• 3 months ago
Meta CEO Zuckerberg sits down with Verge deputy editor Alex Heath during the company’s Meta Connect conference to talk all things AR, AI, and social media. The two discuss Meta’s new Orion AR glasses, the company’s partnership with Ray-Ban, why Zuckerberg is done with politics, and much more.

Category

🤖
Tech
Transcript
00:00:00Mark, we just tried Orion together.
00:00:05Yeah, what'd you think?
00:00:06Refresh off of it.
00:00:08It feels like true AR glasses are finally getting closer.
00:00:11Orion is a product that you all have been working on for five plus years?
00:00:16Yeah, almost 10.
00:00:17Almost 10?
00:00:18Yeah.
00:00:19So take me back to the beginning when you started the project, when it started in research.
00:00:23Like what were you thinking about?
00:00:24What was the goal for it?
00:00:27I think a lot of it goes all the way back to our relationship with mobile platforms
00:00:33and all that.
00:00:34So we have lived through one major platform transition already because we started on web,
00:00:41not on mobile.
00:00:43And mobile phones and smartphones kind of got started around the same time as Facebook
00:00:49and kind of early social media was getting started.
00:00:52So it didn't really get to play any role in that platform transition.
00:00:56But going through it, we weren't born on mobile.
00:00:59We kind of had this awareness that, okay, web was a thing.
00:01:04Mobile is a thing.
00:01:05It is different.
00:01:06There are strengths and weaknesses of it.
00:01:08There's this continuum of computing where now you have a mobile device that you can
00:01:12take with you all the time.
00:01:13And that's amazing.
00:01:15But it's small, kind of pulls you away from other interactions.
00:01:19Those things are not great.
00:01:21But there was sort of this recognition that just like there was the transition from computers
00:01:25to mobile, mobile was not going to be the end of the line.
00:01:30So as soon as we started becoming like a more, I guess I'd say stable company, like once
00:01:38we found our footing on mobile and we weren't like clearly going to go out of business or
00:01:41something like that, I was like, okay, let's start planting some seeds for what we think
00:01:49could be the future.
00:01:50It's like mobile is already kind of getting defined by 2012, 2014-ish.
00:01:57It was generally too late to really shape that platform in a meaningful way.
00:02:01We had some experiments that didn't succeed or go anywhere.
00:02:05So pretty quickly I was like, okay, we should focus on the future because just like there's
00:02:09the shift from desktop to mobile, new things are going to be possible in the future.
00:02:13So what is that?
00:02:16I think the kind of simplest version of it is basically what you started seeing with
00:02:21Orion, right?
00:02:22It's like the vision is a normal pair of glasses that can do two really fundamental things.
00:02:29One is put holograms in the world to deliver this realistic sense of presence, like you
00:02:35are there with another person or in another place, or maybe you're physically with a person,
00:02:40but just like we did, you can pull up a virtual pong game or whatever.
00:02:45You can work on things together.
00:02:46You can sit at a coffee shop, pull up your whole workstation of different monitors.
00:02:50You can be on a flight or in the backseat of a car and pull up a full screen movie theater
00:02:55and like, okay, all these things, like great computing, full sense of presence, like you're
00:03:00there with people no matter where they are.
00:03:02Thing two is it's the ideal device for AI.
00:03:09And the reason for that is because glasses are sort of uniquely positioned for you to
00:03:15be able to let them see what you see and hear what you hear and give you very subtle feedback
00:03:21back to you where they can speak in your ear or they can have silent input that kind of
00:03:27shows up on the glasses that other people can't see and doesn't take you away from the
00:03:31world around you.
00:03:32And I think that that is all going to be really profound.
00:03:34Now when we got started with this, I had thought that kind of the hologram part of this was
00:03:39going to be possible before AI.
00:03:43So it's sort of an interesting twist of fate that the AI part is actually possible before
00:03:49the holograms are really able to be mass produced at kind of an affordable price.
00:03:53But that was sort of the vision.
00:03:56I think that it's pretty easy to wrap your head around and there's already a billion
00:04:00to 2 billion people who wear glasses on a daily basis, just like everyone who didn't
00:04:04have smartphones were kind of the first people to upgrade to smartphones.
00:04:07I think everyone who has glasses is pretty quickly going to upgrade to smart glasses
00:04:11over the next decade.
00:04:12And then I think it's going to start being really valuable and a lot of other people
00:04:16who aren't wearing glasses today are going to end up wearing them too.
00:04:21That's kind of the simple version.
00:04:24And then I think as we've developed this out, there are all these sort of more nuanced directions
00:04:30that have emerged too.
00:04:31So we've started, you know, while that was kind of the full version of what we wanted
00:04:35to build, there were all these things that we said, okay, well, maybe it's really hard
00:04:41to build normal looking glasses that can do holograms at an affordable price point.
00:04:45So what parts of that can we take on?
00:04:47And that's where we did the partnership with SLR Luxottica.
00:04:49So it's like, okay, before you have a display, you can get normal looking glasses that can
00:04:54have a camera, that can have a microphone, great audio, can capture content, you can
00:05:01stream video at this point.
00:05:02But the most important feature at this point is the ability to access meta AI and just
00:05:06kind of have kind of a full AI there, multimodal, because it has camera.
00:05:12And I mean, that product is starting at $300.
00:05:14And, you know, initially, I kind of thought, hey, this is sort of on the technology path
00:05:18to building full holographic glasses.
00:05:20At this point, I actually just think both are going to exist long term, right?
00:05:24I think that there are going to be people who want the full holographic glasses.
00:05:28And I think that there are going to be people who prefer kind of the superior form factor
00:05:32or lower price of a device where they are primarily optimizing for getting AI.
00:05:36I also think there's going to be a range of things in between, right?
00:05:40So there's the full kind of field of view that we that you just saw, right, 70 degrees,
00:05:44really wide field of view for glasses.
00:05:46But I think there are other products in between that, too.
00:05:51There's like a heads up display version, which, you know, for that, you probably just need,
00:05:55you know, 20, 30 degrees.
00:05:58You can't do kind of the full kind of world holograms where you're interacting with things
00:06:05like you're not going to play ping pong and 30 degree field of view.
00:06:08But you can communicate with AI.
00:06:11You can text your friends, you can get directions, you can see the content that you're capturing.
00:06:17So I think that there's a lot there that's going to be compelling and at each step along
00:06:20this continuum from display list to small display, to kind of full holographic, you're
00:06:25packing more technology in.
00:06:26So each step up is going to be a little more expensive, is going to have a little more
00:06:31constraints on the form factor, even though I think we'll get them all to be attractive.
00:06:35You'll be able to do the kind of simpler ones and much smaller form factors permanently.
00:06:41And then, of course, there's the mixed reality headsets, which kind of took a different direction,
00:06:44which is going towards the same vision.
00:06:47But on that, we said, okay, well, we're not going to try to fit into a glasses form factor
00:06:52for that one.
00:06:53We're going to say, okay, we're going to really go for all the compute that we want.
00:06:57And we're going to say, okay, this is going to be more of like a headset or goggles form
00:07:00factor.
00:07:01And my guess is that that's going to be a thing long term too, because there are a bunch
00:07:04of uses where people want the full immersion.
00:07:08And if you're sitting at your desk and working for a long period of time, you might want
00:07:11the more computing power than you're going to be able to get.
00:07:13But I think that there's no doubt that kind of what you saw with Orion is the, I think,
00:07:19kind of quintessential vision of what people, what at least kind of I thought and continue
00:07:25to think is going to be the next multi-billion person major computing platform.
00:07:31And then kind of all these other things are going to get built out around it.
00:07:33It's my understanding that you originally hoped Orion would be a consumer product when
00:07:37you first set out to build it.
