Hey everyone, it's Josh and it's another Monday interview. Mike Isaac is a long-time technology writer and has been a technology correspondent at The New York Times since June 2014. During that time, he's covered a variety of technology companies including OpenAI, SpaceX, Anthropic, Apple and Uber, which he published a book about in 2019 called Superpumped: The Battle for Uber. But the bulk of his recent reporting has been about Meta, the company that owns Facebook. So last week, I interviewed Mike about what exactly is going on over there. What prompted them to make such a big bet on a VR metaverse? And what, from wearable smart glasses to prediction markets, we can expect Mark Zuckerberg to do next. Hope you enjoy.
Mike, thank you so much for taking some time to talk with me today.
Thank you for having me. Sorry it took me this long to get on here, but you dealt with my scheduling.
Happy to have you no matter what. We did an episode, at this point I think it was like a month and a half or so ago, about Meta Horizon Worlds and the metaverse as a concept and how that concept cost the company Meta $80 billion. We are now at the point where Meta is trying to figure out what they are going to do for their business next. And I wanted to have you on because you have for a long time reported on this specific company for The New York Times. So before we get into the meat and potatoes of it, how long have you been working on this subject as a reporter? And what are the sort of big things you've noticed in your coverage that maybe make this company, Meta, different from the other ones?
Yeah, so I'm out in San Francisco, well, Oakland, and been out here since 2004, but I started covering tech in 2010. And really jumped around a few different outlets, but basically had been covering Facebook and Twitter really and kind of social media. Kind of still back in the era where it was like this question of is this going to be the next MySpace, right? Like depending on how old you are. I don't know how old everyone in your audience is, but I'm 41, so I've been around with that first era of you had AOL dial-up internet, you had instant messenger, you had MySpace where the question was because MySpace kind of died, everyone was like, is this just a fad, you know?
Were you a MySpace kid back in the day?
I was. I was. I'm praying to God no one can find my old profile. I have like, it, I had hair. I had like an emo haircut, I think. Good lord.
But I think to your question of like what I've noticed is like Facebook and Zuckerberg have been different in, I would argue, evading for a long time evading some of the pitfalls that like other companies have gone down, whether it's, you know, like figuring out how to hook themselves into the bones of the internet in a lot of ways. Basically from the bottom level like Facebook as an app, and before it was an app, it was a desktop site. That's how they kind of got hooked in everywhere. It's very much hard to, I guess what I would say is it's hard for them to die in the traditional sense that we think of a company like backs up and goes out of business, which is why later on as they've gone down some different holes, I don't think it's necessarily lights out for them forever. But yeah, I think like the biggest thing is they have a lasting power that is not common among many Silicon Valley companies, what I'd say.
And I think that's an interesting point about how they contribute to the architecture of what we consider the modern internet to be.
Because they managed to be in the right place at the right time, release a whole bunch of open standards that a lot of other people, a lot of other companies, a lot of other developers immediately signed on to, and they were able to sort of reshape the web in their own image during Web 2.0.
Yep. No, I mean that's it's I'm glad you get into the nuts and bolts of that because I really think that this is like, look, I have a lot of criticisms of Zuckerberg and the management team over there, but I do think they're pretty astute students of like tech and business history and study how different companies from like Linux and open source sort of like software standards back in the day managed to win out against some other types of software or companies that were building things that ended up seeming to have like a real strong, you know, moat or defense around them and their margins were really high and then quickly sort of had the ground underneath them eroded. And exactly to your point, they built PyTorch, which is like this like very widely used, I believe, standard. They've done a lot of moves around the web of yesteryear and then like certain development tools that I think have been pretty smart, if that makes sense.
What are the tools that Facebook developed that have become standards?
