Snap’s Evan Spiegel on AI, AR, and – yes – TikTok 57:43

On September 28th, the Global Citizen Festival will gather thousands of people who took action to end extreme poverty. Join Post Malone, Doja Cat, Lisa, Jelly Roll, and Raul Alejandro as they take the stage with world leaders and activists to defeat poverty, defend the planet, and demand equity. Download the Global Citizen app today and earn your spot at the festival. Learn more at globalcitizen.org.com.

On September 28th, the Global Citizen Festival will gather thousands of people who took action to end extreme poverty. Join Post Malone, Doja Cat, Lisa, Jelly Roll, and Raul Alejandro as they take the stage with world leaders and activists to defeat poverty, defend the planet, and demand equity. Download the Global Citizen app today and earn your spot at the festival. Learn more at globalcitizen.org slash bots. It's on!

Hi, everyone from New York Magazine and the Vox Media Podcast Network. This is Donald Trump with 100% less Ron DeSantis and 100% more legal problems. Just kidding. This is On with Kara Swisher and I'm Kara Swisher. And I'm Naima Razabai.

Bob Iger is going nuclear on Ron DeSantis. I think that Donald Trump has to send him some flowers. I think he's doing appropriate things. I mean, Ron DeSantis is abusing his office. There's a joke that Disney is lawyers that run a theme park, and that's what they are with IP and all kinds of things. These people don't come to

Yeah, I wasn't saying nuclear is in outsized. I just meant he's pulling out the big guns because Walt Disney has sued, obviously, Florida Governor Ron DeSantis, alleging that the Republican governor has waged a, quote, relentless campaign to weaponize government power. And the suit alleges that DeSantis himself has orchestrated the effort. Super interesting to see this showdown between two people, one of whom is a Republican.

I would just say, as my opinion, seems more presidential than the other. Yes, and that would be Bob Iger, who there was a word that he was going to think about it. And of course, he didn't end up doing it. But he's certainly so far and above Ron DeSantis. The issue with Ron DeSantis, which makes everybody realize is he didn't anticipate this. This is just shoot from hip. It just, you know, it doesn't work. Going after Disney and beer,

I mean, I expect Dolly Parton next. I mean, what in the world is he thinking? There are much better targets, but this one is not one of them, especially since they're the biggest employer in Florida. And for him to do this just because they disagree with him on his stupid don't say gay bills is just, it's really chilling. And I'm glad Disney is doing this. Bob Iger, he told us why he didn't run for president. Do you remember that? No, I don't. He said he asked his wife and I loved his wife's response. She said, you

You can have any job you want, Bob, but not with this wife. Yeah, exactly. Yeah, she's great. Willow Bay, she was on the air and she's now an academic at USC. She's a dean, actually. And so, you know, she's got her life. She doesn't want to run for president. He's like, I don't want to be Michelle Obama complaining about this inside the gates. Do you think he's going to run? You asked him if he was going to run. He said no again. No, no.

100%. No, he's got more power than, he's got tons of power as CEO of Disney. So it's fine. He's got his hands full too. But there's no power like the presidency. No, there is. The head of Disney, there is. Last time we saw Bob Iger in California, he talked about Disney's once upon a time idea to buy Twitter.

which would be so different than Elon Musk's Twitter today. And then I think we were talking about how Snap would be a better fit for Disney than Twitter. Yeah, 100%. It still is, actually, in a weird way, if they really wanted to go in this direction, because it's a much better product, period, and it's something that would fit much better with Disney. But I think he's made his forays into online, so he's probably not going to make them anymore. Probably not. But it's relevant for our guest today, who is Evan Spiegel, the CEO of Snapchat. Snapchat has stayed...

formidable even though it has stayed little. It has under a billion monthly active users versus say Facebook's 3 billion and its market cap at the time we're taping this is around 15 billion compared to Meta's 500 billion for scale. But they've managed to stay relevant. Why do you think that is? Well, I

I think there's two things. Look, it's businesses under huge pressure and it may not survive, right? Because it's small, but they're trying to be innovative. They're trying to do things. They've always been at the forefront of new and fresh things in social media. I think they're unfortunately caught in a situation where there's secular changes around social. It's sort of merging with generative AI. There's all kinds of things going on. So even if it's creative and wonderful, it also faces huge business challenges, which is reflected in its really low stock price.

I'm surprised someone hasn't picked this company up and Evan Spiegel, who's such a creative. And so, you know, they have stayed at the forefront of creativity, but it may be a losing game, unfortunately, for what is a very good product. I think you're right. I think they always have gotten, I don't even want to say creators, but creative people. These days, they've been betting on augmented reality and artificial intelligence, AR glasses. They're going to face a lot of competition from Apple. Yeah. And then with AI, it's especially tricky because

your user base is so young, right? Yeah. Asa Raskin has done some tests of the MyAI Snap product. And I remember reading that Washington Post article, kind of surveying that and other tests of Snap AI. And, you know, it gave Raskin this kind of wildly inappropriate advice. You know, he was posing as a 13-year-old. The AI was basically giving him advice on how he could set the mood for sex with a 30-something-year-old. So,

really kind of no guardrails, right? Yeah, they fixed it. They fixed it. Do you think that this is going to be their Achilles heel? No, I think they have to try this. I just think you don't roll these things out without perfection, but there's not going to be perfection because what the problem is humans asking the questions. Like if you ask a really twisted question, you can get a twisted answer. You know, they're a very small company, so it's much harder to catch everything. And it's going to be, it's hard for all of them. You know, you could do that with anybody. Like, I don't know, Kevin Roost had a, had a

relationship with an AI, wrote a column about it. It's like, you can do it over and over. Get a relationship with an AI. You know, he made a big, you know, he wants to break up. That was what the piece was about. It was a stunt. There's a lot of the stunt stuff and it's going to happen over and over again. I mean, I see why you say it's a stunt because the headline is so, you know, is so click worthy. But I also think it's a service because it's proving the limits of the product. Not the limits. You have to fix it. And they did. They fixed it. They fixed it.

