1 00:00:02,520 --> 00:00:08,520 Speaker 1: Bloomberg Audio Studios, podcasts, radio news. Well, let's head back 2 00:00:08,560 --> 00:00:11,160 Speaker 1: to the screen Time main stage, the theater here at 3 00:00:11,200 --> 00:00:14,040 Speaker 1: screen Time and to Adam Asari, he's the head of Instagram. 4 00:00:14,200 --> 00:00:15,320 Speaker 1: Over at Meta Platforms. 5 00:00:15,320 --> 00:00:16,720 Speaker 2: He's speaking with Blueberg News. 6 00:00:16,720 --> 00:00:20,320 Speaker 1: Is Kurt Wagner mister Beast. Everyone in this audience already 7 00:00:20,320 --> 00:00:21,720 Speaker 1: knows who that is. I don't need to explain who 8 00:00:21,720 --> 00:00:24,520 Speaker 1: mister Beast is. He posted actually on Threads. This was 9 00:00:24,560 --> 00:00:27,960 Speaker 1: his Threads post, which is also your platform, that it's 10 00:00:28,000 --> 00:00:31,560 Speaker 1: scary times for creators because of these gen AI tools 11 00:00:31,560 --> 00:00:34,920 Speaker 1: that are coming out. What do you think should creators 12 00:00:34,960 --> 00:00:36,480 Speaker 1: be scared as the scary times? 13 00:00:37,200 --> 00:00:41,320 Speaker 2: I think things are changing really quickly, and I think 14 00:00:41,360 --> 00:00:44,400 Speaker 2: that means that a lot of us, whether you're using 15 00:00:44,400 --> 00:00:47,599 Speaker 2: a platform like Jimmy is or you're building a platform 16 00:00:47,680 --> 00:00:50,960 Speaker 2: like we do on our team, are having to think 17 00:00:51,280 --> 00:00:54,120 Speaker 2: about how we need to evolve how we do what 18 00:00:54,160 --> 00:00:57,680 Speaker 2: we do. The jobs are changing, and so I don't 19 00:00:57,680 --> 00:00:59,920 Speaker 2: think that technology from the creator's side is going to 20 00:01:00,160 --> 00:01:04,160 Speaker 2: place creativity. I think it's going to change who can 21 00:01:04,240 --> 00:01:07,000 Speaker 2: be creative. I think it's going to change how people 22 00:01:07,040 --> 00:01:09,559 Speaker 2: tell their stories. But at the end of the day, 23 00:01:09,720 --> 00:01:13,720 Speaker 2: I think that people have creators and creatives. They have 24 00:01:13,800 --> 00:01:17,520 Speaker 2: their taste, they have their processes, they have their judgments, 25 00:01:17,560 --> 00:01:20,560 Speaker 2: they have their cultural commentary, and they're going to use 26 00:01:20,640 --> 00:01:25,080 Speaker 2: those in different ways than before. Because to produce a 27 00:01:25,200 --> 00:01:29,120 Speaker 2: video isn't going to require what Jimmy has historically done 28 00:01:29,120 --> 00:01:31,800 Speaker 2: with Mister Beast and these huge sets and these amazing 29 00:01:31,800 --> 00:01:36,440 Speaker 2: elaborate production teams. It's going to be something that anybody 30 00:01:36,480 --> 00:01:38,759 Speaker 2: can do with access to the right model. 31 00:01:39,240 --> 00:01:42,520 Speaker 1: So on the one hand, it will make certain things easier, 32 00:01:43,120 --> 00:01:45,160 Speaker 1: it will lower the bar for what it takes to, 33 00:01:45,680 --> 00:01:51,080 Speaker 1: you know, produce maybe something that feels highly like produced. 34 00:01:51,760 --> 00:01:57,040 Speaker 1: Does this create more creators? Does this create an upper 35 00:01:57,080 --> 00:02:00,360 Speaker 1: crust of creators, you know, because there's now so many 36 00:02:00,360 --> 00:02:01,320 Speaker 1: more people who can do this. 37 00:02:02,520 --> 00:02:06,280 Speaker 2: I think it'll mean that today's creators, if they lean 38 00:02:06,320 --> 00:02:08,320 Speaker 2: into the tools, are the ones who do will be 39 00:02:08,320 --> 00:02:11,000 Speaker 2: able to create more or even better content. But I 40 00:02:11,000 --> 00:02:13,839 Speaker 2: think it also means that they're going to be people 41 00:02:13,840 --> 00:02:17,560 Speaker 2: who couldn't be creators before who can now actually produce 42 00:02:17,600 --> 00:02:20,040 Speaker 2: content of a certain quality or at a certain scale. 43 00:02:20,360 --> 00:02:22,560 Speaker 2: But it also means they're going to be bad actors 44 00:02:22,600 --> 00:02:26,560 Speaker 2: who try to leverage these technologies for nefarious purposes as well. 45 00:02:26,639 --> 00:02:30,959 Speaker 2: So it's going to be well if you think a 46 00:02:30,960 --> 00:02:34,120 Speaker 2: big step back, what the Internet did, among other things, 47 00:02:34,320 --> 00:02:38,200 Speaker 2: was allow almost anyone to become a publisher, right, reducing 48 00:02:38,200 --> 00:02:42,720 Speaker 2: the cost of distributing content to essentially zero. And what 49 00:02:42,960 --> 00:02:45,000 Speaker 2: some of these generative AI models look like they're going 50 00:02:45,040 --> 00:02:46,600 Speaker 2: to do is they're going to reduce the cost of 51 00:02:46,639 --> 00:02:52,320 Speaker 2: producing content to basically zero. So that will be a 52 00:02:52,400 --> 00:02:54,880 Speaker 2: multiplier effect on the things that we're already seeing, which 53 00:02:54,919 --> 00:02:57,799 Speaker 