Yeah, so I think PyTorch is probably the one that they get the most credit for. They had this guy there. He had created this, you know, essentially free software that a lot of developers in the AI field or in like just sort of like folks who are building what kind of became a bottomish layer, bottom to middle layer of the internet ended up using. And like I think this as well as like a few open source hardware projects. They have like, I forgot the exact name, but it's kind of like a foundation where they're like, here we're open sourcing a lot of our models for how we build routers, I want to say, or different like data center type things, like before it got way competitive. And the philosophy I think is sound, which is, and they can afford to do this because they print money in their ads business, but like give away as many tools as we can in order to get people hooked on them, you know, and then that builds into how we last. Does that make sense?
Absolutely. I mean, I think another great example of that would of course be internet.org or later Facebook Free Basics where they're like, 'Oh yeah, we'll provide free internet to people in India.' But of course the catch is that you get access to maybe not quite the entire internet and you definitely get a lot more Facebook on your internet than anything else.
God, man. Internet.org feels so long ago, but you're exactly right. There was a huge moment for them and they were also, Zuckerberg was like realizing, you know, we've conquered a or we've taken over a lot of the developed world, the US and Canada like our businesses where our ad dollars are really, the margins are much higher and the average revenue per user is much higher, but like there are limits to that. So, they got to go everywhere. Where can they? And India is one of the most populous countries, the most populous country in the world. Yeah, yeah, yeah. So, you know, that's essentially, okay, this is obviously where we go next, and that's the sort of bargain that they make. We want to give it everywhere. They ended up running into some real strong...
they failed ultimately in that because, you know, their government under Modi and stuff was like, this is probably not smart for us to let Westerners kind of control the internet, but it was a sound theory in terms of like, you know, conquest and sort of digital colonialism, if you want to call it that, you know, and it just didn't work out for them. But, you're exactly right. They're just like, here, we give away internet, but the internet is Facebook, basically.
Well, into the point of things that they tried to build new frontiers on, I think this is a great time to pivot into talking about the metaverse, or at least Meta's conception of the metaverse.
If you had to sum up why Mark Zuckerberg invested so much time and resource into this thing, what would you say?
I would argue it boils down to his real insecurity of not owning the ground beneath his feet. And by that, I mean platform... This is going to get nerdy, but like platform risk, right? Ultimately, Facebook is an app built atop someone else's real estate. Their most popular apps are still Facebook, Instagram, WhatsApp. Those are iOS and Android apps, basically, you know, like the desktop is no longer, I mean, like it's a meaningful revenue source for them materially, but is clearly not the primary driver, and Mark found out very quickly in a world in which the world is rapidly changing from desktop, you know, internet access to mobile devices, that he no longer controlled that experience. He went from browsers, which again, he didn't own the browser, but he still had way more control inside of a browser, to the app environment, which is just you play by Apple's rules, you play by Google's rules, and over the years, Apple got increasingly restrictive on what you were allowed to do. So, ultimately, what is that, the thing he's been searching for? What is the next frontier? What is the next platform? And through sheer force of will on his behalf, he said, 'This is my bet,' but also really just what he wanted, you know? He wanted this.
Well, that's another question that I have, right? Is why was this his bet? We talked a little bit in our episode about Oculus, Palmer Luckey, the way that Palmer Luckey was kind of the hot whiz kid for a little bit. I have to assume there was a level of professional envy there from Mark Zuckerberg, but it also does seem that he really truly thought that VR was imminent as the next big thing. Why?
Mark, I think, is one of those people who actually, he kind of lacks focus in ways, but also the person who put this in front of him was Marc Andreessen, who invested, one of his partners invested in Oculus.
Yeah, he got a very healthy exit when Oculus was acquired by Facebook, if I recall correctly.