they fixed it. But should they even be in this business? Snap has seen a big spike in one-star reviews on the App Store since launching MyAI. And I think the question is when your audience is so young, and obviously people say they're above 13 when they sign on to Snap or any of these products, but

should kids be in the guinea pig arena for AI? Or is it inevitable? You know, I don't know. There's just this bill that Chris Murphy and Josh Hawley just introduced, the social media bill that's trying to limit anyone under 13 to use social media and then parental permission to 18, a very controversial bill, I think. I like Chris Murphy a lot. Not so much Josh Hawley, but it's a bipartisan bill. And it's to protect kids. It feels a little like Chinese Communist Party, but we'll see.

Why do you think it's controversial? Because this is the United States of America, right? I mean, on one can, you can argue, you know, kids can't get cigarettes, they can't drink. I mean, they try, but there's laws against it, right? And so why shouldn't the same thing be from a social media? And it's a good argument. It's a good argument. It just feels, given it's media, it's a little less... They can't see rated R movies, et cetera. Et cetera. So we do that all the time. We age gate things all the time. And the question is, social media is a little different than all those things, but maybe

Maybe not, maybe it isn't, and this bill will pass. But it's definitely, it's a little bit of a step down a road that feels like the Chinese Communist Party. And also a lot of this is giving parental permissions and there's concerns. I know that some advocacy groups have said LGBTQ kids will suffer because that kind of parental scrutiny. That's right. Parental permissions. That has been a big problem with the COSA bill that was introduced earlier.

Blumenthal and others. So let's see how they get around it. Certainly one to watch. And Snap is at the forefront of these questions with kids. So it's a great time for this conversation, which you tape live in an airport hangar. Yes. It's a fantastic venue in Los Angeles. And Snap put on quite a show. It was really wonderful, actually. They showed off all kinds of shopping stuff and

You know, trying things on, you could see the clothes on you and these mirrors and AR stuff and glasses. And they did a great, at Nike, you imagine shoes on your feet and customize them. Oh, no, Kara, they were trying to woo you with all their fancy tech. No, but it was cool. It was creative. It was fun to see creativity. I'm sorry, you don't see a lot of creativity from tech people. That's true. But you like Evan. I do like Evan. Why? Why?

Why do I like Evan? Because he's evolved as a person and an entrepreneur. And I appreciate it. We had a very rough beginning of our relationship. Because? Oh, he had a lot of controversy around things, emails he wrote when he was in college. And he decided I was all press. I never even wrote about the controversy. He wrote some terrible things in emails.

And then he said, oh, it was when I was younger. And I pointed out to him quite correctly, well, it was three years ago. So you can't say when it was younger. So he just got, he had a little rough going with the press initially, but I think he's really evolved and has really embraced a little more,

Just listening to more people. That's all. And, you know, a lot of people don't do that in tech. They just forge ahead. But that's why I appreciate it. We don't always agree. Just like with Mark Benioff, there's a couple people like that, but they also listen and evolve. And so I appreciate anyone who's willing to have a

real discussion. To engage. Yeah, rather than just dunk. Anyways, at this technology-filled conference that was taped days before Snap's Thursday earning call, so Evan was necessarily circumspect. I'm sorry it's a little bit tinny because it's in an airplane hangar, but it's not plates like the last one with Ron Klain. Yeah. Yeah.

Plates or planes, those are the options. But thank God for our producers and engineers who made it sound so much smoother than it actually probably sounded in the room. Next episode, we're going to have you tape in the middle of a wind tunnel. I'll do a car as I'm going. I'll be in like a Lucid or something and taping. But I love doing live events. And so this is what it is. Anyways, let's take a quick break and we'll be back with Evan Spiegel. ♪♪♪

This episode is brought to you by Shopify.

So

Let's start with your stock price. We spoke just over six months ago. The stock was hovering around $10 to $12 then. It's between $10 to $11 today, but still below your IPO price. In the pandemic, it went up to $80 a share. I think a lot of tech stocks did. How are you feeling now versus six months? I know six months ago, it was a tougher time, but the stock price still doesn't reflect that. Can you talk a little bit about that?

Yeah, well, the stock price has certainly had its ups and downs over the years. And really, if you think about the past year, we had to act very decisively and quickly to respond to a very rapidly changing environment, following the war in Ukraine, of course, rising inflation.

rapidly rising interest rates, the economic picture changed very, very quickly. And so as a business, we had to reprioritize all the projects we were working on to make sure they all supported our goals of growing our community, growing revenue, and of course, investing in AR. And then of course, as you all saw today, we've been working really hard to innovate and then build a ton of

products that our community really loves to continue growing our community and boost their engagement. And ultimately, we believe that'll translate into revenue over time. So talk about that because innovation is not something you've ever lacked. You've always tried new things. You've tried things that don't work and do work, but every year it comes up with something else. You do get copied or someone supplants you. How

How do you look at that competition right now? Because you're in a very competitive environment where everybody sort of copies everybody else's thing. Meta is trying to copy TikTok in their reels and things like that. How do you look at the wider competitive landscape right now?

Well, I think at our core, each of these companies really serve their customers in different ways. So if you think about Snapchat, we've always focused on communication, visual communication between friends and family. And that's always been the core driver of engagement on Snapchat. People come back time after time. In the US, it's nearly 40 times a day that people open Snapchat to communicate and to express themselves.