2: is that there's more people producing more content and reaching 54 00:02:57,800 --> 00:03:01,240 Speaker 2: more people, and there's a lot of good and bad 55 00:03:01,280 --> 00:03:02,560 Speaker 2: that comes downstream from that. 56 00:03:02,800 --> 00:03:05,120 Speaker 1: Yeah, So, as someone who creates an algorithm and intended 57 00:03:05,160 --> 00:03:08,240 Speaker 1: to show people things that they want to see, do 58 00:03:08,280 --> 00:03:11,000 Speaker 1: you anticipate I don't know if you can break down 59 00:03:11,040 --> 00:03:13,280 Speaker 1: what the average person has seen right now in terms 60 00:03:13,280 --> 00:03:16,000 Speaker 1: of stuff that's AI generated versus you know, maybe more 61 00:03:16,000 --> 00:03:18,280 Speaker 1: traditional human generated. But what's that going to look like 62 00:03:18,360 --> 00:03:20,320 Speaker 1: in a year or in five years. Are we going 63 00:03:20,400 --> 00:03:24,040 Speaker 1: to be seeing a feed just full of AI stuff 64 00:03:24,160 --> 00:03:26,360 Speaker 1: or is there a human element that's going to be 65 00:03:26,400 --> 00:03:27,080 Speaker 1: important here? 66 00:03:27,160 --> 00:03:30,760 Speaker 2: Well? Today on the big platforms like Instagram and TikTok 67 00:03:31,000 --> 00:03:34,399 Speaker 2: and Facebook and YouTube, I think there's actually a lot 68 00:03:34,400 --> 00:03:37,600 Speaker 2: of hybrid content. And too often we talk about synthetic 69 00:03:37,640 --> 00:03:41,160 Speaker 2: content and organic content and never the chee shop meet. 70 00:03:41,240 --> 00:03:43,560 Speaker 2: But there's a lot of content where AI is part 71 00:03:43,560 --> 00:03:45,880 Speaker 2: of the workflow, right, Like you might not even know it. 72 00:03:45,960 --> 00:03:49,040 Speaker 2: Might be whatever app you're using to do color correction, 73 00:03:49,160 --> 00:03:52,240 Speaker 2: might be using models to do so, you know, stamp 74 00:03:52,280 --> 00:03:54,280 Speaker 2: tool and photoshop in the old days to get rid 75 00:03:54,320 --> 00:03:57,400 Speaker 2: of a pimple, those types of things. But that line 76 00:03:57,440 --> 00:03:59,600 Speaker 2: is going to be more and more blurred, so it's 77 00:03:59,600 --> 00:04:02,760 Speaker 2: going to be little bit less like what is organic 78 00:04:02,800 --> 00:04:05,920 Speaker 2: content and what is AI synthetic content and what the 79 00:04:05,960 --> 00:04:07,760 Speaker 2: percentages are. I think it's going to be actually more 80 00:04:07,760 --> 00:04:10,520 Speaker 2: in the middle than pure synthetic content for a while. 81 00:04:10,920 --> 00:04:14,840 Speaker 2: But you will also see experiences that are purely AI, right. 82 00:04:14,960 --> 00:04:18,000 Speaker 2: You see this in the last two weeks with Vibes 83 00:04:18,040 --> 00:04:21,320 Speaker 2: in the medi app and with Soa too from open Ai. 84 00:04:22,160 --> 00:04:23,320 Speaker 2: So I think you're going to see both. 85 00:04:23,920 --> 00:04:27,000 Speaker 1: You mentioned there's a lot of potential for bad actors 86 00:04:27,080 --> 00:04:30,360 Speaker 1: out there. We talked a few weeks ago, actually Instagram 87 00:04:30,400 --> 00:04:34,080 Speaker 1: announced three billion users and we had a conversation then, 88 00:04:34,200 --> 00:04:35,960 Speaker 1: and one of the things you said was, you know, 89 00:04:36,480 --> 00:04:39,679 Speaker 1: you rely especially with AI on a lot of self reporting, 90 00:04:39,800 --> 00:04:43,440 Speaker 1: people have to sort of let their audience know I 91 00:04:43,560 --> 00:04:46,080 Speaker 1: used AI for this. A lot of people don't do that. 92 00:04:47,160 --> 00:04:51,159 Speaker 1: I'm curious what responsibility you think Instagram and Meta have 93 00:04:52,120 --> 00:04:55,400 Speaker 1: versus how much of it is on the content creator themselves, 94 00:04:55,480 --> 00:04:58,159 Speaker 1: Like do you need to be identifying that something is 95 00:04:58,200 --> 00:05:01,240 Speaker 1: AI and alerting the audience to it or is it 96 00:05:01,279 --> 00:05:02,359 Speaker 1: all on the creator. 97 00:05:03,680 --> 00:05:07,240 Speaker 2: I think that we can do more to identify content 98 00:05:07,240 --> 00:05:09,359 Speaker 2: that was created with AI. We actually when we first 99 00:05:09,400 --> 00:05:11,800 Speaker 2: did it, we were doing we were using the xf 100 00:05:11,920 --> 00:05:13,320 Speaker 2: data and a bunch of people got mad at us 101 00:05:13,400 --> 00:05:17,200 Speaker 2: because Photoshop was exporting content and the AI was part 102 00:05:17,240 --> 00:05:18,640 Speaker 2: of the flow, and then we were saying it was 103 00:05:18,680 --> 00:05:20,560 Speaker 2: AI content. But the creator was like, this is an 104 00:05:20,560 --> 00:05:23,240 Speaker 2: AI content. But if you actually looked at how the flow. 105 00:05:23,080 --> 00:05:26,280 Speaker 1: Was made, they thought Photoshop was AI. 106 00:05:26,839 --> 00:05:29,599 Speaker 2: They had used some tool and Photoshop that triggered that. Okay, 107 00:05:29,640 --> 00:05:31,320 Speaker 2: I think it was Photoshop, or maybe it was Premiere 108 00:05:31,320 --> 00:05:35,240 Speaker 2: I forget which AAP. So it's sort of a delicate matter. 