right this time.
- [51:22]
A
So, okay, so that, that's sort of getting to wwdc. You know, we neither of us expect the preview of the fold. I think there's just like that's the, the product with the most excitement in the Apple lineup. But we might see a Siri rollout. We might see. Do you think we'll see a preview of AirPods with cameras? Do you think we'll see a preview of glasses? I mean, is this going to be so. So there's a lot of pressure for them to deliver this updated Siri this time. But part of it also feels like there is a chance that it can just be like another transitional year of WWDC where like, like maybe Siri is not completely ready. There have been delays on that front. There's going to be the CEO transfer. Is it just like a celebration of, you know, of, of the, you know, Apple's history. Right. And legacy. Yeah. And not much on the product front.
- [52:13]
B
My guess would be that it's more that you would. That talking about it now you would feel underwhelmed. I think that it will be pretty straightforward. I think that they'll, you know, show off the new Siri. I. It's curious how they'll talk about it if, you know, a lot of reports suggested that they won't want to play up the Google element and the Gemini element too much, but you would presume that they sort of want to play it up so that people trust that, you know, they're going to do it right this time. So how they brand that will be interesting. I would be shocked if they show off any sort of newfangled hardware. I think that there's a chance that they show off maybe like a new home pod that's capable of doing it. There's reports that those have been in the works for a while and been ready for a while. They've just been waiting on AI, you know, the AI to sort of catch up. And so maybe there's ways that they do that, but I think that they'll probably show off the new use cases just on an iPhone and show off things like visual intelligence, which they've had for some time, but sort of a version that's actually working well this time. And showing off. Imagine that Gen Moji that looks better this time and whatever and image playgrounds that looked better this time. And, and again, the, the famous getting mom home from the airport. Like, will they actually be able to get her home this time and pull it off?
- [53:25]
A
Oh my God. They're gonna get trashed if they, if they come out and they give you Genmoji and getting mom back from the airport again two years later.
- [53:34]
B
I just, I don't. I have a hard time. They're not going to show off anything that's like, they're not going to show off the new glasses. They're not going to show off the fold. The AirPods, like, are maybe the closest, you know, thing that we have to an AI device that's likely ready. More ready to roll. I, I'd just be shocked if they do that. I think that they're going to keep it pretty straightforward and say like, hey, this is sort of our do over AI event. And this time we got it right. And therefore you should be happy like to use all these things and get excited and maybe they'll say, maybe they'll throw out a bone and say like, hey, it's going to be ready sooner than you're expecting. Like, like you don't necessarily have to wait for the fall. We'll have it in beta in the late summer so all of you guys can test it out. Because by a lot of accounts, like Garmin in particular, this version of iOS is not going to be totally different, you know, not, not a big sweeping change, right. It'll largely be predicated around Siri again. And so maybe it'll be ready to roll a little bit early and maybe that's a surprise or something like that.
- [54:33]
A
I mean, like what the. The pitch is. Siri, that almost works now. More now, more almost than ever.
- [54:42]
B
I think they'll have something a little more, a little more elegant than that. But yeah, I mean, I wouldn't be shocked if you and I are catching up in a month and sort of we're talking about like similarly to Google's. Right? Because again, it's sort of the same idea of our. Do these big events, showcase events make sense in the age of AI, Right? Because AI just rolls at a different cadence than what product announcements do. And you know, to, to go to Apple's credit, WWDC was always meant to be a developer conference. It sort of is morphed into this more consumer facing thing over time. But it's not the iPhone event. It's not meant to be fully, you know, consumer oriented. It's going to be about APIs, it's going to be about all these different things. I mean, the big thing here's a big wild card. What if they did something, something like this? What if with the transition that you're talking about is we're moving away from the Cook era and into the Turnus era and he wants to set the stage and Cook wants to set him up. Well, what if they make a big announcement around the App Store and changing the fee structure like that would bring down the house, right? Like if they were able to actually finally change, you know, tear down those walls, as it were, and change the, you know, the cuts. Change it from say, you know, 70, 30 to 80, 20 or you know, 85, 15 even and you know, make that sort of get ahead of some of these legal battles that they're in right now. Right. It's not like fully magnanimous that they would be doing this, there's other reasons why they would do it, but I feel like something like that could be perceived, at least make the perception of such a conference good, even if they don't have to have, you know, they don't have that much to talk about from an AI perspective, let alone a new device perspective.
- [56:22]
A
Perspective, yeah. I wonder what the stock market would do because that would take a hit directly from the services business.