I think following on all that engagement around communication, we built other parts of our business, like our content platform with Stories or Spotlight, of course, our AR platform. And of course, we've been working really hard on our map and now MyAI. So we've been able over time to build on this engagement around communication to build other services. And I do think that has continued to differentiate Snapchat compared to social media or entertainment products.

and why we've been able to continue to grow our community. - But they've been veering into your area too and also sucking up time, 'cause time is what people don't have, right? That's the one thing that you all have to compete with. Do you think time is a problem for you?

Time is an important metric for success with a content business. So as you mentioned, we do compete with other content businesses, of course, that have platforms like social media companies or entertainment companies. We also compete in short form video and in social media content to a degree with products like Stories. But in other parts of our business, like our camera or with our communications service,

we're less focused on growing time spent and instead are thinking about monetization in other ways. So if you look at our map, for example, our goal is actually to help you find a place that's interesting to you as quickly as possible. And so I do think you can think about building value in your business in different ways, but certainly in the content piece of our business,

That is a place we're focused on growing time spent and why we're so excited to see time spent growing in Spotlight, time spent growing with creator stories, especially creators who are part of our revenue share program, because folks are really loving the content that people are creating on Snapchat. So when you think about the competitive landscape, do you think there should be a ban on TikTok? We'd love that.

We'd love that in the short term. I think there are some big questions about what that would mean longer term, you know, to single out a single technology company instead of developing a more comprehensive regulatory framework. Well, some places they want to ban all of them, for example, but go ahead.

So I do think it is important for us to be thoughtful and really develop a regulatory framework to deal with national security concerns, especially around technology. And I think based on the information that is publicly available, there are legitimate national security concerns far above my pay grade or maybe security clearance. But I do think it's— They actually, to be fair, they haven't proved it yet. It's just Congress people screaming about it without any proof. But that's sort of par for the course. But that's okay. Yeah.

Again, at our security clearance level, that certainly seems to be the case. So, you know, I do think that there needs to be a regulatory framework that's developed more broadly to deal with this issue. But yeah, of course, in the short term, that is something that would help us out. What else could help you grow? You know, you can't rely on a ban. I mean, it might help you, for example, but a ban is not, you know, how I assume you want to do business. What's the most important part of growth right now for you?

well i think there's sort of short medium and longer term efforts in the short term it really is about our advertising platform and continuing to evolve our platform to drive roi for advertisers and in this sort of economic environment that is the thing that they care the most about and so i think we're really well positioned to continue evolving our direct response platform and by demonstrating results for advertisers attract more dollars you know if you think about uh

Snap, I think we generated, what, a bit over $4 billion last year. As a percentage of the overall digital advertising industry, we're very low single digits. So just in terms of growing our share relative to other advertising businesses. I will point out, you're bigger than Twitter now. It's great news. Yeah.

Yeah. We think there's a lot of progress to make in our direct response business and to gain share. You know, we have 750 million people using our service every month. So just in terms of the overall digital advertising market, we see a lot of opportunity to take share, especially with our direct response business and our ad platform. So that...

That's the core short-term focus over the next 12 to 18 months, let's say. Then longer term, we're doing a lot with augmented reality, both on our platform with sponsored augmented reality lenses, but also off-platform with AR enterprise services. We talked a little bit about

And then longer term, we're thinking about the ways that we can monetize our map because we've shown when we highlight specific places, we can increase the number of visitors to those places. So we think that could be a good idea. - It's sort of going back to the past. It feels like Foursquare. I was like, that looks like Foursquare when they were showing, a beautiful version of it, but in a lot of ways. Do you remember Foursquare? No, you're too young. Okay.

So, but let's talk about ads first. So a lot of tech stock fall off, including snaps, has been a correction of the pandemic boom, no question. But, and some reaction to Apple's new privacy settings that targeted ad revenue. You're very bullish on that. And I know most people I talk to in digital advertising are now, someone just said, well, it's not getting worse. Like, because it's stopped going down. How do you look at where it's,

settling out, especially with the pandemic, Apple, and then just the economy in general. Yeah, more broadly speaking, obviously, the end of the low interest rate era, I think, has hit technology companies very, very hard for pretty obvious and I think well understood reasons. And I don't know if there's an end in sight to this inflationary environment. You know, I do think we're going to get some positive short term news.

know about inflation as we start lapping the inflation numbers we saw last year so you know by about the summer time i think people are going to feel like the year-over-year inflation numbers are coming down that'll make people feel more comfortable that these rate rises are kind of going to pause i think that'll create you know some some short to medium-term optimism more broadly in the economy but you know i think some of the structural issues we're seeing in the economy like the very very tight labor market for example that's one of the things that i think is going to be an input into you know inflation

that's going to run a little hotter than we're used to, probably for a longer period of time. So we've entered a new economic era, I would say, and businesses are adapting. Are you still feeling the impact of Apple's ad targeting?

Yes, we've been working through that for quite some time. I think one of the big lingering effects of that is that advertisers went from measuring, you know, advertising relatively the same way. You know, they would focus on different metrics, but they would measure it all relatively the same way. And it was pretty easy to do that.

And now advertisers have lots of different methodologies that they've developed. That makes it a lot harder for us to optimize our platform in a way that can drive results for advertisers that they're now measuring very differently and with different tools. So I think over time as we work through this, advertisers will start to zero in on a few different methodologies. Such as? I mean, here you could say selling, for example, with retailers. For example, we're selling more sweaters, we're selling more glasses, whatever. Whatever retailer you happen to work with.