109 00:05:35,440 --> 00:05:37,520 Speaker 2: For what is worth though, I think that trying to 110 00:05:37,680 --> 00:05:40,880 Speaker 2: automatically label everything that was created in some part with AI, 111 00:05:41,320 --> 00:05:43,440 Speaker 2: particularly when there's a lot of hybrid content where it's 112 00:05:43,480 --> 00:05:46,600 Speaker 2: only part of the process is a little bit of 113 00:05:47,760 --> 00:05:50,200 Speaker 2: it's not a fool's erran, but it's it's maybe the 114 00:05:50,240 --> 00:05:55,640 Speaker 2: wrong focus. I think what's gonna happened quickly is that 115 00:05:56,200 --> 00:05:58,240 Speaker 2: you can still kind of tell when content was created 116 00:05:58,240 --> 00:06:00,560 Speaker 2: by AI. Not always but usually pretty good. 117 00:06:00,600 --> 00:06:02,719 Speaker 1: It's hard, it's hard, harder. 118 00:06:02,320 --> 00:06:04,960 Speaker 2: It's getting harder, and I think pretty soon it'll be 119 00:06:05,640 --> 00:06:08,359 Speaker 2: almost impossible. And so I think what we need to 120 00:06:08,400 --> 00:06:11,840 Speaker 2: do is, sure, we can work more on labeling, but 121 00:06:11,880 --> 00:06:17,160 Speaker 2: we also need to provide more context for people to 122 00:06:17,200 --> 00:06:19,640 Speaker 2: make more informed decisions about what to believe in to 123 00:06:19,760 --> 00:06:23,320 Speaker 2: not so like the way you think about, you know, literacy. 124 00:06:23,520 --> 00:06:25,360 Speaker 2: For like my kids, my kids are young, they're nine 125 00:06:25,400 --> 00:06:28,400 Speaker 2: to seven and five. I need them to understand as 126 00:06:28,400 --> 00:06:30,520 Speaker 2: they grow up and they get exposed to the Internet 127 00:06:30,520 --> 00:06:32,880 Speaker 2: that just because they're seeing a video of something doesn't 128 00:06:32,920 --> 00:06:37,000 Speaker 2: mean it actually happened. And they when I grew up 129 00:06:37,080 --> 00:06:38,760 Speaker 2: and I saw a video, I could assume that that 130 00:06:38,839 --> 00:06:41,560 Speaker 2: was a capture of a moment that happened in the 131 00:06:41,560 --> 00:06:43,920 Speaker 2: real world. And what they're going to need to do 132 00:06:44,000 --> 00:06:45,760 Speaker 2: is they're going to need to think about who is 133 00:06:45,800 --> 00:06:48,880 Speaker 2: saying it, who's sharing it in this case, and what 134 00:06:48,880 --> 00:06:52,400 Speaker 2: are their incentives and why might they be saying it? 135 00:06:52,760 --> 00:06:55,320 Speaker 2: And there's going to be way more importance if you're 136 00:06:55,360 --> 00:06:59,800 Speaker 2: going to make good decisions about what to trust and 137 00:06:59,839 --> 00:07:04,600 Speaker 2: what to allow to inform your own opinions, we're gonna 138 00:07:04,600 --> 00:07:08,840 Speaker 2: have to shift emphasis away from the content itself and 139 00:07:09,400 --> 00:07:11,360 Speaker 2: not to ignore it, but just away from it and 140 00:07:11,480 --> 00:07:13,520 Speaker 2: put more on who's saying it and why they might 141 00:07:13,560 --> 00:07:14,040 Speaker 2: be saying it. 142 00:07:14,160 --> 00:07:16,840 Speaker 1: Yeah, but you're doing that as an involved parent. 143 00:07:16,920 --> 00:07:19,160 Speaker 2: But as a platform, we can provide context too, right, 144 00:07:19,560 --> 00:07:21,559 Speaker 2: and we do today, but we can do more like today. 145 00:07:21,600 --> 00:07:23,840 Speaker 2: If you go to my profile, you can go to 146 00:07:23,880 --> 00:07:25,880 Speaker 2: about this profile. You can see how long I've had 147 00:07:25,880 --> 00:07:29,440 Speaker 2: this profile, what country I'm based in, how many times 148 00:07:29,440 --> 00:07:32,680 Speaker 2: I've changed my user name, a series of signals that 149 00:07:32,760 --> 00:07:36,320 Speaker 2: you can use to decide how much you want to 150 00:07:36,360 --> 00:07:39,440 Speaker 2: trust or not trust. Or maybe it's not even about 151 00:07:39,520 --> 00:07:43,560 Speaker 2: positive miniccative. Maybe it's just about my political bias or 152 00:07:43,680 --> 00:07:46,600 Speaker 2: my financial bias or any other kind of bias of it. 153 00:07:46,720 --> 00:07:50,240 Speaker 1: Yeah, it feels like it's evolving. Obviously the Instagram app 154 00:07:50,320 --> 00:07:52,200 Speaker 1: is evolving as well. That was a big part of 155 00:07:52,240 --> 00:07:55,400 Speaker 1: our last conversation. I think people still think of I 156 00:07:55,440 --> 00:07:58,200 Speaker 1: shouldn't say I certainly still part of me thinks of 157 00:07:58,240 --> 00:08:02,200 Speaker 1: Instagram as the feed of beautiful, curated photos and videos 158 00:08:02,240 --> 00:08:05,120 Speaker 1: like that is maybe it's certainly an older way of 159 00:08:05,160 --> 00:08:08,480 Speaker 1: thinking about it. You're leaning very hard into this sort 160 00:08:08,480 --> 00:08:13,200 Speaker 1: of newer version of Instagram, which is DMS, stories, reels. 