00:07:38Yeah.
00:07:39Yeah.
00:07:40Orion was meant to be our first consumer product, and we weren't sure if we were going to be
00:07:44able to pull it off.
00:07:45I mean, in general, it's probably turned out significantly better than our kind of
00:07:4950-50 estimates of what it would be.
00:07:51But we didn't get there on everything that we wanted to.
00:07:54I think we still want it to be a little smaller, a little brighter, like a little bit higher
00:08:00resolution, and a lot more affordable before we kind of put it out there as a product.
00:08:05And look, we have line of sight to all those things.
00:08:07So I think we'll probably have the thing that was going to be the V2 end up being the consumer
00:08:14product, and we're going to use Orion with developers to basically cultivate the software
00:08:19experience.
00:08:20That way, by the time we're ready to ship something, it's going to be much more dialed
00:08:22in.
00:08:23But to be clear, you're not selling Orion at all.
00:08:25I guess what I'm wondering is, when you made the call, I think it was around 2022, to say
00:08:30Orion is going to just be an internal kind of dev kit, how did you feel about making
00:08:34that call?
00:08:35Was there any part of you that was like, man, I really wish this could have just been the
00:08:38consumer product we had built for years to make?
00:08:41I always want to ship stuff quickly and all that, but I think it was the right thing.
00:08:50On this product, I think that there's a pretty clear set of constraints that I think you
00:08:56want to hit, especially around the form factor.
00:08:59I mean, it is very helpful for us that sort of chunkier glasses are kind of ascendant
00:09:06in the fashion world, because that allows us to build glasses that are going to be fashionable,
00:09:10but also tech forward.
00:09:12But even so, I'd say these are unmistakably glasses, they're reasonably comfortable, they're
00:09:21under 100 grams.
00:09:22I wore them for two hours, and I couldn't really tell.
00:09:26But I mean, I think we aspire to building things that look really good, and I think
00:09:32this is good glasses, but I wanted it to be a little smaller so it can fit within, really
00:09:38fashionable.
00:09:39When people see the Ray-Bans, there's no compromise on fashion.
00:09:43Part of why I think people like them is, yeah, you get all this functionality, but even when
00:09:46you're not using it, they're great glasses.
00:09:49I think for the future version of Orion, that's the target too.
00:09:53We want to make it so that most of the time you're going through your day, you're not
00:09:57computing, or something is happening in the background or something.
00:10:01So it just needs to be good in order for you to kind of want to keep it on your face.
00:10:06I feel like we're almost there.
00:10:08I think we've made more progress than anyone else in the world that I'm aware of, but it
00:10:12didn't quite hit my bar.
00:10:14And similarly, on price, these are going to be more expensive than the Ray-Bans.
00:10:20There's just a lot more tech that's going in them, but we do want to have it be within
00:10:25a consumer price point, and this was outside that range.
00:10:29So I wanted to wait until we could get to that range in order to have something that
00:10:33we ship.
00:10:34So when you get the first commercial version, whenever it is in the next couple, few years,
00:10:37will it be a developer-focused product that you're selling publicly?
00:10:41You want it to be a consumer-ready.
00:10:43And that's why I'm asking about the strategy, because Apple, Snap, others have decided to
00:10:48do developer-focused plays and get the hardware kind of going with developers early.
00:10:52But are you saying you're skipping that and you just want to go straight to consumer?
00:10:56We are using this as a developer kit, just primarily internally and maybe with a handful
00:11:00of partners.
00:11:01But I mean, I think at this point, Meta is by far the kind of premier developer of augmented
00:11:07reality and virtual and mixed reality software and hardware in the world.
00:11:11So you can think of it as a developer kit, but we just have a lot of that talent in-house.
00:11:15And then we also have well-developed partnerships with a lot of folks externally who we can
00:11:19go to and work with them as well.
00:11:21So I don't think we need to go announce a dev kit that kind of arbitrary developers
00:11:27can go buy to get access to the talent that we need to go build out the platform.
00:11:30I think we're kind of in a place where we can work with partners and do that, but that's
00:11:33absolutely what we're going to do over the next few years.
00:11:35We're going to hone the experience and figure out what we need to do to really nail it when
00:11:39it's ready to ship.
00:11:40A lot has been written about how much you're spending on Reality Labs.
00:11:43And you probably can't have an exact number, but if you were to guesstimate the cost of
00:11:50just building Orion over the last 10 years, are we talking five plus billion?
00:11:56More than that?
00:11:57Yeah, probably.
00:11:58Probably.
00:11:59Yeah.
00:12:00So overall, for Reality Labs, for a while, a lot of people thought of all of that budget
00:12:06is going towards virtual and mixed reality.
00:12:09And I actually think we've said publicly that our glasses programs are a bigger budget than
00:12:16our virtual and mixed reality programs, but that goes across all of them, right?
00:12:20So that's the kind of full AR, that's the display-less glasses, like all the work we're
00:12:26going to do on Ray-Ban.
00:12:28And we just announced the expanded partnership with Luxottica, SLR Luxottica, great company.
00:12:33We've had a great experience working with them.
00:12:35They designed so many great glasses, and I think working with them to do even more is
00:12:41going to be really exciting.
00:12:42So there's a lot more to do there on all of these things.
00:12:45How does this partnership work?
00:12:47And this new renew that you just did with them, how is it structured?
00:12:52What does this deal look like?
00:12:55I think it was more of just a kind of commitment from the companies that we're feeling pretty
00:13:01good about how this is going, and we're going to build a lot more glasses together.
00:13:06And part of the way it works is, you know, rather than having sort of doing one generation
00:13:13and then designing the next generation, I think by having a longer term partnership,
00:13:19it allows the teams to not just have to worry about one thing at a time, then like, okay,
00:13:25is this one going to be good?
00:13:26And then how do we build on that for the next one?
00:13:29Now we can start like a multi-year roadmap of many different devices, knowing that we're
00:13:34going to be working together for a long time.
00:13:36So I'm optimistic about that.
00:13:38That's sort of how we work internally, where we don't just, I mean, sometimes when you're
00:13:42early on, you definitely want to learn from each device launch.
00:13:45But when there are things that you're committed to, I don't think you want the team to feel
00:13:51like, okay, if we don't get the short term milestone, then like, we're going to cancel
00:13:55the whole thing.
00:13:56All right.
00:13:57So...
00:13:58Are you buying a stake in Elosora Luxottica?
00:14:01Yeah, I think we've talked about investing in them.
00:14:04It's not going to be a kind of major thing.
00:14:06I'd say it's more of a symbolic thing.
00:14:10We want to have this be a long term partnership.
00:14:13And as part of that, I thought that this would be kind of a nice gesture.
00:14:18And I fundamentally believe in them a lot.
00:14:20I think that they're going to go from being the premier glasses company in the world to
00:14:26I think one of the major technology companies in the world.
00:14:30My vision for them and how I think about it is that if you think about how Samsung in
00:14:36Korea kind of made it so that Korea became one of the main hubs of building phones in
00:14:43the world, I think that this is probably kind of one of the best shots for Europe and Italy
00:14:48in particular, to become a major hub for manufacturing and building and designing the
00:14:53next major category of smart glasses and sorry, of computing platforms overall.
00:15:00And I think that they're kind of all in on that now.
00:15:03And it's been this interesting question, because, you know, they have such a good business and
00:15:08such deep competence and the areas and I've gotten more of an appreciation of how strong
00:15:13of a technology company they are in their own way, right?
00:15:17So, you know, designing lenses, designing the materials that you need to make fashionable
00:15:21glasses that can be light enough, but but also kind of like feel good.
00:15:29They have they bring a huge amount that I think is people in kind of our world, the
00:15:35tech world probably don't necessarily see, but but I think that they're really well set
00:15:39up for the future.