- [56:30]
B
But again, think back to when Satya Nadella took over at Microsoft, right? One of the first things he did introduced I Office for iPad. Now there was talk that that had been the worst royal under Ballmer, but it was Nadella's moment to make it, put it out there and sort of set the 10. The tenor of how his administration, for lack of a better phrase, was, would go from there. And so what if, you know, Cook handing off to Turnus, setting a new stage for the next 10, 15 years of Apple.
- [56:58]
A
Yeah, that'll be, we'll have so much to talk about next month because we can recap it, but clearly, like a lot a big moment. It will be a transitional wwdc. The question is how big the transition will be. Of course, Cook to Turnus, but also, you know, home built Siri to Gemini Siri. It'll be really interesting. All right, before we go, let's talk a little bit about what's going on at Meta where, like the word mess has been associated with many things that they're doing from branding, which you pointed out. Let's just read it, folks. If you want to, you know, do a meta subscription, you can subscribe to WhatsApp+ for 2.99, Facebook+ for 3.99, Instagram+ for 3.99, Meta1+ for 7.99, MetaOne Essential for 14.99, MetaOne Premium for 1999 and MetaOne Advanced for 49.99. That's marketing in particular. You call this out. And Spyglass mg I mean, it does. This is where I wanted to take our discussion about Meta. You wrote we're not quite in Microsoft Ian territory, but we're close. Maybe Meta should just stick with ads and between sort of the uninspired product direction and now subscriptions, which you say, okay, this is what you do when you're late stage and you need to like the profit out of your users and these layoffs where, you know, people have, you know, within the company, the morale is terrible. State of Meta looks kind of rough right now.
- [58:31]
B
It does. And you know, also not helped by the fact that every Time there's some sort of negative story. Certain, certainly culturally, you know, the negative report, they are so adamant about pushing back against it that it just feels disingenuous. Right. Like, obviously things are not, it's a big company. Things are not always going to be hunky dory. There's going to be, you know, turf disputes like we talked about earlier, between big factions of companies that are sort of moving in different directions and maybe have different priorities. And so not to acknowledge that is, is and pretend like it's, everything's going fine, it's sort of a little bit gaslighty. It just it that doesn't help their whole, I think, vibe going on right now. And look, they release the first, you know, new models, early versions of them. They're. They're first to admit that they're not quite at the frontier yet. And so sort of like Google, they're gonna promise those down the road. Right. And you know, do they actually. How long does it take them to get there? We'll see. And do they get there with coding? We'll see. It seems like they're behind there as well to our earlier discussion. But yeah, like this whole pricing thing, you, you brought up the, just the, the various different levels of the, of the tiers, but the fact that you named all of those and all those, almost all those things do different things. It's not like they're all, they're all just for a different AI tiers. They're all just for different product tiers. It's so confusing like, in terms of like what each tier offers you that I, I don't really understand why they would roll it out that way other than to your, you know, earlier point. Like, I do think that they are under some pressure to certainly from the highest level to move towards a, a sort of a more sustainable model for their business, which is not right now. I think it's still 98 predicated around ads. And if you believe that we're sort of at like at the top and everyone famously has always predicted, like, you know, we're at the top of, of Facebook and we're at the top of all these services and then they just keep growing, growing. But the reality is like, they're not really growing anymore in Western markets. Maybe they're growing in some, you know, some pockets of different parts of the world, but they're not growing in the place that, you know, that they feel like they can monetize as effectively with the ad product that they've been doing. And so that plus all the spend, famous capex spend that they're doing, all the spend that they obviously spent on the metaverse, which may or may not end up, you know, coming, coming to anything. It feels like that they're, they're probably under some pressure to figure out these new models. Zuckerberg, even this week or this past week was in the news again saying, like, basically acknowledging that with all the CapEx spend, if they need to, they can roll out a cloud, right? They can roll out a cloud that other businesses can use. And, and I think that's a direct result of what we saw with, with elon Musk and SpaceX of leading up to the SpaceX IPO. It's all of a sudden, oh, they're a NEO cloud. That's great. Like, don't worry about this, the how much money Xai is losing because it's just a NEO cloud now for Anthropic and, and for Cursor and this is great. Like these are great growth businesses now. And so can you imagine Meta becomes like a NEO cloud business based off of all the capex spend that they're doing. And like we, I could joke about that, but I don't, I wouldn't put it fully past them if they need to and say a year from now, you know, suggest to Wall street, like, look, we're being a little bit more prudent with how we're trying to do the spend and, and we want to bring money in beyond just the advertising layer, which again AI is helpful with, but it's not the same return that the other peer group is seeing with their capex spend.
- [62:00]
A
Okay, one more thing before we go. Do you have like another minute or two?
- [62:03]
B
Yeah, yeah, keep going.