- Unfortunately, that level of measurement is just less available today than it was before, especially for smaller advertisers. Larger advertisers can afford to do privacy safe data integrations to pass data back and forth in a privacy safe way and really clearly measure their advertising campaigns.

smaller advertisers don't necessarily get that level of measurement anymore. And I think that's one of the real challenges that digital advertisers are facing right now. But, you know, platforms are working through it. We've been investing a huge amount in overcoming some of those challenges. And we've always built our platform in a privacy safe way from the beginning. So we're very, you know,

excited to see that the industry overall is really prioritizing privacy in this way. So when you think about that idea of differentiation via privacy and other ways, at the same time, you've moved more towards the advertising thing. I think you read that Ellis Hamburger and The Verge had a piece saying the reason he went to Snap is because they cared about privacy, cared about things. But you can't inevitably avoid the advertising conundrum of privacy.

putting advertising things or things that will get people to stay on longer. How do you resist that without hurting your own business? Well, I don't think you have to distribute advertising in a way that compromises user privacy. And a lot of what we've done for a very long time, whether that's never offering products to target advertising on a device-specific level, we've always worked really hard to provide tools for advertisers to help them grow their business without compromising the privacy of our community.

I don't think you have to do one or the other. And ostensibly over time, especially as we get better and better at optimizing our advertising and showing advertisements that are more relevant, people are actually delighted to learn about a new product or service. I mean, it's one of the primary ways that people find out about new products today. If it's done creatively, if it's done somebody... I mean, I find a lot of the stuff you do with Vogue very creative, a lot of the try-on stuff...

The other way you're trying to make money is paid subscription services. People are trying. They're trying it at Twitter. It's not working out. IG Verified might work out. Yours is called Snapchat Plus. Everything seems to be called Snapchat Plus. Everything pluses. Talk about that model and how it's going. One of the things I think is one of the original sins of the Internet is focused on advertising too much and not subscriptions necessarily. There are good reasons for that. There's bad reasons for that. But it was an easy...

easy thing to do, advertising. Talk about your paid subscription thing, and is it the future of social media from your perspective? One of the things that we realized, you know, we were sitting in a lot of these design meetings, making up all sorts of new fun features, and they were really features for the people who love our product the most. You know, people use our product every day of the week to talk with their friends and family. And

Sometimes our power users, our most engaged users, use our product in a different way than the broader Snapchat community. And so we really wanted to be able to release these features that we knew that our most engaged members would really, really love. But we were having trouble prioritizing them against features that we knew would benefit everyone in the Snapchat community. And so as we worked on Snapchat+, we realized we could really align our

love of innovating and making new products, especially for the most engaged members of our community with generating revenue and a subscription service. So our goal has just been to constantly create new features. We try to release new features every month or two and give them to our most engaged Snapchatters through our subscription service. - Do you see that being the most important part of your revenue stream? Subscription? Replacing the advertising?

I think it'd have to grow pretty significantly to outpace our opportunity in advertising, but I do think it's a really important part of our service and it's also a really great way to experiment with new products. So when we were working on MyAI, we released it to the Snapchat Plus community first. We got a ton of feedback, made a bunch of changes to the MyAI product before rolling it out to our broader community. So one of the ways you're doing the things you're investing in is very clear is this AR. And you've been, you and I have talked about this for years. The only two people that...

talk about AR with me or Tim Cook and you. And he's a lot less interesting on the topic. But in any case, let's talk about the main technologies as you've invested in augmented reality. It's been not just two years, it's longer than that, including tools for creators to sell digital AR goods. I want to do a quick lightning round to see what's working from your perspective. Talk about

the spectacles. I know you said it would take a long time. A lot of what we found is that, you know, AR is really best suited to experiential use cases. So really changing the world around you, augmenting it, bringing it to life. Right. Instead of these more informational use cases that a lot of people have pursued. So if you think about, you know, telling you the weather or taking a photo or something like that, a lot of that can be done better with your smartphone. So where we've really focused with augmented reality are actually all the things your smartphone can't

do. And that's where we're seeing a lot of success. What's changed? Because it had a little of that. It had a little take a picture and stuff like that. And yours were much more attractive than Google Glass. So talk about the form factor right now, because they're still heavy, whether it's the meta stuff. We don't know what Apple's coming up with yet. So how do you look at it? So experiential in a place or moving around the world?

So one of the real benefits of AR glasses is that instead of sitting behind your desktop computer or staring down at your phone, you can actually use computing out in the world, move around freely, use your hands to interact, which is a ton of fun. I would say as we've developed our glasses over time, I can maybe talk about a couple of the things we learned. One of the things you touched on right away, our first version of glasses were just camera glasses.

- They were. - And that was our goal, was just to see if people would wear technology on their face in that way. And what we learned very, very quickly was that it's just not enough to have a camera in your sunglasses. It's a lot of fun. There's great point of view video and for really passionate and talented creators, they enjoy playing around with that product. But as a mass market consumer product,

I don't think camera glasses are a particularly compelling value proposition when your phone is so capable and always right in your pocket. But what it did teach us a lot about was building technology into glasses. And so,

As we started evolving that technology, we built the first version of AR glasses with displays to bring 3D graphics into the world. That's been a ton of fun. We've learned so much building new experiences together with creators. One of the big things that that taught us actually was that,

We had originally been so focused on almost competing with glasses. Instead of really recognizing that what we're really building is a new computing experience. So it's actually much less about the form factor and much more about what wearing the glasses enables. So we focused a ton on the software that powers all these different lens experiences, the way that people actually interact with those lenses. So you don't see these spectacles that are listed right now at $380 as spectacles?

You don't see them as like an iPhone kind of thing, correct? Like just a comparable. I think in many ways they're far more capable than a phone, but they're also very, very different. So I don't think about AR glasses as cannibalizing a phone. I think about them as enabling a new way to use computing. Such as a watch does. An Apple Watch works with the app. Finally does work pretty well.