00:15:40So I believe in the partnership.
00:15:41I'm really excited about the work that we're doing together.
00:15:43And fundamentally, I think that that's just going to be a massively successful company
00:15:46in the future.
00:15:48Is it set up to where they control the designs or and you provide the tech stack or do you
00:15:52collaborate on the design?
00:15:53I think we collaborate on on everything.
00:15:56Okay, it's actually I mean, part of working together is you kind of build a joint culture
00:16:02over time.
00:16:03And there were a lot of really sharp people over there, who I think it took maybe a couple
00:16:10versions for us really to gain an appreciation for how we each of us approach things because,
00:16:16you know, they really think about things from this like fashion, manufacturing, lens,
00:16:21selling optical devices perspective, then we obviously come at it from a kind of consumer
00:16:26electronics, AI, software perspective, but I think over time, we just kind of appreciate
00:16:33each other's perspectives on things a lot more.
00:16:35And I mean, I'm like constantly talking to them about things to get their ideas on different
00:16:40things.
00:16:41I mean, but you know, when partnerships are working well, when you reach out to them,
00:16:45to get their opinion on things that are not actually currently in the scope of what you're
00:16:49working on together and, and because I do that frequently with with Rocco, who runs
00:16:55the wearables and Francesco, who's their, their, their CEO, it's, and our team does
00:17:00that too, with like, a large part of the working group over there.
00:17:03It's, it's a good crew there, they kind of they share good values, they're, they're really
00:17:07sharp.
00:17:08And like I said, I believe in them a lot.
00:17:09And I think it's gonna be a very successful partnership and company.
00:17:12Yeah.
00:17:13How many Ray-Ban Metas have you sold so far?
00:17:15I don't know if we've given a number on that, but yeah, it's, it's going very well.
00:17:22You know, one of the things that I think is interesting is we, we underestimated demand.
00:17:27So part of, you know, one thing that is very different in the world of consumer electronics
00:17:32than software is there are fewer kind of supply constraints in software.
00:17:39There are some, I mean, like some of the stuff that we're rolling out, like the voice on
00:17:42meta AI, we need to meter it as we're rolling it out because we need to make sure we have
00:17:46enough inference capacity to, to handle it.
00:17:48But fundamentally we'll resolve that in weeks, right?
00:17:51But for manufacturing, it's like, you make these concrete decisions like, okay, are we
00:17:55setting up four manufacturing lines or six?
00:17:59And you know, each one, it was a big upfront CapEx investment.
00:18:04And you're basically just deciding upfront the velocity at which you're gonna be able
00:18:08to generate supply before you know what the demand is.
00:18:10So on this one, we thought that Ray-Ban meta was probably going to sell three or five times
00:18:20more than the first version did.
00:18:23And we just dramatically underestimated what the, so now we're in this position where it's
00:18:30actually been somewhat hard for us to gauge what the real demand is because they're sold
00:18:34out and you can't get them.
00:18:35So if you can't get them, like then, then how do you know where, where the actual curve
00:18:37is?
00:18:38So that's, we're basically getting to the point where that's resolved.
00:18:42Now we, we kind of adjusted, we made the decision to build more, more manufacturing lines.
00:18:47It took some time to do it.
00:18:48They're online now.
00:18:49We're kind of, you know, it's, it's not just about being able to make them, you need to
00:18:51get them into all the stores and get the distribution right.
00:18:55We feel like that's in a pretty good place now.
00:18:56So I'd say over the rest of this year, we're going to start getting a real sense of the
00:19:00demand.
00:19:01But while that's going on, the glasses keep getting better because of over the air AI
00:19:06updates.
00:19:07The hardware doesn't necessarily change, even though we keep shipping new frames and new
00:19:11type of, you know, they're adding more transitions lenses because people want to wear them indoors.
00:19:17And that's an interesting thing because I mean, people, sunglasses are a little more
00:19:22discretionary.
00:19:23So I think a lot more people early on, we're thinking, Hey, I'll try, I'll, I'll experiment
00:19:27with this with sunglasses.
00:19:28I'm not going to make these my primary glasses.
00:19:32Now we're seeing a lot more people say, Hey, this is actually really useful.
00:19:35I want to be able to wear them inside.
00:19:36I want them to be my primary glasses.
00:19:37So whether that's kind of working with them through the optical channel or the transitions,
00:19:40that's an important part.
00:19:41But the, the AI part of this, um, is also, it just keeps getting better.
00:19:47I mean, we talked about it at, um, at connect, I mean, basically the ability to now, you
00:19:54know, for the next few months, when we roll this out, real time translations, you're traveling
00:19:58abroad, someone speaking Spanish to you, you just get it translated into English in your
00:20:02ear and just roll out to more and more languages over time.
00:20:04I think we're starting with a few, we'll, we'll, we'll kind of hit more over time.
00:20:07Um, I tried that.
00:20:09Well, actually I didn't try real time, but I tried looking at a menu in French and it
00:20:12translated into English.
00:20:13And then I was like, at the end, I was like, what is Euro and dollar actually?
00:20:17And it did that too.
00:20:18And then I'm also starting to see the continuum of this to Orion in the sense of the utility
00:20:23aspects.
00:20:24You could say, look at this and remind me about it at 8pm tonight.
00:20:28And then it syncs with the companion app.
00:20:30So I guess it's not replacing the phone, but it's augmenting what I would do with my phone.
00:20:35Like how are these glasses going to be more deeply tied to Med AI over time?
00:20:39Um, yeah, it seems like they're getting closer and closer all the time, but yeah, well, I
00:20:43think Med AI is becoming a more and more prominent feature of the glasses, right?
00:20:49There's more stuff that you can do.
00:20:50So you just mentioned reminders, another example, it's like now that is just going to work and
00:20:54now your glasses can remind you of things.
00:20:56And so, okay.
00:20:57So you can look at a phone number and say, call this phone number and then it falls on
00:21:00the phone.
00:21:01Yeah.
00:21:02So it's just to add more capabilities over time.
00:21:03And some of that are model updates, right?
00:21:04So like, okay, now it has Lama 3.2, but, um, but some of it is, is kind of software development
00:21:10around it.
00:21:11Like reminders you don't get for free just because we updated the model.
00:21:13That's kind of, we have this big software development effort and, um, and, and we're
00:21:18kind of adding features continuously and developing the ecosystem, right?
00:21:21So you get more apps.
00:21:22So Spotify and all these different things kind of, kind of can work more natively.
00:21:27So the glasses just get more and more useful, um, which I think is also going to create
00:21:31increased demand over time.
00:21:33And how does it interact with phones?
00:21:35I mean, like you said, I don't think people are getting rid of phones anytime soon.
00:21:42Um, the way I kind of think about this is that when phones became the primary computing
00:21:48platform, we didn't get rid of computers.
00:21:51We just kind of shifted, right?
00:21:54So I don't know if you have this experience, but at some point in the early 2010s, I noticed
00:22:01that I'd be sitting at my desk in front of my computer and I just pull out my phone to
00:22:05do things.
00:22:06Yeah.
00:22:07And I think what's going to happen, it's not like we're going to throw away our phones,
00:22:11but I think slowly, um, we're just going to start doing more things with our glasses and
00:22:18leaving our phones in our pockets more.
00:22:20And it's not like we're done with our computers and I don't think we're going to be done with
00:22:24our phones for a while.
00:22:25But there's a pretty clear path where, um, where you're just going to use your glasses
00:22:30for more and more things.
00:22:32And over time, I think the glasses are also going to be able to be powered by wrist-based
00:22:37wearables or other wearables.