161 00:08:14,080 --> 00:08:16,720 Speaker 1: Walk us through a little bit about where your priorities 162 00:08:16,720 --> 00:08:19,720 Speaker 1: are with Instagram right now, Like, is the traditional feed 163 00:08:19,920 --> 00:08:20,400 Speaker 1: a thing in. 164 00:08:20,360 --> 00:08:24,600 Speaker 2: The past at this point, it's certainly changing. All the 165 00:08:24,640 --> 00:08:27,920 Speaker 2: growth in Instagram over the last five years or more 166 00:08:28,600 --> 00:08:32,880 Speaker 2: has really come from dms and reels and recommendations and stories. 167 00:08:34,360 --> 00:08:39,160 Speaker 2: But we started as a feed of hypersaturated photos, you know, 168 00:08:39,440 --> 00:08:43,720 Speaker 2: glossy landscapes. You guys can all picture the aesthetic, and 169 00:08:43,760 --> 00:08:47,480 Speaker 2: that's just not how people use Instagram anymore. Our biggest 170 00:08:47,559 --> 00:08:50,439 Speaker 2: risk is that the world evolves, we don't keep up 171 00:08:50,480 --> 00:08:53,480 Speaker 2: and we just become irrelevant. If we hadn't built dms 172 00:08:53,480 --> 00:08:57,040 Speaker 2: and stories and reels, you wouldn't be interviewing me on 173 00:08:57,120 --> 00:08:59,880 Speaker 2: stage today. We wouldn't be relevant enough to show up 174 00:09:00,080 --> 00:09:04,080 Speaker 2: at screen time. But moving it forward is also tricky. 175 00:09:04,200 --> 00:09:07,280 Speaker 2: People get frustrated when you have an when they have 176 00:09:07,320 --> 00:09:10,560 Speaker 2: an experience that they use for call it half an 177 00:09:10,559 --> 00:09:12,640 Speaker 2: hour a day, and then someone comes and changes it 178 00:09:12,640 --> 00:09:15,520 Speaker 2: without asking them. So it's a balance. But we have 179 00:09:15,600 --> 00:09:18,720 Speaker 2: to because if we didn't, we would be in trouble. 180 00:09:18,760 --> 00:09:20,760 Speaker 2: I mean, like I'll do like probably shouldn't do this 181 00:09:20,760 --> 00:09:23,520 Speaker 2: because nobody can see the audience on camera. But how 182 00:09:23,559 --> 00:09:25,240 Speaker 2: many of you use Instagram? It's okay if you don't 183 00:09:25,280 --> 00:09:27,880 Speaker 2: raise your hand really high so I can see. Cool. 184 00:09:28,400 --> 00:09:31,240 Speaker 2: How many of you keep your hand up? Have messaged 185 00:09:31,240 --> 00:09:36,280 Speaker 2: someone in the last two days? Great? How many? Now? 186 00:09:36,320 --> 00:09:38,439 Speaker 2: How many of you have posted a story in the 187 00:09:38,480 --> 00:09:42,080 Speaker 2: last couple of days? All right, so maybe about half? 188 00:09:42,080 --> 00:09:44,280 Speaker 2: From about it was like eighty percent. How many of 189 00:09:44,320 --> 00:09:46,480 Speaker 2: you posted something to feed in the last two days? 190 00:09:47,800 --> 00:09:50,280 Speaker 1: Those watching, the number of hands is very minimal. 191 00:09:50,440 --> 00:09:51,280 Speaker 2: We have seven. 192 00:09:51,360 --> 00:09:53,160 Speaker 1: We went from like almost everybody to seven. 193 00:09:53,360 --> 00:09:58,000 Speaker 2: Yeah, that's how Instagram is used. We connect with friends, 194 00:09:58,040 --> 00:10:00,920 Speaker 2: we explore our interests. But we so by seeing a 195 00:10:00,960 --> 00:10:02,760 Speaker 2: reel and sending it to someone who we think might 196 00:10:02,880 --> 00:10:05,679 Speaker 2: enjoy it, or responding to a story, or sharing a 197 00:10:05,720 --> 00:10:08,840 Speaker 2: fleeting moment, or posting a note. Actually, if you're you know, 198 00:10:09,000 --> 00:10:13,600 Speaker 2: younger than I am, it's posting to your feed is 199 00:10:13,640 --> 00:10:15,920 Speaker 2: just not the primary way people express themselves. 200 00:10:16,000 --> 00:10:18,199 Speaker 1: Yeah, and that's evidenced by some of the stuff you're doing. 201 00:10:18,440 --> 00:10:21,600 Speaker 1: So I want to get into this sort of test 202 00:10:21,640 --> 00:10:24,679 Speaker 1: you're doing with reels. You launch an iPad app, which 203 00:10:24,720 --> 00:10:26,760 Speaker 1: people have been clamoring for for a long time. If 204 00:10:26,800 --> 00:10:29,920 Speaker 1: you the iPad ad opens directly into reels, so it 205 00:10:29,920 --> 00:10:32,760 Speaker 1: doesn't open into the traditional feed, you're now kicked into 206 00:10:32,760 --> 00:10:36,319 Speaker 1: this short form video experience. You're testing that in India 207 00:10:36,440 --> 00:10:40,800 Speaker 1: as well. Ye is this the future? Like a year 208 00:10:40,840 --> 00:10:43,319 Speaker 1: from now, are we all going to open up Instagram 209 00:10:43,320 --> 00:10:45,720 Speaker 1: and instead of seeing that feed, are we just going 210 00:10:45,800 --> 00:10:48,199 Speaker 1: to be kicked into you know what? Maybe feels like. 211 00:10:48,160 --> 00:10:50,800 Speaker 2: TikTok so when you know Yeah, So the test in 212 00:10:50,880 --> 00:10:53,000 Speaker 2: India right now is an opt in and it opens. 