Yeah, I think that's a possibility. Yeah. So one of the things you also is virtual try on. It was a big push for you. You did the Vogue stuff. You tried to get me to wear some sort of Stellum Cartney thing, which I declined to do. And then you put it on, which was very you look much better than I did. You're helping sellers get people to try things or imagine things. How are the numbers on this? How big a business do you imagine this is going to be for you?

I think what we're really excited about is just the way that consumers are responding when they're shopping in our partners' apps and websites, and the way that virtual trial and fit-in-size finder are changing their conversion rates, encouraging them to...

and more because they can actually visualize how they look when they're wearing a product. And I think that visualization process is really, really important and it's missing while we're shopping online. And so a lot of our work has just been taking the strengths of our AR platform, all the investments we've made in our AR platform and really driving business value with things like AR Try On. So that's an area we're very excited about. We're investing a lot more. We've, of course, created this new way with one image

enable real-time trial. - So you had the hands, you had the face, and then the body, correct? Which I assume the last one's the hardest. - It's very challenging, especially with different fabrics and the way that they drape, et cetera. It can be quite challenging. - What's the problem there?

when you think of AR? What is it? It just doesn't look real or it doesn't look what? Because they have AR looking into furniture. IKEA does it. So what's the big challenge from your perspective? Well, if you think about furniture, the frequency of purchase is not high enough, in our opinion, to really accelerate consumer adoption of AR, right? It's not very often that you buy a new couch. And so while that AR experience is really useful, you may not

use that you know particularly often whereas what we found with our community you know young people i think they spend about a third of their disposable income on on fashion right so they're constantly trying on new clothing new accessories experimenting with self-expression in that way and and i think that's why you know it's been such a fun area for us to focus because it's a behavior that people do really frequently it's something they really enjoy and it's so tied to self-expression which is really at the core of the snapchat platform

So that's a promising business from your perspective. When you think about the most important applications of AR, and look, Meta's trying it slowly, but it's mostly gaming. They're aiming at gaming. Who is your biggest competitor in this? Is it Apple? These reality one glasses, which are now going to be at some time in 2027. I don't know. They keep delaying it. Is that something you're worried about with them coming in?

To date, they've been a really helpful partner. So if you think about it, we're one of the largest augmented reality services on Apple devices. And I hope that continues to be the case as they continue to build great new devices and work on things in augmented reality. So I'm optimistic we can continue to partner with them there. We're usually one of the earliest adopters of whatever new AR technology they're working on, like LiDAR, for example.

That was a really cool way that we were able to enhance AR experiences for iPhone customers. - So are you worried about them in the way, I recall Daniel X saying this before he said something different, you know, when they went into music more heavily. Are you worried about them becoming a competitor in a way that's problematic for you? - I think, you know, it would be unwise not to worry about any of these large technology companies. - They're pretty large and they're pretty key to your business right now.

We've competed with a lot of very large technology companies for a long time, and so we always keep a healthy dose of paranoia. But so far, I think we've really been able to bring to life really engaging AR experiences for iPhone customers. And because of the investments they make in their hardware, we can do some really advanced and cool stuff in augmented reality, which is one of the things that makes iPhones appealing. You can really bring out some of the best of Snapchat using an iPhone. So you would be on their glasses?

you'd want to be on their glasses. If they'll have us, I think that'd be a really cool thing to experiment with. They don't have the glasses yet, so we'll see. We'll be back in a minute.

So let's get to AI. You released MyAI. It's an internal chatbot powered by ChatGPT. Explain the product and whether AI is a game changer or just an increment thing, because I know everyone's sort of cheerleading around it and sort of rushing into it. I think there's a bit of an arms race around this. My worry is more the arms race aspect of it.

Well, artificial intelligence is definitely a game changer for Snapchat. So if you think about the evolution of our use of AI, we've long used AI to power our content platforms in terms of recommendations. Of course, our ad platform is powered by artificial intelligence. A lot of our

Lenses are powered by generative artificial intelligence. So AI is a huge driver of our business today. What's so exciting about MyAI is it's the first time we're bringing artificial intelligence into communication, into the chat page on Snapchat, which is really, as we discussed, the core of our business. Snapchat's really about communicating with your friends and family. And so to be able to bring artificial intelligence into Snapchat, I

think makes me believe that we could be one of the best ways, if not the best way, to communicate with AI. And so it is very, very early. So this is your new friend. Maya is your new friend, your new pal. I think the way that our community has been using it so far is more as a creative tool than a friend, because it actually really enriches your conversations with your friends. That's one of the reasons why we're so excited to release at

mentioned, so you can bring my AI into conversations with friends. And whether that's coming up with an itinerary for a trip that you have, or I use my AI to get feedback on a wedding speech I'd written, which was a lot of fun. Or for our kids, for example, I use it for story time. We can kind of do these generative stories where we make up stories. And I'm exhausted at the end of the day, so I come home. And it's really fun to do that with our kids. So I think

where I see these large language models really succeeding today is in creativity. They're incredibly good at coming up with new ideas, new concepts, new stories. Where they're struggling still is around informational use cases and accuracy because some of the things that make them so good at creativity actually make them not so great at retrieving perfectly accurate information. So I think that's why my AI is the right

place for this product today. There's two issues I see. One is the broadly speaking IP issues around because it's not search, it's take. Again, technology companies taking content from all over and reusing it. That's one issue and I think that's going to be a big legal problem for all of you going forward. It's sort of like a super powered YouTube or something else. And the second thing is the creepy factor. Which one would you like to discuss first? Uh...

Say, I'll take creepy for a hundred. That's exactly what I was thinking. I'll take copyright. Copyright. We're going to do both. So talk about copyright. Are you worried about that issue of it pulling in? And who's responsible, really, for that? Well, I think what's interesting to your point is actually that it's not reusing the content. It's actually generating totally new content. From content, right?