00:22:38So you're going to wake up one day, 10 years from now, and you're not even going to need
00:22:43to bring your phone with you.
00:22:44Now you're still going to have a phone, but I think like more of the time people are going
00:22:48to leave it in their pocket or leave it in their bag or eventually even some of the time
00:22:51leave it in their, at home.
00:22:53And I think there will just be this sort of gradual shift to glasses becoming the main
00:22:58way that we do computing.
00:22:59It's interesting that we're talking about this right now because I feel like phones
00:23:02are becoming kind of boring and stale.
00:23:05You know, like I, I just was like looking at the new iPhone and it's basically the same
00:23:09as the year before.
00:23:11People are doing foldables, but like, it feels like people have kind of run out of ideas
00:23:14on phones and that they're kind of at their natural end state.
00:23:18And I wonder, you know, when you see something like the Ray-Bans and how people have really
00:23:22gravitated to them in a way that's surprised you guys and I think surprised all of us,
00:23:26but it's also just like people want to interact with technology in different ways now.
00:23:31And I think AI, like you said at the beginning, like the way that AI is intersected with this
00:23:36is just kind of like an aha thing for people that honestly, for me, I didn't expect it
00:23:42to click as quickly as it did, but like when I got whitelisted for the AI, I was like walking
00:23:47around my backyard and like using it and I was like, Oh, like it's obvious now where
00:23:51this is going.
00:23:52So it feels like we're fine.
00:23:53Like I was saying at the beginning, it feels like things are finally, you can see where
00:23:57it's going.
00:23:58Whereas before it's been a lot of like R&D and talking about it, but like these, the
00:24:01Ray-Bans are kind of a signifier of that.
00:24:04And I'm wondering if you agree with that.
00:24:06I agree.
00:24:07I mean, I still think it's early.
00:24:08I think you really want to be able to not just ask the AI questions, but ask it to do
00:24:16things and know that it's going to reliably go do it.
00:24:22And we're starting with simple things, right?
00:24:24So voice control of your glasses, although you can do that on phones too.
00:24:28Things like reminders, although you can generally do that on phones too.
00:24:31But I think as the model capabilities grow over the next couple of generations and you
00:24:37get more of what people call these agentic capabilities, I think it's going to start
00:24:43to get pretty exciting.
00:24:44For what it's worth, I also think that all the AI work is going to make phones a lot
00:24:48more exciting.
00:24:51It's the most exciting thing I think that has happened to our family of apps roadmap
00:24:56in a long time is all of the different AI things that we're building.
00:24:59So if I were trying to design what the next few versions of iPhone or Google's phone should
00:25:05be, I mean, I think that there's a long and interesting roadmap of things that they can
00:25:10go do with AI that as an app developer, we can't.
00:25:15So I think that that's a pretty exciting and interesting thing for them to go do, which
00:25:18I assume that they will.
00:25:19On the AI social media piece, one of the wilder things that your team told me you guys are
00:25:24going to start doing is showing people AI generated imagery personalized to them in
00:25:29feed.
00:25:30I think it's starting as an experiment, but if you're a photographer, you would see
00:25:34meta AI generating content that's maybe personalized for you alongside content from the people
00:25:40you follow.
00:25:41Yeah.
00:25:42So that's the idea that I've been thinking about of AI and kind of invading social media,
00:25:46so to speak.
00:25:47Maybe you don't like the word invading, but you know what I mean?
00:25:50And what does that do to how we relate to each other as humans?
00:25:56Like how much AI stuff and AI generated stuff is going to be filling feeds in the near future
00:26:02in your view?
00:26:03Well, here's how I come at this.
00:26:05So in the history of running the company where we've been building these apps for 20 years,
00:26:11we call it three to five years.
00:26:14There's some new major format that comes along that is typically additive to the experience.
00:26:21So initially people kind of updated their profiles, then they were able to post statuses
00:26:26that were texts, then links, then you got photos early on, then you added videos.
00:26:35Then with mobile, basically Snap invented stories, the first version of that, and that
00:26:40became a pretty kind of widely used format.
00:26:46The whole version of short form videos, I think, is sort of a still ascendant format.
00:26:52But you basically, at each step along the way, you want to, right, it's like you keep
00:27:02on making the system richer by having more different types of content that people can
00:27:08share and in different ways to express themselves.
00:27:10And when you look out for the next 10 years of, okay, if this trend seems to happen where
00:27:15every three, five years, whatever the pace is, that there are new formats.
00:27:23I think given the pace of change in the tech industry, I think you'd bet that that continues
00:27:26or accelerates.
00:27:27And I think you'd bet that probably most of the new formats are going to be kind of AI
00:27:34connected in some way, given that that's the kind of driving theme for the industry at
00:27:37this point.
00:27:39Given that kind of set of assumptions, we're sort of trying to understand what are the
00:27:46things that are most useful to people within that.
00:27:49There's one vein of this, which is helping people and creators make better content using
00:27:55AI, right?
00:27:56So that's, that I think is gonna be pretty clear, right?
00:27:59Just make it like super easy for like aspiring creators or advanced creators to make much
00:28:05better stuff than they would be able to otherwise.
00:28:10That can take the format of like, all right, like my daughter is writing a book and she
00:28:16wants it illustrated.
00:28:17And like we sit down together and work with meta AI and imagine to help her come up with
00:28:22images to illustrate it.
00:28:24It's like, okay, that's like a thing that's like, she didn't have the capability to do
00:28:27that before.
00:28:28She's not a graphic designer, but now she kind of has that, that, that, that ability.
00:28:31I think that that's going to be pretty cool.
00:28:34Then I think that there's a version where you have just this great diversity of AI agents
00:28:42that are as part of this system.
00:28:43And this I think is a big difference between our vision of AI and most of the other companies
00:28:47is, yeah, we're building meta AI as, as kind of the main assistant that you can build.
00:28:52That's sort of equivalent to, you know, the singular assistant that may be like a Google
00:28:55or an open AI or different folks are building, but it's not really the main thing that we're
00:29:01doing.
00:29:02Our main vision is that we think that there are going to be a lot of these, right?
00:29:06It's every business, all the hundreds of millions of small businesses, you know, just like they
00:29:12have a website and an email address and a social media account today, I think that they're
00:29:17all going to have an AI that helps them interact with their customers in the future that does
00:29:21some combination of sales and customer support and all that.
00:29:24I think all the creators are basically going to want some version of this that basically
00:29:29helps them interact with their community when they're just limited by, they don't have
00:29:32enough hours in the day to interact with all the messages that are coming in and they want
00:29:37to make sure that they can show some love to people in their community.
00:29:42And those I think are just the two most obvious ones that even if we just did those, that's
00:29:45many hundreds of millions, but then there's going to be all this more creative stuff that's
00:29:50UGC that people create for better kind of wilder use cases that they want.
00:29:54And our view is, okay, these are all going to like live across these social networks
00:29:57and beyond.
00:29:58And I don't think that they should just be constrained to waiting until someone messages
00:30:06them.
00:30:07Right?
00:30:08I think that they're going to have their own profiles.
00:30:09They're going to be creating content.
00:30:11People will be able to follow them if they want.
00:30:14You'll be able to comment on their stuff.
00:30:16They may be able to comment on your stuff if you're connected with them.
00:30:18I mean, there will obviously be different logic and rules, but that's one way that there's
00:30:22going to be just a lot more kind of AI participants in the kind of broader social construct that
00:30:28we have.
00:30:30And then I think you get to the test that you mentioned, which is maybe the most abstract,
00:30:34which is just having kind of the central meta AI system directly generate content for you
00:30:44based on what we think is going to be interesting to you and putting that in your feed.
00:30:48And on that, you know, I think that there's this been, there's been this trend over time
00:30:52where the feed started off as primarily and exclusively content for people you, you followed
00:30:59friends.