213 00:10:53,040 --> 00:10:54,600 Speaker 2: You have stories at the top still, but if you 214 00:10:54,640 --> 00:10:58,400 Speaker 2: scroll down, it scrolls into reels. But if you look 215 00:10:58,440 --> 00:11:01,120 Speaker 2: at the reels tab and the main feed tab, there 216 00:11:01,120 --> 00:11:05,440 Speaker 2: are two primary differences. The one is full screen media. 217 00:11:05,440 --> 00:11:07,640 Speaker 2: At times you see one piece, one photo or video 218 00:11:07,679 --> 00:11:11,000 Speaker 2: at a time, and that is the reels tab, whereas 219 00:11:11,000 --> 00:11:15,240 Speaker 2: Feed is free scrolling and then reels has more videos 220 00:11:16,080 --> 00:11:18,960 Speaker 2: and more recommendations and feed has less, but they both 221 00:11:19,040 --> 00:11:22,560 Speaker 2: have recommendations. They both have photos, they both have carousels, 222 00:11:22,559 --> 00:11:24,880 Speaker 2: and they both have videos. It's just a balance, and 223 00:11:24,920 --> 00:11:27,400 Speaker 2: so it's really a design difference and then a ranking difference. 224 00:11:27,960 --> 00:11:31,240 Speaker 2: I think just swapping them doesn't really make sense. But 225 00:11:31,360 --> 00:11:35,640 Speaker 2: I think that if a more immersive media is really 226 00:11:35,679 --> 00:11:37,760 Speaker 2: driving all our growth, if we can figure out a 227 00:11:37,840 --> 00:11:41,280 Speaker 2: way to transition the app to that being the experience 228 00:11:41,280 --> 00:11:45,120 Speaker 2: that you swipe into, then that would probably be a 229 00:11:45,120 --> 00:11:46,800 Speaker 2: good thing if we can make it happen. The questions 230 00:11:46,840 --> 00:11:51,079 Speaker 2: can we make it happen responsibly without you pissing everybody off, 231 00:11:51,120 --> 00:11:53,959 Speaker 2: without really hurting the business because the monetization efficiency of 232 00:11:54,040 --> 00:11:57,040 Speaker 2: video is lower by finding the right balance. It's probably 233 00:11:57,040 --> 00:11:59,800 Speaker 2: not ranked exactly as one is or the other is today, 234 00:12:00,440 --> 00:12:02,200 Speaker 2: but that is definitely something I want to explore. 235 00:12:02,400 --> 00:12:04,040 Speaker 1: Yeah, I mean, from the outside, it's hard not to 236 00:12:04,080 --> 00:12:07,240 Speaker 1: look at that and think tik TikTok is driving this 237 00:12:07,440 --> 00:12:10,440 Speaker 1: change for you. How much do you think about TikTok 238 00:12:11,400 --> 00:12:13,560 Speaker 1: as a It's obviously a major competitor, But I'm wondering 239 00:12:13,679 --> 00:12:17,680 Speaker 1: how much something like this is driven maybe by what 240 00:12:17,679 --> 00:12:19,880 Speaker 1: you're seeing from competition. 241 00:12:19,880 --> 00:12:22,080 Speaker 2: Curious about this data, what is this data aduls eighteen 242 00:12:22,120 --> 00:12:25,000 Speaker 2: to thirty four average audience by hour. I don't know 243 00:12:25,040 --> 00:12:30,080 Speaker 2: what that means. Cool. Definitely think about TikTok, Definitely think 244 00:12:30,120 --> 00:12:33,800 Speaker 2: about YouTube. Definitely think about Snapchat. I mean, the immersive 245 00:12:33,880 --> 00:12:39,319 Speaker 2: viewer is pre dates TikTok, right. I think Cacao Stories 246 00:12:39,400 --> 00:12:42,320 Speaker 2: pioneered the stories format, which was an immersive viewer. Yeah 247 00:12:42,559 --> 00:12:46,760 Speaker 2: in Korea. It was popularized by Snapchat worldwide, so they 248 00:12:46,760 --> 00:12:49,439 Speaker 2: should get a lot of credit for that. TikTok was 249 00:12:49,480 --> 00:12:54,160 Speaker 2: applying that immersive viewer that native first experience too short 250 00:12:54,160 --> 00:12:57,520 Speaker 2: from video, and they should get a lot of credit 251 00:12:57,520 --> 00:12:59,720 Speaker 2: for that. And now obviously you've got YouTube shorts, Instagram 252 00:12:59,760 --> 00:13:02,360 Speaker 2: real Facebook reels and the rest of it. We think 253 00:13:02,400 --> 00:13:04,480 Speaker 2: a lot about TikTok because they are really good at 254 00:13:04,520 --> 00:13:08,120 Speaker 2: what they do. They are better than I think the 255 00:13:08,160 --> 00:13:09,800 Speaker 2: rest of the industry at this point at a couple 256 00:13:09,920 --> 00:13:13,720 Speaker 2: things at breaking new talent at. 257 00:13:14,160 --> 00:13:16,800 Speaker 1: In like identifying up and coming creators. Is that what 258 00:13:16,800 --> 00:13:17,079 Speaker 1: you mean? 259 00:13:17,240 --> 00:13:21,360 Speaker 2: Yeah? So the way it works is when you're ranking 260 00:13:21,480 --> 00:13:24,439 Speaker 2: content and you want to drive engagement, it's easier to 261 00:13:24,520 --> 00:13:27,160 Speaker 2: drive engagement by showing content that you know lots of 262 00:13:27,200 --> 00:13:31,120 Speaker 2: people like than to try and discover content that you 263 00:13:31,200 --> 00:13:33,040 Speaker 2: might love that you didn't even know about yet, that 264 00:13:33,120 --> 00:13:35,320 Speaker 2: might be from a really small name or small creator. 