Yeah, yes. If you're a songwriter, you're not going to love it. If you're a pick, like, look, if you took the whole Godfather and started playing with it, I'm sure Francis Ford Coppola would have a thing to say or two about it. But I think as artists, we all get inspired by other people's work, right? Their, you know, their music, their paintings. That's, I think, long been a feature, actually, of...

of art and creativity. And so, to me, I actually see it in a continuation of that vein, that I think the way that these systems have been architected, they, of course, learn from a lot of content that's out there. And in the case of our models, for example, we use first-party data, we use licensed third-party data,

Of course, we use synthetic data as well, and that's how we think about managing the rights issues. But I think more broadly, when you look at this technology in the way that it's actually generating something totally new, not reusing, not copying, you know, a piece of content. It's using from it. It's an interesting problem from a legal point of view. I've talked to a lot of lawyers about this because some of it is taking.

and remaking. And so photographers are starting to have issues, artists are starting to have issues, movie makers soon. In the visuals where you are, it could be a problem as they begin to take content.

You don't think that's something you're right. That's why, as I mentioned, when we build our own models internally, we use license data or we use our own first party data or we generate data. Sometimes we create synthetic data that we train on. So I think that's how we're thinking about the risk management of this rights issue. But I think more broadly speaking, in response to your question, the thing that is so fascinating about these systems is that they are generating totally new

content from what they've learned. - Yeah, and the issue is sometimes it's not correct. You were talking about the misinformation. It's called hallucinations. I just call it wrong, but that's fine. But let's get creepy. The launch of MyAva is controversial for everybody, not just you.

Asa Raskin from the Center for Human Technology did a test. Every company, they've been testing them all out on. I think ChatGPT tried to break up Kevin Roos' marriage and date him. Okay. This one happened to you. It did a test as a 13-year-old girl when asked to make losing her virginity with a man 18 years older than her special. The AA provided tips on how to set the mood with candles and music rather than saying, I don't know, call the police. Talk

Talk about that launch, because every single company, Google's had this problem again. I don't want to just blame it on you, but every one of these has been releasing these things and reporters are able to do this and generate this. So in that specific scenario, as you mentioned, I believe that was a researcher who was adversarially using my AI to try to get it to say things that were inappropriate. The longer you're with it, the

the more it loses information previously, from what I understand. And in a way, I think that sort of research is actually very helpful. That's exactly what people should be doing with it. And I think humans, whenever they come across new technology, one of the first things we try to do as humans is break it.

which is why building a service that is safe is so important to us. So if you think about how we've architected MyAI, whether it's storing the conversations that people have with MyAI and reviewing them, 99.5% of MyAI responses comply with our community guidelines. Right. I'm worried about the 0.5 part. You know, even though, like, I mean, many tech companies have said this to me, 99%, and I'm like, well, the 1%.

you know, say, died in Myanmar. So I'm going to go with 100%. When we dive into that 0.5%, actually, what we find is oftentimes when it's not complying with our community guidelines, it's either repeating what somebody said to it that was inappropriate, or oftentimes it could be citing an inappropriate song lyric, for example. So when we dive into the reasons why it's breaking or not working, what we find, again, even in the failure cases,

makes us feel comfortable with a broader role. And that's not to say that it's perfect, it's going to make mistakes, but we can learn from that and continue to evolve the technology. - I'm not particularly worried about Snapchat. I think a couple weeks ago you had a long blog post about the early learnings from my AA on safety enhancements where you talk about guardrails and age appropriate design. I think you are concerned about it. Summarize what you're doing differently and more importantly tell me how you're gonna know what you don't know because you have the responsibility for a young audience at Snapchat. And should,

It should be age-gated. The use of my... You just released it to all the Snapchatters. I would be comfortable with my older kids using it. I definitely wouldn't be comfortable with younger kids using it.

So in order to use Snapchat, you have to be over the age of 13. When you're interacting with MyAI, we pass the age of the Snapchatter to MyAI to make sure that conversation is age-appropriate for the age of the person who is communicating with MyAI. And I think that's an example of the way that we've worked to make the service age-appropriate. So your experience does change over time.

based on your age. We've also built my eye on the foundation we already have, whether it's things like Family Center so parents can see who their teens are chatting with, or change the content settings, for example, or just the ease of

know reporting very quickly an inappropriate message so that you can get help and of course we can improve our service i think that strong foundation we have you know managing you know trust and safety at snapchat is again one of the reasons why we feel confident in in rolling out my eye more broadly even though we know it's not perfect you have i think more responsibility to a young audience than others since your audience is so young

How do you make the calculations with this rush towards AI? I do think it's a rush. I think it's another thoughtless rush. Not necessarily you, but in general, just to win, just to win ground. Is that a dangerous attitude? You know Tristan Harris has talked about it and others have talked about it.

I see it pretty differently. I mean, I can't remember the last time a new technology was actually rolled out this thoughtfully and that there's been this much thoughtful debate this early, I think is actually very promising. I actually think it shows we've learned a lot about the evolution of the internet, for example, or the evolution of social media. So I actually think people are asking a lot of harder questions sooner than with any other piece of technology I've ever seen.

in my life. So that actually makes me quite optimistic about how thoughtful folks are being rolling out these products. And these transformer models are, what, more than five years old at this point. So I'm not sure that I agree with the sort of rush narrative. But I do think the sense of excitement people are feeling is very real because folks all over the world are embracing this technology. What's your worst case scenario here, though? I mean, you have to be thinking, I can...

I remember when one of those Facebook Live came out where I said, what about if people strap a thing on their helmet, start killing people and putting it on Facebook Live? What's your tools to stop it? And literally the room was like, Carrie, you're a bummer. And I was like, yeah, I'm a bummer. People are terrible. And so when I think about it, I can think of 20 cases of things that could happen in this thing. What is yours?