00:31:00I guess it was friends early on, then it kind of broadened out to, okay, you followed a
00:31:06set of friends and creators, and then it got to a point where the algorithm was good enough
00:31:11where we're actually showing you a lot of stuff that you're not following directly because
00:31:16in some ways that's a better way to show you more interesting stuff than only constraining
00:31:19it to things that you've chosen to follow.
00:31:21And I think the next logical jump on that is like, okay, we're showing you content from
00:31:26your friends and creators that you're following and creators that you're not following that
00:31:30are generating interesting things.
00:31:31And you just add onto that a layer of, okay, and we're also going to show you content from,
00:31:37that's generated by, by an AI system that might be something that you're interested
00:31:40in.
00:31:41Now, how big do any of these segments get?
00:31:44I think it's really hard to know until you build them out over time, but it feels like
00:31:48it is a category in the world that's going to exist.
00:31:51And how big it gets is kind of dependent on the execution and how good it is.
00:31:55Why do you think it needs to exist as a new category?
00:31:57I'm still wrestling with why people want this.
00:32:01I get the companionship stuff that character AI and some startups have already shown there's
00:32:05like a market for, and you've talked about how meta is already being used for role playing.
00:32:11But the big idea is that AI has been used to intermediate and feed how humans reach
00:32:17each other.
00:32:18And now all of a sudden, AIs are going to be in feeds with us.
00:32:22Well, I think the main difference, and that feels big.
00:32:25Yeah.
00:32:26But in a lot of ways, the big change already happened, which is people getting content
00:32:30that they weren't following.
00:32:32And the definition of feeds and social interaction has changed very fundamentally in the last
00:32:3610 years.
00:32:38Now in social systems, most of the direct interaction is happening in more private forums
00:32:45and messaging or groups.
00:32:47And this is one of the reasons I think why we were late with reels initially to competing
00:32:51with TikTok is because we hadn't made this mental shift where we kind of felt like no
00:32:55feed is where you interact with people.
00:32:58Actually increasingly, feed is becoming a place where you discover content that you
00:33:02then take to your private forums and interact with people there.
00:33:06So a lot of the way that I interact with people, it's like, yeah, I'll still have the thing
00:33:10where a friend will post something and I'll comment on it and engage directly in feed.
00:33:16Again, this is additive, you're adding more over time.
00:33:21But the main way that you engage with reels isn't necessarily that you go into the reels
00:33:26comments and comment and talk to people you don't know.
00:33:29It's like you see something funny and you send it to friends in a group chat.
00:33:32Right.
00:33:32And I think that that paradigm will absolutely continue with AI and all kinds of interesting
00:33:38content.
00:33:39So it is facilitating connections with people.
00:33:44But I think already we're in this mode where our connections through social media are shifting
00:33:49to more private places.
00:33:52And the role of feed in the ecosystem is more as a, you know, I call it a discovery engine
00:33:57of content to kind of, you know, icebreakers or interesting kind of topic starters for
00:34:02the conversations that you're having across this broader spectrum of places where you're
00:34:06interacting.
00:34:07Do you worry about people interacting with AIs like this, making people less likely to
00:34:13talk to other people, like reducing the engagement that we have with humans?
00:34:18The sociology that I've seen on this is that most people have way fewer friends physically
00:34:24than they would like to have.
00:34:25I think people cherish the kind of human connection that they have.
00:34:29And the more we can do to make that feel more real and give you more reasons to connect,
00:34:33whether it's through something funny that shows up so you can message someone or a pair
00:34:36of glasses that lets like your sister show up as a hologram in your living room when
00:34:41you, you know, she lives across the country and you wouldn't be able to see her otherwise.
00:34:44Like that's always kind of our main bread and butter and the thing that we're doing.
00:34:48But, you know, in addition to that, I mean, if the, you know, the average person, I think,
00:34:54you know, maybe they'd like to have 10 friends.
00:34:57And I mean, there's the stat that it's like, it's sort of sad.
00:35:00But I think the average American feels like they have fewer than three, like real kind
00:35:05of close friends.
00:35:06So does this take away from that?
00:35:09My guess is no.
00:35:10I think that what's going to happen is it's going to help give people more of the support
00:35:17that they need and give people more kind of reasons and ability to connect with either
00:35:22a broader range of people or more deeply with the people that they care about.
00:35:26How are you feeling about how Threads is doing these days?
00:35:29I mean, Threads is on fire.
00:35:31It's great.
00:35:31I mean, it's, I mean, these things, it's like, there's only so quickly that something can
00:35:34get to a billion people.
00:35:35So it's going to, you know, we'll kind of keep on pushing on it over.
00:35:38I've heard it's still using Instagram a lot for growth.
00:35:40Like, I guess I'm wondering when you see it getting to like a standalone growth driver
00:35:47on its own.
00:35:48I mean, I think that these things all connect to each other.
00:35:50So I don't, I mean, I think Threads helps Instagram.
00:35:53I think Instagram helps Threads.
00:35:54I don't, I don't know that we have some strategic goal, which is like, make it so that Threads
00:35:59is completely disconnected from Instagram or Facebook.
00:36:01I actually think we're going the other direction.
00:36:02It started off just connected to Instagram, and now we also connected it so that the content
00:36:05can show up.
00:36:06You know, taking a step back.
00:36:07I mean, we just talked about how for most people, they're interacting in more private
00:36:11forums.
00:36:11If you're a creator, what you want to do is have your content show up everywhere, right?
00:36:15Because you're trying to build the biggest community that you can in these different
00:36:18places.
00:36:18So it's this huge value for people if they can generate a reel or a video or some text-based
00:36:24content, and now you can post it in Threads, Instagram, Facebook, and more places over
00:36:30time.
00:36:30So I think the direction there is generally kind of more flow, not less, and kind of more
00:36:37interoperability.
00:36:39And that's why I've been kind of pushing on that as a theme over time.
00:36:43I'm not even sure what X is anymore, but I think what it used to be and what Twitter
00:36:47used to be was a place where you went when news was happening.
00:36:50I know you and the company seem to be distancing yourself from recommending news, but with
00:36:55Threads, it feels like that's what people want and people thought Threads might be,
00:37:01but it seems like you all are intentionally saying, we don't want Threads to actually
00:37:05be that.
00:37:05Yeah, I mean, it's interesting.
00:37:06There are different ways to look at this.
00:37:09I always looked at Twitter not as primarily about real-time news, but as a kind of short
00:37:15form, primarily text, discussion-oriented app.
00:37:18To me, the fundamental defining aspect of that format is that when you make a post,
00:37:25the comments aren't subordinate to the post.
00:37:28The comments are kind of at a peer level, and that is a very different architecture
00:37:33than every other type of social network that's out there.
00:37:38And it's a subtle difference, but within these systems, these subtle differences lead to
00:37:44very different emerging behaviors.
00:37:46So because of that, people can take, they can fork discussions, and it makes it a very
00:37:50good discussion-oriented platform.
00:37:52News is one thing that people like discussing, but it's not the only thing.
00:37:56I always looked at Twitter and I was like, hey, this is such a wasted opportunity.
00:38:01This is clearly a billion-person app.
00:38:05Maybe in the modern day, when you have many billions of people using social apps, it should
00:38:09be multiple billions of people.
00:38:11And for whatever reason, I mean, there are a lot of things that have been complicated
00:38:15about Twitter and the corporate structure and all that, but they just weren't quite
00:38:19getting there.
00:38:19And eventually, I kind of thought, hey, I think we can do this, right?
00:38:25I think we can get this, build out the discussion platform in a way that can get to a billion
00:38:31people and be more of a ubiquitous social platform that I think achieves the fullness
00:38:37of its full potential.