265 00:13:36,000 --> 00:13:38,600 Speaker 2: And that's called exploration based ranking, when you try to 266 00:13:38,640 --> 00:13:43,200 Speaker 2: literally almost test out things to see if they have legs, 267 00:13:43,520 --> 00:13:47,120 Speaker 2: if there's an audience for them. They are really good 268 00:13:47,120 --> 00:13:51,240 Speaker 2: at exploration based ranking. We're closing that gap, but we're 269 00:13:51,240 --> 00:13:54,320 Speaker 2: not all the way there yet. YouTube isn't actually great 270 00:13:54,400 --> 00:13:56,480 Speaker 2: at that. YouTube is the best I think in the 271 00:13:56,480 --> 00:14:01,520 Speaker 2: market at paying out creators through revshare. That's because they're 272 00:14:01,679 --> 00:14:04,240 Speaker 2: still mostly long form and long form. They've done a 273 00:14:04,240 --> 00:14:07,400 Speaker 2: really good job with their in or their pre rules 274 00:14:07,400 --> 00:14:10,719 Speaker 2: and their mid roles. But I think TikTok has been 275 00:14:10,800 --> 00:14:13,560 Speaker 2: like really culturally relevant because they've been able to break 276 00:14:13,600 --> 00:14:17,400 Speaker 2: small names. They've been really timely, so trends happen really quickly, 277 00:14:17,640 --> 00:14:20,280 Speaker 2: and they've just been reliably really fun and entertaining. 278 00:14:20,400 --> 00:14:25,520 Speaker 1: Yeah, there's obviously I'm sure you followed. There's a deal now, 279 00:14:25,640 --> 00:14:28,880 Speaker 1: a proposed deal for TikTok to stay in the United States. 280 00:14:29,640 --> 00:14:32,120 Speaker 1: It's kind of unique because it calls for the US 281 00:14:32,200 --> 00:14:36,320 Speaker 1: government to be somewhat involved and you know, if not managing, 282 00:14:36,360 --> 00:14:41,560 Speaker 1: at least overseeing this new US TikTok does this worry 283 00:14:41,640 --> 00:14:45,000 Speaker 1: you at all? Like that the US government might perhaps 284 00:14:45,400 --> 00:14:48,800 Speaker 1: have an incentive for TikTok to do better than Instagram. 285 00:14:51,800 --> 00:14:52,960 Speaker 2: Be careful of my words. 286 00:14:53,960 --> 00:14:57,200 Speaker 1: We're all friends here. You're welcome to say whatever you want. 287 00:14:57,520 --> 00:15:03,360 Speaker 2: What could go wrong? Yeah, I'll say a few things. 288 00:15:04,000 --> 00:15:06,280 Speaker 2: I do think it's good that TikTok is still going 289 00:15:06,320 --> 00:15:08,280 Speaker 2: to be in the US. It would certainly make our 290 00:15:08,360 --> 00:15:11,520 Speaker 2: lives easier on the professional side to have less competition, 291 00:15:11,600 --> 00:15:14,000 Speaker 2: But I think competition is fundamentally healthy. They've pushed us 292 00:15:14,000 --> 00:15:15,920 Speaker 2: to do better work. I think it's good for the 293 00:15:15,960 --> 00:15:18,800 Speaker 2: consumer and it's good for competition, So I'm glad that 294 00:15:18,800 --> 00:15:21,320 Speaker 2: they're still going to be around. On the specifics of 295 00:15:21,360 --> 00:15:26,120 Speaker 2: the deal, it just seems all very hard to parse. 296 00:15:26,960 --> 00:15:31,240 Speaker 2: It looks like more or less there's a bit of 297 00:15:31,280 --> 00:15:34,440 Speaker 2: an ownership and a rev share change, but not an 298 00:15:34,480 --> 00:15:39,400 Speaker 2: actual real change in how it's built and in you know, 299 00:15:39,440 --> 00:15:43,920 Speaker 2: because you know by dancing TikTok share code. Historically, most 300 00:15:43,960 --> 00:15:46,240 Speaker 2: of the most the strongest ranking work has come out 301 00:15:46,240 --> 00:15:51,360 Speaker 2: of China. Like, I don't see anything that suggests that 302 00:15:51,440 --> 00:15:54,680 Speaker 2: any of how it's actually built it's going to meaningfully change. 303 00:15:55,080 --> 00:15:58,680 Speaker 2: It just seems like an ownership stake and a rev 304 00:15:58,720 --> 00:16:04,240 Speaker 2: share going that's split between obviously by dancing the new 305 00:16:04,280 --> 00:16:07,360 Speaker 2: owners here. But it's the same map, the same ranking system, 306 00:16:07,920 --> 00:16:10,680 Speaker 2: you know, the same creators that you're following, the same people. 307 00:16:10,720 --> 00:16:13,520 Speaker 2: It all sort of seamless. So it seems a little 308 00:16:13,520 --> 00:16:15,840 Speaker 2: bit hard to I'm not sure what to make of that. 309 00:16:15,880 --> 00:16:18,040 Speaker 2: It doesn't seem like it's a major change in terms 310 00:16:18,120 --> 00:16:20,960 Speaker 2: of incentives. Yeah, I mean we always want to be considered. 