- [62:04]
A
Okay. Love it when news breaks. While we're recording news just broken, Thropic has filed to go public.
- [62:11]
B
Really? Wow.
- [62:12]
A
Today, Anthropic confidentially submitted a draft registration Statement on Form S1 to the U.S. securities and Exchange Commission for a proposed initial public offering of our common stock. So they have now it looks like they've jumped the gun on OpenAI, I
- [62:28]
B
was going to say. So there was the report last week, right? That, that and it was right ahead of SpaceX filing their S1. It seemed very not coincidental that OpenAI let it perhaps leak out that they would be filing soon in one of the reports that suggested that it would be last Friday. As far as we know, that didn't happen and, or at least it never leaked out. Obviously they would file confidentially too. So either they filed and no one found out about it or oh, Anthropic just pulled a quick one on them and got out ahead of them, which would be wild because we've talked about this before. If OpenAI goes out, if anthropic goes out before OpenAI does. OpenAI is in a lot of trouble from public listing perspective, just because we even talked about this when it was still just the notion that they would have a better bottom line. Right? Like that they wouldn't be burning money for as long as OpenAI would and they would get to profitability quicker. Now the reports are that Anthropic has already turned in some in at least at some level, perhaps dipped into profitability because of just the crazy growth that they're seeing. And if they're now growing faster at the top line and have a higher valuation, like this is going to be a narrative nightmare for OpenAI to try to also go public in.
- [63:52]
A
This almost reminds me of the book publishing games where if you have a,
- [63:57]
B
that's ahead of a.
- [63:58]
A
Somewhat similar to someone else's, you might announce your publication date to sort of clear the deck. But they also might, you know, keep theirs secret until they're ready. And I wouldn't be totally stunned if OpenAI has also filed and now lets it out in silence.
- [64:16]
B
And now they would have to, of course, put it out there. Don't worry, we filed on Friday. We're ahead of them by two days or whatever. The silly thing about this is, of course, like by filing, you know, this is filing confidentially, you don't necessarily have to go out. You're just like, you're basically putting the wheels in motion so that if you want to, you can have that optionality. And when I saw that leak last week about OpenAI doing that, I thought it was both for optics because obviously they, they would take any, any chance they can get against going after Elon, their, their foil, you know, to try to take some, steal some momentum or just some buzz out of that out of those sales. But it also gives them the optionality to go out if they feel like they needed to against Anthropic. That doesn't necessarily means that they would. And I feel like Anthropic doing it now though, puts so much pressure on them to actually have to do it because now they're going to just be afraid that if they don't go out, Anthropic could go out at any moment and then like, again, that's a big, big problem for them.
- [65:13]
A
Yeah. Talk briefly again and then we'll head out. Why is it important for OpenAI to go before Anthropic.
- [65:21]
B
So so much of this to your book publishing point is around the narratives and around drumming up up investor interest in, in these stocks. And while SpaceX and OpenAI are interesting because obviously SpaceX has XAI now which is directly analogous in ways to, to OpenAI's business, they're very different businesses. Obviously OpenAI doesn't have a rocket launching arm nor do they have Twitter for, probably for better in that case. And so they don't have a lot of overlap app. And so there's, there's going to be comps that are done like, you know, market comparables done between the two companies on their AI businesses, but they're not direct. Anthropic and OpenAI would be much more viewed as direct, obviously competitors between the two. Now their businesses, as we've been talking about, are different in ways, but as we've also spent a lot of this conversation about, they're converging, right, because they have, because OpenAI feels like they have to go after what, what Anthropic's been able to tap into with developers and yeah, leaning into, you know, coding and, and then eventually agentic work. And so they are converging. And the comps between Anthropic and OpenAI that we're seeing with this, these most recent fundraises just increasingly do not look good for OpenAI. And so when public investors see that, they are going to say, okay, we're going to invest in one of these or the other, which one are we going to invest in in? Previously when you said chatgpt was by far and away the leader in yeah like top line revenue and yeah, they might like be behind in bottom line because they're spending so much on servers and they want to grow into a bigger market opportunity than what Anthropic could do. That's the narrative they would have projected with that. But now the fact that again they're converging businesses and that Anthropic has sort of zoomed ahead on the top line too is a real, real big problem. Now I would assume that OpenAI would try to counter by saying, look, we still have 900 million MAUs, which is also slightly problematic because that number has not shifted in a while and clearly they've wanted it to shift above a billion so that they can announce that. And it seems like the IO point Sundar was on stage announcing that Gemini is also at 900 million MAU, so that's not great. But still they would probably say, look, we have a lot more regular and consumer usage than what Anthropic has. And so that's going to be our narrative. But I don't, I again, when you're, when you're talking to and out there pitching investors and they're going to look for comps, and those two are the, you know, the most direct comps to one another and one of them is just overtaking the other one. And so unless in the next six months, you know, before they were to go out, unless OpenAI has a way to zoom back ahead heavily relying or pushing codecs, you would assume like that's going to be a big problem. Now one other thing I would just add. This is highly controversial, you know, in the past like few months and it's playing out right now as, as SpaceX gears up to go any, any day now, there have been changes to the indice rules where apparently SpaceX, when they, when they list, are going to be included in. And maybe not the S&P 500 right away, but they're also trying to change the rules for inclusion into that so that they no longer have a holding period. And so what that means for both OpenAI and Anthropic potentially is that they're big enough where all of them plus SpaceX would all go into these indices. And that just means that the bunch of different mutual funds and all sorts of big funds automatically have to rebalance and buy into these. And so you can see a world in which maybe the comps and, and all the numbers don't matter so much because these, these basically these, they're going to be included in these indices and these mutual funds are going to have to buy in in major ways regardless. And so maybe that's an argument for going out now, I guess. But again, I'm trying to paint like, yes, the most like, roundabout picture of how, like how you can make an argument for OpenAI going out at the same time as anthropic. Anthropic, maybe OpenAI just waits as a public company for longer, but would they, they need to raise money still. And so. But are they able to do that?