00:38:40But our version of this is we want it to be a kinder place, right?
00:38:44We don't want it to start with the direct head-to-head combat of news, and especially
00:38:51politics.
00:38:54Do you feel like that constrains the growth of the product at all?
00:38:57I mean, I think we'll see.
00:38:57We'll run the experiment.
00:38:59That needs to exist in the world, because I feel like with X's seeming implosion, it's
00:39:04not really existing anymore.
00:39:06Maybe I'm biased to someone in the media, but I do think people want, when something
00:39:10big happens in the world, they want an app that they can go to and see everyone that
00:39:13they follow talking about it immediately.
00:39:15Yeah.
00:39:16Well, we're not the only company.
00:39:20There are a ton of different competitors and different companies doing things.
00:39:23And I think that there's a talented team over at Twitter and X, and so I wouldn't write
00:39:30them off.
00:39:32And then, obviously, there's all these other folks.
00:39:34There's a lot of startups that are doing stuff.
00:39:36So I don't feel like we have to go at that first.
00:39:42I think that maybe we get there over time, or maybe we decide that it's enough of a zero
00:39:47sum trade, or maybe even a negative sum trade, where that use case should exist somewhere.
00:39:52But maybe that use case prevents a lot more usage and a lot more value in other places,
00:39:59because it makes it a somewhat less friendly place.
00:40:01And I don't think we know the answer to that yet.
00:40:06But I do think the last 10 years, 8 years of our experience has been that the political
00:40:13discourse is tricky.
00:40:19On the one hand, it's obviously a very important thing in society.
00:40:23And on the other hand, I don't think it leaves people feeling good.
00:40:26So I'm torn between these two values.
00:40:29On the one hand, I think people should be able to have this kind of open discourse,
00:40:33and that's good.
00:40:34On the other hand, I don't want to design a product that makes people angry.
00:40:39There's an informational lens for looking at this.
00:40:42And you're designing a product, and what's the feel of the product?
00:40:46I think anyone who's designing a product cares a lot about how the thing feels.
00:40:52But you recognize the importance of that discussion happening in the world.
00:40:55I think it's useful.
00:40:56And look, we don't block it.
00:40:58We just make it so that for the content where you're following people, if you want to talk
00:41:04to your friends about it, if you want to talk to them about it in messaging, there can be
00:41:07groups about it.
00:41:08If you follow people, it can show up in your feed.
00:41:10But we don't go out of our way to recommend that content when you're not following it.
00:41:15And I think that that has been a healthy balance for us and for getting our products to generally
00:41:22feel the way that we want.
00:41:24And culture changes over time.
00:41:26Maybe the stuff will be a little bit less polarized and anger-inducing at some point.
00:41:31And maybe it'll be possible to have more of that while also at the same time having a
00:41:36product where we're proud of how it feels.
00:41:39But until then, I think we want to design a product that, yeah, people can get the things
00:41:46that they want.
00:41:47But fundamentally, I care a lot about how people feel coming away from the products.
00:41:55Do you see this decision to downrank political content for people who aren't being followed
00:42:02in feed as a political decision, I guess?
00:42:05Because, I don't know, you're also at the same time not really saying much about the
00:42:11election this year.
00:42:11You're not donating.
00:42:12You've said you kind of want to stay out of it now.
00:42:14And I see the way the company is acting, and it reflects your personal way you're operating
00:42:20right now.
00:42:21And I'm wondering, how much more of it is also about what you and the company have gone
00:42:25through and the political environment, and not necessarily just what users are telling
00:42:30you?
00:42:30Sure.
00:42:33Is there a through line there?
00:42:35I'm sure it's all connected.
00:42:36I think in this case, it wasn't a trade-off between those two things, because this actually
00:42:40was what our community was telling us.
00:42:41And people were saying, generally, we don't want so much politics.
00:42:47We don't feel good.
00:42:50We want content.
00:42:51We want more stuff from our friends and family.
00:42:52We want more stuff from our interests.
00:42:56So that was kind of the primary driver.
00:42:58But I think it's definitely the case that our corporate experience on this shaped this.
00:43:05And there's a big difference between something being political and being partisan.
00:43:09And the main thing that I care about is making sure that we can be seen as a nonpartisan,
00:43:17and as much as something can in the world in 2024, be sort of like a trusted institution
00:43:22by as many people as possible.
00:43:24And I just think that the partisan politics is so tough in the world right now that I've
00:43:33made the decision that I kind of feel like for me and for the company, the best thing
00:43:37to do is to try to be as nonpartisan as possible in all of this and kind of be kind of as neutral
00:43:44and distance ourselves as much as possible.
00:43:46And it's not just the substance.
00:43:47I also think the perception matters.
00:43:50So that's why maybe it doesn't matter on our platforms whether I endorse a candidate or
00:43:55not, but I don't want to go anywhere near that.
00:43:58And yeah, sure.
00:43:59I mean, you could say that's a political strategy.
00:44:02Um, but I think it's, um, I think for where we are in the world today, it's very hard.
00:44:09Almost every institution has become partisan in some way.
00:44:12And we are just trying to resist that.
00:44:15And maybe I'm too naive, and maybe that's impossible, but that's, we're going to try
00:44:20to do that.
00:44:21On the Acquired podcast recently, you said that the political miscalculation was a 20-year
00:44:26mistake.
00:44:26Yeah, from a brand perspective.
00:44:28And that it was going to take another 10 or so for you to fully work through that cycle.
00:44:33Yeah, yeah.
00:44:34What makes you think it's such a lasting thing?
00:44:37Because you look at, like, how you personally have kind of evolved over the last couple
00:44:40of years, and I think perception of the company has evolved.
00:44:42And I'm wondering, like, what you meant by saying it's going to take another 10 years.
00:44:46I'm just talking about where our brand is and our reputation are compared to where I
00:44:52think they would have been.
00:44:54I mean, there's no doubt that even now here, and okay, yeah, sure.
00:44:57Maybe things have improved somewhat over the last few years.
00:45:00You can feel the trend.
00:45:01But it's still significantly worse than it was in 2016.
00:45:05You know, I mean, the internet industry overall, and I think our company in particular, just
00:45:13were seen way more positively.
00:45:16And, you know, look, there were real issues, right?
00:45:19So I think that it's always very difficult to talk about this stuff in a nuanced way,
00:45:23because I think to some degree before 2016, everyone was sort of too rosy about the internet
00:45:30overall and didn't talk enough about the issues.
00:45:31And then the pendulum sort of swung, and people only talked about the issues and didn't talk
00:45:35about the stuff that was positive.
00:45:36And it was all both there the whole time.
00:45:39But so when I talk about this, I don't mean to, you know, come across as simplistic or,
00:45:47you know, like, you guys didn't do anything wrong.
00:45:49Or that there weren't issues with the internet or things like that.
00:45:51I mean, obviously, every year, you know, whether it's politics or other things, there
00:45:55are always things that you look back at.
00:45:56And you're like, hey, yeah, like, if I were playing this perfectly, I would have done
00:45:59these things differently.
00:46:00And but I do think it's the case that I didn't really know how to react to something as big
00:46:08of sort of a shift in the world as what happened.
00:46:11And it took me a while to find my footing.
00:46:13And I do think that it's tricky when you're caught up in these kind of big debates.
00:46:18And you're not kind of experienced or sophisticated in engaging with that.
00:46:23I think you can make some big missteps.
00:46:26And I do think that some of the things that we were accused of over time, it just, you
00:46:33know, I think it's just been pretty clear at this point, you know, now that all the
00:46:36investigations have been done, that like, they weren't true.
00:46:41And you're talking about like Cambridge Analytica.