311 00:16:21,640 --> 00:16:23,720 Speaker 2: We always want to consider all the incentives of the 312 00:16:23,720 --> 00:16:26,840 Speaker 2: most relevant parties involved. But I just want to stay 313 00:16:26,840 --> 00:16:28,720 Speaker 2: focused on how do we make sure Instagram does a 314 00:16:28,720 --> 00:16:31,160 Speaker 2: great job where they're better than us, how do we 315 00:16:31,240 --> 00:16:33,880 Speaker 2: catch up where we think we've got a competitive advantage, 316 00:16:33,880 --> 00:16:36,840 Speaker 2: how do we double down? And how do we keep 317 00:16:37,000 --> 00:16:39,280 Speaker 2: moving forward and trying to make Instagram as valuable as 318 00:16:39,320 --> 00:16:42,920 Speaker 2: possible for people, and then trying not to get too 319 00:16:43,000 --> 00:16:46,720 Speaker 2: sucked into the noise of the of the drama of. 320 00:16:46,720 --> 00:16:50,200 Speaker 1: The week, maybe supporting your direct competitors, all that stuff. 321 00:16:50,240 --> 00:16:52,040 Speaker 2: We think about it, but we try to focus on 322 00:16:52,080 --> 00:16:54,800 Speaker 2: what we can control and spend less time on what 323 00:16:54,840 --> 00:16:55,320 Speaker 2: we cannot. 324 00:16:55,720 --> 00:16:57,440 Speaker 1: We'll only have a few more minutes. It's always flies 325 00:16:57,480 --> 00:17:01,040 Speaker 1: by there was a report that you guys are building 326 00:17:01,080 --> 00:17:05,160 Speaker 1: a TV app? Are you building a TV app? Let's 327 00:17:05,240 --> 00:17:05,840 Speaker 1: make some news. 328 00:17:06,160 --> 00:17:09,879 Speaker 2: This is it. We're exploring TV. TV is an increasingly 329 00:17:09,920 --> 00:17:14,560 Speaker 2: important surface. It's been very important for YouTube. Neil's talked 330 00:17:14,560 --> 00:17:17,439 Speaker 2: about that publicly, the CEO there. It's been very important 331 00:17:17,440 --> 00:17:20,240 Speaker 2: for TikTok. So we'd like to figure out how to 332 00:17:20,240 --> 00:17:23,320 Speaker 2: make sure that we have we show up in a 333 00:17:23,320 --> 00:17:27,320 Speaker 2: compelling way on all the relevant devices. So we're definitely 334 00:17:27,320 --> 00:17:28,920 Speaker 2: exploring it, but nothing to announce today. 335 00:17:29,600 --> 00:17:33,280 Speaker 1: Why now, though, Why explore it now versus a few 336 00:17:33,359 --> 00:17:35,680 Speaker 1: years ago, because, as you point out. 337 00:17:35,560 --> 00:17:37,840 Speaker 2: I wish we had explored a few Okay, why didn't you? 338 00:17:38,760 --> 00:17:44,440 Speaker 2: We've been busy. It seems like every year is a 339 00:17:44,440 --> 00:17:46,480 Speaker 2: little bit more intense and a little bit more wild. 340 00:17:47,640 --> 00:17:49,640 Speaker 2: I see that as a mistake though on my part, 341 00:17:49,720 --> 00:17:53,240 Speaker 2: so I'll take ownership for that. I know we've been 342 00:17:53,240 --> 00:17:55,800 Speaker 2: focusing a lot on ranking. We've been focusing a lot 343 00:17:55,800 --> 00:17:58,360 Speaker 2: on evolving the experience forward. We've been focusing a lot 344 00:17:58,400 --> 00:18:01,240 Speaker 2: on safety and parents. We've been focusing a lot of 345 00:18:01,400 --> 00:18:05,280 Speaker 2: just compliance. Like I said, the biggest risk that we 346 00:18:05,400 --> 00:18:08,320 Speaker 2: face is that the world changes too quickly. Not too 347 00:18:08,359 --> 00:18:10,800 Speaker 2: quickly it's going to change quickly, and we don't change 348 00:18:10,880 --> 00:18:15,040 Speaker 2: quickly enough to adapt. And if behavior in the consumption 349 00:18:15,119 --> 00:18:17,680 Speaker 2: of these platforms is moving to TV, then we need 350 00:18:17,680 --> 00:18:20,399 Speaker 2: to move to TV too as a form factor. And 351 00:18:20,440 --> 00:18:21,720 Speaker 2: if we're late, that's on us. 352 00:18:22,720 --> 00:18:26,919 Speaker 1: Right now, on you, right now, these videos. 353 00:18:26,920 --> 00:18:28,840 Speaker 2: Your deep voice that sounded it did. 354 00:18:28,720 --> 00:18:31,920 Speaker 1: Say it came off a little more dramatical. When I'm 355 00:18:31,920 --> 00:18:34,520 Speaker 1: on Instagram looking at video, it's almost all vertical video. 356 00:18:34,680 --> 00:18:37,919 Speaker 1: When I look at a TV, it's not. Have you 357 00:18:38,280 --> 00:18:42,159 Speaker 1: through this exploration, have you thought about whether the stuff 358 00:18:42,160 --> 00:18:45,040 Speaker 1: that's on Instagram today works on TV? Or do you 359 00:18:45,080 --> 00:18:48,080 Speaker 1: need to go out and you know, basically source new 360 00:18:48,160 --> 00:18:50,520 Speaker 1: video that fits a TV screen. 361 00:18:50,960 --> 00:18:53,160 Speaker 2: I think the vertical content can work on TV because 362 00:18:53,160 --> 00:18:55,800 Speaker 2: I think we can also show there's a couple, we 363 00:18:55,800 --> 00:18:58,040 Speaker 2: can show social context, we can show you the comments, 364 00:18:58,119 --> 00:18:59,760 Speaker 2: we can show us as watching. There's a bunch of 365 00:18:59,800 --> 00:19:02,040 Speaker 2: things you can do with that space. We might we 366 00:19:02,080 --> 00:19:04,280 Speaker 2: do still have horizontal content. We might need to lean 367 00:19:04,320 --> 00:19:07,280 Speaker 2: into that too, but I'd rather first figure out how 368 00:19:07,280 --> 00:19:12,879 Speaker 2: we can embrace the content ecosystem that already exists and 369 00:19:12,920 --> 00:19:15,399 Speaker 2: the creatives that already exist on the platform. And if 370 00:19:15,400 --> 00:19:17,760 Speaker 2: we have to expand. If it turns out we have 371 00:19:17,800 --> 00:19:19,280 Speaker 2: to expand, then we'll certainly consider that. 372 00:19:19,320 --> 00:19:20,879 Speaker 1: Then it was going to say, it's our world in 373 00:19:20,920 --> 00:19:25,560 Speaker 1: which you go and you know, license more kickback content. 