That's right. That's exactly right. They, I want to say they paid 4 billion for Oculus, and Andreessen was like the main or primary, let's say, investor. And Andreessen is on the board of Facebook. There's like, folks were like, 'Are you self-dealing?' And da da da da da. But, Andreessen, who, if you don't follow him, approaches everything like with the world is on fire and is a zealot and is like fully convinced of whatever is the next thing, comes to Zuckerberg and says, 'Look, this is huge. You got to try this platform.' And Zuckerberg, I remember this, too, because it was like 2012, and I had tried the Oculus at a tech conference. Um, actually, Shane Smith was there of Vice and was also putting it on. It was really weird. It was like a cool like thing, like, 'Oh, this is a like, oh it's crazy like you have this immersive VR thing.' It was definitely more sophisticated than past iterations of it had been, and so I was like, okay, that's cool, but also you felt nauseous, like I couldn't do it for more than 15 minutes, like, and like the limitations were there. But Zuckerberg, I think was just, I imagine part of it was wanting so badly to believe that this was the next thing because he could like sort of acquire it and then sort of pour a bunch of money into it and fuel the fire to kind of make it the next thing. I think the flaw I would say in that thesis is like he wasn't chasing user behavior, he was sort of trying to prognosticate, does that make sense?
Yeah, it sort of feels like almost magical thinking where it's like if enough people get on board with this thing, it will be the thing. Time and time again they ran up against the limitations of technology and physics, frankly, that there are certain things that we are simply not capable of doing in electronics and there are other things that we will probably never be capable of doing, and for some reason they, at least from my perspective, just felt so passionate about this that they truly believed that they could defy the force of gravity itself.
I really like that you say physics because that's what I kept... when I, in every interview I was like, they have this thing that where the momentum of, they look at like, oh Moore's law tells us like scaling is going to like get us like, you know, a couple of years from now or a couple of years from...
on the show like Moore's law is not a law, it is a projection that people have been following and now transistors can't really get any smaller.
Exactly, exactly, but they latch on to it as like gospel and say, 'This is...' It's never about where we are now, it's where I know we're going to be 5 to 10 years from now. And they have the conviction of people who, I mean, it's almost like religious sort of level of conviction of like this is going to happen. I was just thinking back on the interviews over those years, you could never really talk to them about what is capable now or like the numbers of people who are using or buying these devices now, which were low and continue to be low for a long time because it was always, 'Well, if we put enough money or seed the ecosystem with enough games like any platform, it's a chicken and egg problem,' so they always wanted to try to like prop it up on both sides, whether it's apps or people. And but it was just, it never was there, you know.
Well, a question I have for you about that is you were very much on the ground during all of the attempts for this to become a thing. What were you seeing in terms of people's thoughts, opinions? To what extent do you think that they were just trying to project an image of confidence versus to what extent were they true believers in the vision? Because I feel like with stuff like this, it's always a little bit of column A and a little bit of column B. But I'm curious what you were seeing and feeling.
I mean, look, I've lived here, I've lived in the Bay Area since 2004, and so there's definitely, this is definitely like the locus of people who enjoy new technology for the sake of it, right? You know, they... And like I respect the idea of people who are sort of pushing boundaries on things that are improvements to how tech works with a specific purpose in mind, right? Like there's, I'm not a Luddite in the sense that I think all tech is bad. I think there's a lot of things that it can be used for good. I think a lot largely in the realm of science and drug research and things like that. But I, but like I think in general the attitude out here is look what I can do and look how I can push this forward and like see what we can do because we can. And so a lot of those people were pouring into VR, AR research because they were genuinely interested in it, and Meta, once Zuckerberg sort of made this a priority, was recruiting a lot of those folks. But then I would say that is a vocal minority of people who in general, like, you know, Meta was also willing to lose a lot of money putting, plowing cash into this, and so they hired entire divisions for marketing. They hired, you know, partnerships people to convince Disney they needed a chief metaverse officer or whatever, you know, like it became a whole industry of folks who were... there was like a lot, a vocal minority of people who really loved and believed in the tech and were like true believers, and then a much larger majority of people who were like, well, I'm told this is where it's going, and also I need a job, and this is what we're going to do. So like let's do it, you know.