I'm much more concerned about the way that humans will misuse this technology than I am about a response that my AI might provide. I see. And I can't imagine necessarily what they might do, but, you know, for example, fraudsters might use, you know,

this sort of GPT technology to write a really convincing phishing email and be able to do that at scale for all sorts of different people. So I worry that this sort of technology will be very useful to bad actors. And that's why it's so important for us, again, to monitor the conversations people are having with MyAI so that we can detect that behavior, learn from that misuse. And we've rolled out timeouts so that if people are misusing MyAI, we can stop

that conversation and slow that conversation down. So AI is not killing people. People are killing people as usual. As per usual, it's true though. I think you're right. So lastly, I want to finish talking about

You've come under fire for your filters, for example. There's several lawsuits before about the speed filter and car accidents. Lower courts concluded that you were immune from liability due to Section 230, but the Ninth Circuit Court of Appeals found differently. When you think about where we're going with AI, should these companies, including yours,

liable finally for everything? Or do you think it falls under, I don't think Section 230 covers generative AI. And Sam Altman at OpenAI told me that he doesn't think there should be immunity. How do you look at that? And you called Section, in 2021, Section 230 a brilliant rule. How do you look at that now?

So I think there are a couple different and important pieces of this. The first one is that regulation is just never a substitute for taking responsibility. From what we've seen, regulation lags new technology by sometimes more than- Well, there isn't any regulation on- In the United States of America, I think that that's accurate.

And that was going to be my second point. I actually think we learned a lot from our quite thoughtful and deliberate approach to the internet. When the internet was first created, there was a very specific decision in the US to actually let it grow unregulated. Yes. And I think in the early days, that was a very wise approach because we weren't necessarily sure what the internet would become, what shape it would take because it was so early. And

You know, I think as a result, the United States is full of leading internet companies in the world and has been really at the forefront of technology because we created space for people to innovate and try things. But I think if you look at what's happened since then, the fact that we're one of the only countries in the world that's still lacking comprehensive federal privacy regulation, it's clear that we've now waited too long.

And so what I'm hopeful for is that we could take this learning, that it is wise to observe early, learn a lot early, try to understand the way these systems can be misused, and then create a regulatory framework as that technology evolves without stifling it. Do we need a new regulatory body? Some people have talked, I don't want to call it the Ministry of Information, but some people have talked about an AI department to regulate this.

Well, one of the things that I think could be very useful is to build more expertise within our government on these topics. One of the challenges we've had since the 1980s is we've repeatedly reduced the budget for these sort of centers of excellence inside our government, whether it's folks that study technology or tax policy, that we're really advisors to Congress.

we could really benefit from government employees who can really provide expertise and understanding of these complex technical issues. So I think the most important thing we can do right now, especially with the technology that's this new, that's just emerging, is do the best we can to learn as quickly as possible and really build that expertise inside of our government so that we can be thoughtful about regulation. I do know the Justice Department is hiring AI experts right now, and I think they're doing it with an eye towards AI.

possible litigation if they need to. But speaking of the last thing about regulation, I have two more questions. We discussed TikTok and national security earlier, but obviously antitrust is a big issue. You are a small company. You are, comparatively speaking, you're a small company. You're supportive of this legislation? The idea? It didn't pass in the last Congress. Antitrust legislation. Sorry.

I think at this point, in the evolution of our business, we've just learned not to rely on government intervention in that way and instead just focus on serving our community.

you know, the history of technology would suggest that that's the right approach. Because just about at the time that the antitrust regulators, you know, intervene with, let's, you know, use the Microsoft example, there's another new technology company that's taking flight. And so, you know, our view has really just been to stay focused on our community, to continue to innovate and really try to grow that way. So you, of all companies, I would say have been the most

abused by bigger tech companies if I had to pick a company in terms of your creativity getting snatched out and redone by them. It hasn't been that bad. Oh, it has for you. Now, so you don't think there needs to be antitrust or worry about that. You're just going to keep innovating.

I think it's an interesting moment when we've got a community of 750 million people that love our products and a ton of exciting innovation in our business, and we're a really small company that should need antitrust intervention. That's not necessarily, if I were a regulator, where I would focus. I do think there are small startups and developers that probably would benefit from some antitrust intervention.

Legislation, for example, there's some new regulation I think coming that allows side loading of applications on devices to circumvent the app stores. That sort of regulation I think will create new opportunities for startups and new developers to get distribution in different ways. So I do think that antitrust legislation will create those sorts of opportunities. Would you like to be out of the app store paradigm, both for Google and Apple?

The App Store has served us really well, as have the developer platforms that both Apple and Google provide. So frankly, we've built our business on that foundation, and we're very grateful for it. But it's still a company town. You still have to use them no matter what.

Yeah, but again, we really enjoy using their tools and their incredible devices that they create, the operating systems that are incredibly sophisticated. So my very last question is, you and I both have small kids. Both of us have young kids. You have a four-year-old. I have a three-year-old. They're close in age. They're 17 years old. They're hanging out. They've turned out to be best friends. You're very upset about the situation. But nonetheless, here we are. What are they doing?

Well, hopefully they're using Spectacles. We'd be really excited about that. What do those look like and what are they doing with them? I think, you know, one of the things that could be... And it better not be creepy, just so you know. But go ahead, because I'm very particular. I think they could be playing a game together. I mean, that's one of the things that we've found to be really enjoyable with the most recent version of Spectacles, that you can have a shared game experience together with your friends. So we'll be wearing a heads-up display playing a game. Outside. It's a lot of fun. All right. Evan Spiegel, everybody. Thank you.

A shocker that a TikTok ban would help Snapchat.