00:46:43I think Cambridge Analytica is a good example of something that it's like, people thought
00:46:47that like, all this data had been taken, and that it had been used in this campaign.
00:46:52And it turns out that it wasn't.
00:46:54And like, yeah, so it's like all this stuff.
00:46:56Okay.
00:46:56And like, the data wasn't even, you know, accessible to the developer.
00:47:00And then we'd fix the issue like five years ago.
00:47:01So but in the moment, it was like really hard for us to kind of have a rational discussion
00:47:08about that.
00:47:08And part of the challenge is that, you know, for the general population, I think a lot
00:47:15of people, they read the initial headlines, and they don't necessarily read the, you know,
00:47:21and frankly, you know, a lot of the media, I don't think was like as loud to write about
00:47:26when all the investigations concluded that said that there weren't that like, a lot of
00:47:30the initial allegations were just completely wrong.
00:47:33So I think that's like a, that's a real thing.
00:47:36So okay, so you take these hits.
00:47:37And I didn't really know how to, how to kind of push back on that.
00:47:41And maybe some of it, you can't, but I like to think that I think we could have played
00:47:47this some of the stuff differently.
00:47:50And I do think it was certainly the case that when you take responsibility for things that
00:47:54are not your fault, you become sort of a weak target for people who are looking to blame
00:48:02other things and find a target for them.
00:48:04And it's sort of like, this is a different part of it's somewhat related to this.
00:48:09But when you think about like, like litigation strategy for the company, one of the reason
00:48:13why I hate settling lawsuits is that it, it basically sends a signal to people that, hey,
00:48:19this is a company that settles lawsuits.
00:48:21So maybe like, we can sue them and they'll settle lawsuits.
00:48:24So you wouldn't write a blank check to the government like Google did for its antitrust
00:48:28case.
00:48:28I think like, I think the right kind of way to approach this is when you believe in something,
00:48:34you fight really hard for it.
00:48:36And I think this is a repeat game.
00:48:38This isn't like this, it's not like there's a single issue and we're going to be around
00:48:42for a long time.
00:48:43And I think it's, it's really important that people know that we're a company that has
00:48:47conviction and that we, we believe in what we're doing and we're going to back that up
00:48:51and defend ourselves.
00:48:52And I think that that kind of sets the right tone.
00:48:56Now, I think over the next 10 years, I think we're sort of digging ourselves back to neutral
00:49:01on this.
00:49:02And, but I like to think that if we hadn't had a lot of these issues, we would have made
00:49:06progress over the last 10 years too.
00:49:07So I sort of give it this timeframe, maybe 20 years is too long, maybe it's 15, but it's
00:49:11hard to know with politics.
00:49:12It feels like mental health and youth mental health may be the next wave of this.
00:49:16That I think is the next big fight.
00:49:17And, and on that, you know, I think a lot of the data on this, I think is just not where
00:49:22the narrative is.
00:49:23The narrative.
00:49:24Yeah.
00:49:24I think the narrative is a lot of people sort of take it as, um, as if it's like an assumed
00:49:30thing that there is some link.
00:49:32And like, I think the majority of the, of the high quality research that's out there
00:49:37suggests that there's no causal connection at like a kind of a broad scale between, between
00:49:41these things.
00:49:42So, um, no, look, I mean, I think that that's different from saying like in any given issue,
00:49:46like was, is someone bullied?
00:49:47Should we try to stop bullying?
00:49:48Yeah, of course.
00:49:49Um, but yeah, overall, um, I, I think that this is, this is one where there, there are
00:49:58a bunch of these cases.
00:49:59I think that there will be a lot of litigation around them.
00:50:02Uh, and it's one where we're trying to make sure that the academic research that, that
00:50:07shows something that, that I think is, you know, it, um, you know, to me, it's sort of
00:50:13foots more with what I've seen of how the platforms operate, but it's counter to what
00:50:18a lot of people think.
00:50:19And I think that that's going to be a reckoning that we'll have to have is, is basically when
00:50:23as the, the kind of the majority of the high quality academic research, um, kind of gets
00:50:29shown is like, okay, can people accept this?
00:50:31And I think that's going to be a really important, um, set of debates over the next few years.
00:50:35At the same time, you guys have acknowledged there's affordances in the product, like the
00:50:38teen rollout with Instagram recently that you can make to make the product a better
00:50:42experience for young people.
00:50:43Yeah.
00:50:43And I think this is an interesting part of the balance is, um, you can play a role in
00:50:49trying to make something better, even if the thing wasn't caused by you in the first
00:50:54place, like, there's no doubt that being a parent is really hard.
00:50:59And there's a big question of in this internet age where we have phones, um, what are the
00:51:09right tools that parents need in order to be able to raise their kids?
00:51:15And like, I think that we can play a role in giving people controls over, um, parental
00:51:21controls over the apps.
00:51:22I think the parental controls are also really important because parents have different ways
00:51:25that they want to raise their kids, or just like schooling and education.
00:51:29People have like very significantly different local preferences for how they want to raise
00:51:33their kids.
00:51:34I don't think that most people want some internet company setting all the rules for this either.
00:51:40So obviously, you know, when there are laws passed, we'll kind of follow the government's
00:51:45direction and the laws on that.
00:51:47But I actually think the right approach for us is to primarily kind of align with parents
00:51:54to give them the tools that they want to be able to raise their kids in the way that they
00:51:57want.
00:51:58And some people are going to think that more technology use is good.
00:52:01That's sort of how my parents, you know, raised me growing up.
00:52:05Um, I think it worked pretty well.
00:52:07Um, some people are going to be kind of more, more, um, you know, want to limit it more.
00:52:12And we want to give them the tools to be able to do that.
00:52:14But, but I don't think that this is, you know, primarily or only a social media thing.
00:52:19Um, even, even the parts of this that are technology, age verification, I think phones,
00:52:24like the phone platforms have a huge part of this.
00:52:26I mean, it's, um, yeah, there's this big question of how do you do age verification?
00:52:31And I mean, I can tell you what the easiest way is, which is like, all right, like every
00:52:35time you go do a payment on your phone, I mean, there already is child.
00:52:37Um, you know, basically like child age verification.
00:52:40So I don't really understand.
00:52:42Well, I guess I understand, but I think it's, it's, it's not very, um, you know, excusable
00:52:48from my perspective, why Apple and, and, and, um, I guess to some extent, Google don't want
00:52:53to just extend the age verification that they already have on their phones to be a parental
00:52:58control for parents to, to basically be able to say what, um, you know, what apps can my
00:53:05kids use, right?
00:53:06It's hard for me to not see the logic in it either.
00:53:08I don't, I don't really think they don't want to take responsibility,
00:53:12but maybe that's on Congress then to pass who has to take responsibility.
00:53:15Yeah.
00:53:16Yeah.
00:53:17So, and, and, and we're going to do our part and we're going to, we're going to build the
00:53:20tools that we can for, for parents and for, and for teens.
00:53:24Um, but at the end of the day, um, I, and look, I'm not saying it's all the phone's
00:53:31fault either.
00:53:32Although I would say that like the ability to get push notifications and get kind of
00:53:36distracted is from my perspective, seems like a much greater contributor to mental health
00:53:41issues than, um, than, than kind of a lot of the specific apps.
00:53:44But, but there are things that, that I think everyone should, should kind of try to improve
00:53:47and work on.
00:53:49Um, but yeah, I mean, I, that, that's sort of, that's sort of my view on all that.
00:53:55On the regulation piece, as it relates to AI, you've been very vocal about what's happening
00:54:00in the EU.
00:54:01And, uh, you recently signed an open letter and I believe it was saying basically that
00:54:06you guys just don't have clarity on consent for training, how it's supposed to work.