That's something that I always find so interesting is you have these narratives that spin up, and the narratives of course are driven by the tech companies because they want to frame whatever it is that they are working on as an inevitability. And to use a contemporary example, the sheer number of times that I've heard that, well, AI is the future. It's going to change the way that you work. It is, I promise, and therefore you need to figure out how to incorporate it into your workflow, as opposed to, well, what is this technology actually doing and is it useful for me in any way, shape or form whatsoever? And these narratives are then able to drive this hype cycle that sometimes are able to bring people on board and sometimes are not. I think though that something that the metaverse thing had in common with the other prevailing hype trend of the time, which is Web3, is that both of them had a pretty high barrier to entry. In the case of VR, it was the headset, which is a huge pain in the ass to wear for a lot of people. On the Web3 side, it was technical knowledge and of course money because actually investing in any of that was so expensive. And I remember I was in 2020, 2021 actually working for a tech company and seeing a lot of this stuff happen in real time where it was like, well, how could we potentially integrate metaverse into our product, integrate crypto into our product. And now that we're sitting where we are now, which is to say that bubble has burst a long time ago and then some of it's still around, particularly on the crypto side, but it's obviously not the wave of the future. I'm wondering what you have seen from that experience or learned from that experience that you can apply to what's happening right now and going forward.
I have to constantly, I'm just as susceptible when everyone out here is like, this is going to happen, this is going to happen. Like I have to be like, is this going to happen? Like maybe it is going to happen.
You see, this is why I live in New York is because I can't.
I know, no, and that's the thing. Like if you're out here, it's just sort of, it's crazy how much of a take-for-granted X thing is going to happen just is on the street. And there was a moment, I remember this specifically in the context of crypto and then later metaverse, but for me I wrote about, you know, all these people inside of Meta and Google or whatever like leaving to start crypto companies because like there was much, much more upside in starting, raising a bunch of money, easy money to build a crypto company compared to doing it for Meta or Google. And I was sort of like, yeah, this is coming, da da da, and then some guy in my mentions was like, what is a good use case for crypto?
And I had to like stop and I was like, well, like how are people using it? Like what beyond, let's say, money transfers or whatever, which is already a thing that exists, you know, like I'm still trying to figure out how it would be more efficient to use Bitcoin for money transfers even though it costs a load of money to do that, and I couldn't come up with anything. And so this idea that I think this thing has really returned to me in my reporting and my career is as this North Star is just like humans have agency. We have the ability to make decisions on what we want the world to look like and what we want our lives to look like. And so whenever someone just gets on their soapbox and starts saying this is what's happening and you better get on the train or not, I immediately recoil. And I think crypto and the metaverse are, it's not a one-to-one to AI now because AI is not a new discipline by any stretch, right? And you know this, you're a technologist so you know that like artificial intelligence has been around for decades in some form or another, and it's also a very bad catch-all term because it means a lot of different things in different contexts.
Yeah, but if we're talking about predictive text modeling, that's been around forever.
100%. So like I think it's more what forms is it going to take? What are we comfortable with and how do we want to, you know, have that exist in our world or not, you know? And so as I was talking about before, like I think there are forms of, let's say, cancer research or drug, I think AlphaFold is like a worthwhile project that Google is doing, you know? Like I think things like where you can model predicted behavior in drug research or cancer trials is actually very interesting if you actually test that. But I think we should have real questions on what that should look like on other parts of the world and how humans should or should not intervene. Does that make sense?
It does, yeah. And I want to get back to the bigger conversation about like frontier models and machine learning in a little bit. But, for now, I want to pull it back to the metaverse again because, you know, one of the things that I noticed was how much of this seemed to just be a scramble for speculative value. This was most overt in the case of out-and-out real estate schemes, basically. Of course, there was like Decentraland and those models of metaverses where it's like, 'Well, here's all this digital land. You better come get more land in Atlas Earth or whatever because you don't want to miss the train.' Mark, on the other hand, I think he understood that there could be some value in that. But, the major value was, as you had been saying at the beginning, sort of getting the standards, the tools in place so that he could own the land under his feet and then, of course, lease that land out to everybody else who wanted to use his technology. But, of course, it didn't work out that way. On March 18th of this year, Meta announced that it would be shutting down Horizon Worlds in VR in June. And they ended up partially backtracking on that, but they did most of it. They sunset the vast majority of Horizon as a platform including Horizon Workrooms and Spark AR, which was their toolkit for building 3D objects in mixed reality environments.