You know, they all say that. He was joking, and I think, but he's right. I mean, they have really cleaned everybody's clocks, so to speak, in terms of creativity. And, you know, TikTok has been very creative, and I think it's been a challenge for Snap and Facebook. I was surprised how little he bit on antitrust, because that also could help him, as a TikTok ban could do. But I guess that can't be part of your kind of forward-facing strategy to rely on regulators, too.

kind of upend businesses around you, right? I think he's sophisticated enough to understand he's going to rise and fall based on his earnings and his products. What do you think Snap's endgame should be? I mean, do they get bought by Apple? He tried to say they're a partner, not a competitor, but they might... You think sale? Yeah. I think when I talked to some people and they said, oh, I think you should sell to Apple. And they're like, oh, please call Tim Cook. And I was like, you know, I think they have a great product. I think that's the problem when it's...

when it's a really great product and a great offering and creative people, but you're still too small. I think Snap's biggest problem is meta. Meta just is too big. It's too big to fight. Yeah, and that way they're well-named. I mean, obviously, they're trying to diversify away from advertising revenue of doing this AR try-and-buy thing. Are you more bullish on shopping, on subscriptions?

What? Well, you know, I thought this Ares thing they had, it stands for an enterprise solution. And I thought that was interesting because it wasn't so much the retail stuff. I do think retail has to find ways to get people back in the stores. You know, they had a Coke machine there at the event. A Coke?

machine, like a vending machine? Yeah, it's a Coke machine that you move your hands in front of and it's all AR and stuff like that. It's really cool. Just a question of whether they can make it into a product because they're small. Like, this is all the stuff. The mirrors were cool. The try-on was cool. So it's just, you know, someone said to me, I'm

It may have been Casey Newton, actually, who was there. He's like, these are all great. How are they going to make money? And I think that's the absolute question. Well, affiliate fees on some of the sales stuff, like they should get a slice of the revenues, you know. Could be. Could be. Takes a lot of stores to do it and to accept it. It's not easy stuff. It's not easy. But stuff like that should be in stores and make the store experience better. So that's kind of cool, but can they make a business out of it?

Yeah, and you have to integrate with like a point-of-sale provider, like a payments company or something. But he said something interesting, which he said, I don't think about AR glasses as cannibalizing a phone. I think of them as enabling a new way to use computing, which you thought he was talking about Apple Watch, but he was almost speaking more about like an experiential way to...

to compute through your vision, compute through what you see, right? I think he's talking about experiencing the world. You go to, you're wearing these glasses, you're in France, you look at the Eiffel Tower, you get all kinds of information. I think that's cool. And I do think that's coming, period. It's coming. It's like layering computing onto the real world, basically. Yes. Yeah. He's been in the forefront. You've seen a lot of these things come through, Foursquare, Google Glass, and you mentioned a couple of them in the interview. What? What? Six.

succeed? And how much is timing? Is it technology? Is it leadership? It's the right idea. It's the right, it's 100% the right idea. It's just timing. Like, listen, I was just talking to someone today about General Magic. General Magic was this company nobody knows about, but it was where Tony Fadell was and where Andy Rubin was. That's iPhone and Android. And so- It was like an iPad, right? It was great. It was, I have one in my basement somewhere, but it was, you know, it's early.

directionally, it's completely correct. Whether it's going, who's going to do it is another question, but eventually we'll all have glasses or some sort of a heads up display that we see in front of us. No, I have no question about that. Final thing, and probably the most important, AI. So Evan made this point that we're in a much better place now with AI than we were in the early days of social media. We're ahead of the curve on regulation. We're ahead of the curve in terms of thinking and talking about the consequences. Do you agree with that? Do you

Do you think we're in a more evolved state? Yes, I think we are. But I think people are aware of it. It's a question if they do anything about it. I never think our regulators ever do anything in this area. So I think people are talking about it, which is good. I think they're talking about it in kind of the wrong way. Is this going to be the terminator? That's not the discussion. It's about the same thing, privacy, guardrails, and everything else. So yes, I think we are ahead of it. And compensation for ideas that I rely on as well.

Copyright, IP, everything. Yeah, I do think he's right in that regard. Let's see what the regulators do. He used that line, regulation isn't a replacement for responsibility, which is, they're always so good. They're always so slick, the CEOs, on saying that because- Yes, it is. It's a replacement for responsibility. That's exact. I would not, I would go, if stop signs weren't there, I would go through all the streets. I'm sorry. It is an enforcer of responsibility. It's not a replacement. It's an enforcer. And I sometimes think that these CEOs can do more. Sam Altman came on and said-

regulate us kind of, right? Regulate us, government, which is what Meta has always said. And yet OpenAI hadn't opened the kimono to the U.S. government between 2018-15 or to now, right? Well, they weren't asking. They weren't asking. So it's a question of our government. But they come in and they're like, well, we want to be regulated. It's like, well, you can also have those conversations so that this country can get ahead of a global regulatory scheme, for example. That's not how it works, though. It's not. I mean, they just, they were making it

I just think it's just like a stop sign doesn't mean stop. It means stop and then go. And so that's what a government is supposed to do. Well, we'll see if they do anything about it. In the meantime, please read us out, Cara. Okay. Today's show was produced by Naima Raza, Blake Nishik, Christian Castro-Rossell, and Megan Burney. Special thanks to Mary Mathis. Our engineers are Fernando Arruda and Rick Kwan. And our theme music is by Trackademics. If you're already following the show, you get to try and buy overalls with Snap Spectacles.

If not, it's Google Glass for you. Go wherever you listen to podcasts, search for On with Kara Swisher and hit follow. Thanks for listening to On with Kara Swisher from New York Magazine, the Vox Media Podcast Network, and us. We'll be back on Thursday with more.