374 00:19:26,000 --> 00:19:29,480 Speaker 2: I don't like licensing content if I can avoid it. 375 00:19:29,640 --> 00:19:32,560 Speaker 2: The look, we pay for lots of licensing in general, 376 00:19:32,640 --> 00:19:34,359 Speaker 2: for people are going to share stuff on our platform, 377 00:19:34,359 --> 00:19:36,480 Speaker 2: whether or not we want them to, music licenses, et cetera. 378 00:19:36,560 --> 00:19:38,040 Speaker 2: We're always going to try to make sure we pay 379 00:19:38,080 --> 00:19:40,800 Speaker 2: people for the rights, you know, when we need to. 380 00:19:41,280 --> 00:19:43,720 Speaker 2: But I wouldn't. I'm not interested in trying to go 381 00:19:43,800 --> 00:19:48,960 Speaker 2: get sports rights or exclusive entertainment rights because I'd much 382 00:19:49,040 --> 00:19:50,920 Speaker 2: rather be the second screen. I'd much rather be where 383 00:19:50,960 --> 00:19:55,880 Speaker 2: you talk about the game or the movie we're in. 384 00:19:56,640 --> 00:19:59,000 Speaker 2: I feel like connecting with your friends and short from 385 00:19:59,080 --> 00:20:02,000 Speaker 2: video can be really simp iiotic, and short form video 386 00:20:02,040 --> 00:20:04,800 Speaker 2: and long form video can be really symbiotic. But long 387 00:20:04,840 --> 00:20:06,840 Speaker 2: form video and connecting with your friends it's really difficult 388 00:20:06,880 --> 00:20:09,439 Speaker 2: to do in one place. If you're just watching a 389 00:20:09,480 --> 00:20:12,560 Speaker 2: basketball game or a movie or an hour long show, 390 00:20:12,800 --> 00:20:14,800 Speaker 2: it's a lot of time that you're not actually talking 391 00:20:14,840 --> 00:20:17,119 Speaker 2: to a friend. A lot of things you're not sending 392 00:20:17,119 --> 00:20:18,640 Speaker 2: to friends or seeing that you might send to friends, 393 00:20:18,640 --> 00:20:23,639 Speaker 2: et cetera. So I'd rather focus on user generated content 394 00:20:23,760 --> 00:20:29,919 Speaker 2: creators AI unless on premium long form licensed content. 395 00:20:30,040 --> 00:20:32,879 Speaker 1: We've hit all the zeros very last, so that I 396 00:20:32,880 --> 00:20:35,639 Speaker 1: don't get in trouble. What's your one tip if you're 397 00:20:35,680 --> 00:20:38,879 Speaker 1: a creator trying to hit a big on Instagram today, 398 00:20:38,960 --> 00:20:42,399 Speaker 1: what's the one piece of advice you would give? And 399 00:20:42,440 --> 00:20:44,000 Speaker 1: again short, so that I don't get you. 400 00:20:44,040 --> 00:20:45,720 Speaker 2: Know, yeah, you don't get too much trouble out and 401 00:20:45,840 --> 00:20:47,160 Speaker 2: I can just fill a buster and go. 402 00:20:47,240 --> 00:20:49,040 Speaker 1: Yeah, exactly, get me back. 403 00:20:49,520 --> 00:20:53,600 Speaker 2: I think it's really important to be intentional, specifically to 404 00:20:53,960 --> 00:20:57,600 Speaker 2: articulate what you're trying to accomplish by using a platform 405 00:20:57,600 --> 00:21:00,760 Speaker 2: like Instagram. So you might be trying to sell a 406 00:21:00,800 --> 00:21:04,240 Speaker 2: product or brand, or raise awareness for a cause or 407 00:21:04,800 --> 00:21:08,440 Speaker 2: build up demand for a brand. What just write down 408 00:21:08,600 --> 00:21:11,040 Speaker 2: what you're trying to accomplish, and then also to be 409 00:21:11,119 --> 00:21:14,399 Speaker 2: very clear about what is going to actually resonate with 410 00:21:14,440 --> 00:21:17,239 Speaker 2: an audience, what's going to work on the platform. And 411 00:21:17,280 --> 00:21:20,320 Speaker 2: you need to find a content strategy that is at 412 00:21:20,359 --> 00:21:23,400 Speaker 2: the overlap of those two things, because reach is great, 413 00:21:23,440 --> 00:21:25,199 Speaker 2: but reaches a means to an end. So if you 414 00:21:25,280 --> 00:21:27,800 Speaker 2: find content that is engaging, but it's not actually helping 415 00:21:27,840 --> 00:21:31,320 Speaker 2: you move towards your goals, it's useless. And if you 416 00:21:31,760 --> 00:21:33,280 Speaker 2: find content that is going to help you with your 417 00:21:33,320 --> 00:21:35,080 Speaker 2: goals but no one sees it, it's not going to 418 00:21:35,080 --> 00:21:37,960 Speaker 2: be effective. It is much harder to find a content 419 00:21:38,000 --> 00:21:41,440 Speaker 2: strategy that does both. But that's the job. That's the work, 420 00:21:41,920 --> 00:21:45,760 Speaker 2: and that's going to require persistence, intentionality, and experimentation. But 421 00:21:45,840 --> 00:21:48,400 Speaker 2: I really, I really preach the overlap