Also on May 19th, and you reported this, Meta laid off 8,000 employees company-wide, and this is making good on a plan that they had announced in April, and your reporting noted that these cuts were actually first rumored to hit Reality Labs back in January. And that is a unit that is a top-level unit in Facebook. It sits adjacent to platform under the Meta banner and has now lost 80 billion dollars. So my question for you is what is going on at Reality Labs right now?
So it's actually interesting. I think if Zuckerberg could go back or hit rewind a few years ago or something, and this is just me sort of speculating, but he probably would have tried to buy or at least made strong overtures to a company like Roblox or Fortnite, Epic Games which owns Fortnite. And for a number of reasons he didn't, probably because he didn't suspect he should do it, but then also like just an antitrust environment now is very different than it was 10-15 years ago. So in the absence of buying that, and because he went down this sort of metaverse in his mind, and to get back to your initial point, their conception of the metaverse was like really focused on hardware. I think they've realized like that conception, this sort of Oculus-based, now called Meta Quest thing is not anywhere near where they wanted it to be or anywhere near...
It never will be. It's impossible to say at this point.
Exactly. Yeah, like that, it's not where they want to be now basically. So what they're truly chasing inside now is trying to restart from a mobile perspective, which is how a lot of people use Fortnite and Roblox and these other platforms, which I would argue is like a real more solid conception of the metaverse, like this idea where you're playing across phones or desktops or whatever and you can be John McClane in Die Hard fighting Lady Gaga in like a level or something. Like that's a more realistic, that's a conception of the metaverse that is actually happening right now where people are using it. And so that's what they're kind of focusing on. How do we build the architecture? They're sort of working on a new engine, I believe, to like, it's not going to be Horizon necessarily, I think. Or it might be Horizon, I can't remember, but like...
Well, I know that they rebuilt Horizon Worlds in Horizon engine because it used to be built on top of Unity, right?
That's right. And they moved from Unity, that's exactly right. And so they're moving to basically what's going to work on phones in a better way than... because, and this is like a move that they probably should have done years ago, you know, when they saw that the headset thing wasn't working out. So, I think that is what that division's kind of working on. And then meanwhile, their happy accident, we don't have to go down this hole, but their happy accident was the glasses and AR actually taking off, Meta Ray-Bans. And so they're pouring a ton of money into that now, basically. That's like their hardware play.
The sense that I get is that Zuckerberg himself and a lot of the other people who were high up, especially Andrew Bosworth, were so starstruck by the idea of this ultra-futuristic thing that they lost track of the potential profitability of a much more prosaic implementation of this stuff. Is that fair to say?
Yeah, they, I mean, I really do think that they love tech for tech's sake in some ways. They love Snow Crash, Ready Player One. There's a famous story. I think it was reported on at the time, but like Mark Zuckerberg read Ready Player One 15 years ago or whatever and was like proselytizing around the office. Everyone has to read a copy of this. Like was very happy to talk about it. And it's funny that like often some of the themes of these books are lost on the people who love them so much.
You're not supposed to do the plot of Snow Crash, guys. That's not the idea.
Right. They love sci-fi. I just don't think they take away... You could...
go like death of the author argument here, but I just think they take away something different. They read a lot of this stuff or absorb a lot of these things and are excited about it rather than go to a dark place, if that makes sense.
Yeah, some of the stuff that really interested me actually about Reality Labs was some of the like pure tech stuff. There's that one video from one of the years of Meta Connect where in like the back corner or so that this is the infamous keynote, the one where he starts by going, "Hey."
[laughter] I remember that.
And uh got the free bottle of Sweet Baby Ray's on the shelf and all that or whatever.