00:54:12Um, and I'm wondering what you think needs to happen there for things to move forward.
00:54:17Cause like med AI is not available in Europe, new Lama models are not, um, is that something
00:54:23you see getting resolved, uh, at all, I guess?
00:54:27And what would it take?
00:54:29Yeah, I don't know.
00:54:30It's, it's a little hard for me to parse the, the European politics.
00:54:36Um, I have a hard time enough with American politics.
00:54:39I mean, it's, and I'm, I'm American, but, um, so in theory, my understanding of the
00:54:46way this is supposed to work is they, they kind of passed this GDPR regulation and you're
00:54:53supposed to have this, this idea of sort of a one-stop shop, like home regulator who
00:55:01can basically on behalf of the whole EU interpret and enforce the rules.
00:55:06And we have our European headquarters and we work with that regulator and, you know,
00:55:12I think they're like, okay, they're pretty tough on us and pretty firm, but at least
00:55:16when you're working with one regulator, you can kind of understand how are they thinking
00:55:20about things and it's, and you can make progress.
00:55:22And the thing that I think has been tricky is there have been, from, from my perspective,
00:55:28a little bit of a backslide where now you get all these other, um, DPAs across the continent
00:55:34sort of also intervening and trying to do things.
00:55:37And it just seems like more of an kind of internal EU political thing, which is like,
00:55:42okay, do they want to have this one-stop shop and have clarity for companies so companies
00:55:45can act, it can kind of like can execute or, or do they just want it to be this kind of
00:55:50very complicated regulatory system?
00:55:52And, um, look, I think that's for them to sort out.
00:55:56Um, but, but I think that there's no doubt that when you have like dozens of different
00:56:00regulators that can ask you kind of the same questions about different things, it makes
00:56:05it a much more difficult environment to build things.
00:56:07So I don't think that that's just us.
00:56:08I think that that's all the company.
00:56:09But do you understand the concern people have about training data and creators even?
00:56:14And this idea that their data is being used for these models, they're not getting compensated.
00:56:20And the models are creating a lot of value.
00:56:22And I know you're giving away Lama, but you're still, you've got mid AI.
00:56:26And I understand the frustration that people have about that.
00:56:29I think it's a naturally bad feeling to be like, oh, my, my data is now being used in
00:56:33a new way that I have no control or compensation over.
00:56:35Do you, do you sympathize with that?
00:56:37Yeah.
00:56:38I mean, I think that there in any new medium and technology, there's like the concepts
00:56:43around fair use and like, where are the boundaries between what you have control over, right?
00:56:50When you put something out in the world, to what degree do you still get to control
00:56:53it and kind of own it and license it?
00:56:57And I think that all these things are basically going to get need to get, you know, re-litigated
00:57:03and re-discussed in the, in the, in the AI era.
00:57:08So, I mean, I get it.
00:57:09I think that these are, these are important questions.
00:57:11I think this is not like a completely novel thing to AI in the grand scheme of things.
00:57:16I think it was a lot of them, there were questions about it with, with the internet
00:57:19overall too, and with different technologies over time.
00:57:24And, but I think getting to clarity on that is going to be important.
00:57:27So that way, the things that society wants people to build, they can go build.
00:57:33What does clarity look like to you there?
00:57:35I mean, I think it starts with having some framework of like, okay, what's the process
00:57:40going to be for working through that?
00:57:42But you don't see a scenario where creators get like directly compensated for the use
00:57:47of their content models?
00:57:49I think that there's a lot of different possibilities for how stuff goes in the future.
00:57:52Now, I do think that there's this issue, which is a lot of like, well, psychologically,
00:57:56I understand what you're saying.
00:57:59I think individual creators or publishers tend to overestimate the value of their specific
00:58:07content, right?
00:58:08So it's like, okay, maybe like in, in, in like in the grand scheme of this, right?
00:58:13Um, so we have this set of challenges with, with, um, you know, news publishers around
00:58:19the world, which is like, okay, they, they're like, a lot of folks are constantly kind of
00:58:23asking to be paid for the content.
00:58:25And on the other hand, we have our community, which is asking us to show less news because
00:58:28it makes them feel bad.
00:58:29Right.
00:58:30And I mean, we talked about that.
00:58:31So it's like, there's this issue, which is okay.
00:58:33It's like, actually, we're showing some amount of the news that we're showing because we
00:58:37think it's socially important against what our community wants.
00:58:41Like if we were actually just following our, our, our, what our, what our community wants,
00:58:44we'd show even less than we're showing.
00:58:46And you see that in the data that people just don't like to engage with this stuff.
00:58:49And we've had these issues where sometimes we like publishers say, okay, if you're not
00:58:54going to pay us, then pull our content down.
00:58:56And it's just like, yeah, sure.
00:58:57Fine.
00:58:57Pull, pull your content down.
00:58:58I mean, that sucks.
00:58:59I'd rather people be able to share it.
00:59:01But, but, but I think some, to some degree, some of these questions have to get tested
00:59:09by their negotiations and they have to get tested by people walking.
00:59:14And, and like, and then at the end, once people walk, you figure out where the value really is.
00:59:22Right.
00:59:22If it, if it really is the case that that news was a big thing that the community wanted,
00:59:27then I mean, look, we're, we're a big company.
00:59:30We could, you know, we could probably, you know, we pay for content when it's valuable
00:59:34to people.
00:59:34We're just not going to pay for content when it's not valuable to people.
00:59:37So I think that you'll probably see a similar dynamic with AI, which is, my guess is that
00:59:45there are going to be certain partnerships that get made when content is really important and
00:59:48valuable.
00:59:50And I guess that there's probably a lot of people who are kind of have a concern about
00:59:57like the feel of it, like you're saying, but then when push comes to shove, if like, if they
01:00:03demanded that we take, that we don't use their content, then we just wouldn't use their content.
01:00:07And it's not like, you know, that's going to change the outcome of the stuff that much.
01:00:11I guess to bring this full circle where we started, as you're building augmented reality
01:00:15glasses and what you've learned over just the societal implications of the stuff you've built
01:00:19over the last decade, how are you thinking about this as it relates to glasses at scale?
01:00:25Because you're literally going to be augmented, augmenting reality, which is a, it's a
01:00:29responsibility.
01:00:30I think that's going to be another platform too.
01:00:32And you're going to have a lot of these questions as well.
01:00:34I mean, I think the interesting thing about holograms and augmented reality is it's going
01:00:38to be this intermingling of the physical and digital much more than we've had in other
01:00:43platforms.
01:00:45Where on your phone, it's like, okay, yeah, we live a primarily physical world, but then
01:00:49you have the small window into this digital world.
01:00:51And I think we're going to basically have this world in the future that is increasingly,
01:00:58call it half physical, half digital, or I don't know, 60% physical, 40% digital.
01:01:02And it's going to be blended together.
01:01:06And I think that there are going to be a lot of interesting governance questions around
01:01:10that, right?
01:01:11In terms of like, is kind of all of the digital stuff that's overlaid physically going to
01:01:18fit within sort of a physical kind of national regulation perspective?
01:01:25Or is it sort of, is it actually coming from a different world or something?
01:01:29And I think these will all be very interesting questions that we will have a perspective.
01:01:35I'm sure we're not going to be right about every single thing.
01:01:39I think like the world will kind of need to sort out where it wants to be.
01:01:42Different countries will have different values and take somewhat different approaches on
01:01:45things.
01:01:45And I think that it's part of the interesting process of this, the tapestry of how it all
01:01:50gets built is you need to work through so that it ends up being positive for as many
01:01:57of the possible stakeholders as possible.
01:01:59More to come.
01:02:00Yeah.
01:02:01A lot to come.
01:02:02Thanks, Mark.

Recommended