- [69:34]
A
Or maybe they do, right? That's sort of the one thing that sort of remains hanging for me is anthropic just raised $65 billion. They announced it last week, right. OpenAI just raised 122 billion. Is the burn that great that they have to go public to keep funding it? And remember, I don't know, you know, what's left after the public market, right, in terms of fundraising well, they just
- [69:54]
B
have easier access to raise debt, you know, if they're public and they'll be able to use their stock sort of in a more liquid way, you know. And there's other more granular reasons that I'm not well versed enough in to know exactly why they would do it. I would imagine, though, at the highest level, they're doing it because the window's open, right? Like, and if and when it changes, like, they could be stuck if they don't go out when the window is open, open, knowing that they need to continually be fundraising. Now, to your point of do they need to keep fundraising? I would say, like, yeah, Anthropic seems like in a good spot, you know, given that talk of profitability and whatnot. But remember, they're also in a bit of a tricky spot, which OpenAI has played up certainly in that they haven't had the capacity to meet the demand. And so now they're striking these deals with SpaceX, as we talked about, and with other Neo clouds. And those deals are not cheap and those cost billions and billions of dollars. What's the report that they're paying like $15 billion potentially a year to SpaceX to lease the Colossus Data Center. And so, like, yeah, I mean, you can burn a lot of your money that you just raised quickly if you're all of a sudden spending it right back to these Neo clouds to, to meet capacity. And so you know how that plays out. It's June now. Do they go out before the end of the year? I don't. I mean, there's so many variables in between now and then, man, I feel like I'm a broken record on this like that A lot could change in between now and, and the time that they would go out, but SpaceX is going to go public, you know, this week, the end of this week or early next and so, or maybe the 11th, you know, at the latest or whatnot. And so they're going to be out there and they're going to test that, that public appetite and they're also going to test the fences of what we were just talking about with regard to these indices buying in and what that will look like. Like there's a world in which SpaceX is stock value. Market cap just zooms way past 2 trillion, like in an instant because of, of this automatic buying that happens, which is wild.
- [71:47]
A
Definitely. Anthropic made its announcement about the confidential filing to go public two paragraphs on his blog post, did not pre brief any news publications. So talk about a mic.
- [71:59]
B
Your move. Open AI Your move.
- [72:02]
A
MG Great to see you again. Is always, always fun to talk about breaking news live. So thank you again for coming on.
- [72:08]
B
Likewise. And you totally hit me with that blind Alex. I didn't even get a push notification about it, so. Wow. Okay.
- [72:13]
A
Oh yeah, email in my inbox about 10 minutes ago, so. But it's good that we were able to address it. All right. Thanks again. All right, everybody, thank you so much for listening and watching. Go check out spyglass. Spyglass.org highly recommend you sign up and we'll see you next time on big Technology Podcast.
- [72:30]
B
Par le tu francais hablas par le italiano? If you've used Babbel, you would Babbel's conversation based technique teaches you useful words and phrases to get you speaking quickly about the things you actually talk about in the real world. With lessons handcrafted by over 200 language experts and voiced by real native speakers, Babbel is like having a private tutor in your pocket. Start speaking with Babel today. Get up to 55% off your Babbel subscription right now at babbel.com acast spelled B A B-B-E-L.com acast rules and restrictions may apply.