The Exchanges

Q Are there any other ways that AI work in your products that I didn't touch on?

A Sure. Uh, we have a few. Uh, so we, uh, use AI in shine on our photos app. To, uh, to do things like deduplicate to overall, uh, understand, um, and, and understand when you might be doing something that's share worthy, either on a photo level or on an album level. You mentioned earlier, the suggested albums. It's one of my favorite features. The fact that the AI can figure out where I tend to take interesting group photos and where I don't and make those kinds of suggestions to me is something that relies on AI. In our events app, we're using generative AI, so different type, um, to make beautiful invitations, and we do have gotten rave reviews for our invitations. They are, my view is they are the best out there in terms of really helping build a mood and a theme and anticipation for your event. Uh, and in things like Sunshine Contacts, we're deploying, ah, text recognition and pattern recognition to overall do extraction around signatures, so we can figure out which of your contacts is this person, and we can ultimately pull in their professional information, phone numbers, ah, and all of those pieces to really make your contact that much more up to date, ah, and enriched.

AI assessment note: “we use AI in shine on our photos app... In our events app”

Q would be one answer that Google might give that was better than the others. And that helped build Google's like long-term trust. And Right now we're in this moment where these models, these AI models are on their way to commoditization. So I'm curious, like what you think will enable these models to differentiate themselves? Is it that better answer every 30 tries or something, or is it something else?

A I think that that brings me back to the first question you asked in the podcast around the stickiness of consumer apps. I think that one of the things we learned as we Built Google and it grew is that yes, there was a lot, there was kind of a head of queries, the crazy people do all the time, and didn't matter what search engine you used for those. They were all pretty much the same. It's the long tail of queries, deep research queries, queries that people had never done before, where Google's flexibility and depth and comprehensiveness and speed really set us apart. We, you get much better answers there. That said, you know, when we saw it, like the difference between us and even the second best search engine was sometimes pretty narrow. Three percent, five percent, as you said, at which point in time you're saying, well, look, you're really only getting a better answer With Google, once out of every, one out of every 20 tries, or one out of every 30 tries. And two things we saw. One, once you become part of someone's routine, the odds they keep turning to you as their search engine is really high. So, as I said, consumer apps are stickier than people realize. The other is that people are so grateful when they get a novel, insightful answer or result, That they, it builds a huge amount of allegiance. The fact that you can just really nail someone's query, you know, one in 30 t…

AI assessment note: “The fact that you can just really nail someone's query, you know, one in 30 times”

Q that this is the direction and with that context, do you think there's a bright future for the web? I mean, there's, there are browsers now that you'll go to the webpage and it just, instead of having you read the article, we'll summarize it for you. So I guess I'm curious if you think that this is the future and if it is what it means for the web.

A Uh, I think that summarization is going to be really important. That's why I bought Summly. I think it ultimately makes you a lot more efficient in your daily routines that helps you consume a lot more content and also ultimately understand where do I want just a surface level summary and where do I want to do a deep dive? Uh, so I think that summarization technology is incredibly helpful. And I also think we, uh, you know, one of the big reasons we bought it was to apply it on video because there's definitely times where you can't watch a video, um, or, you know, you can't listen to a podcast, but if you can basically take all that content and summarize it down, uh, is ultimately really useful. Uh, I do think that one of the concerns I have about the web in the long run is with machine generated content. I'm not confident that the web improves. Maybe it stays the same. Maybe it declines slightly, but for a lot of these technologies, we need the web and we need this base of knowledge and data that everything's learning off of to continue to get better. And I do think that's going to be one of the challenges in the future as we continue to train models and advance these technologies is how do we make sure that we're training on things where the quality is actually monotonically increasing as opposed to staying the same or decreasing.

AI assessment note: “one of the concerns I have about the web in the long run is with machine generated content”

Q users will just turn. So do you think that he's, he's over hyping the challenge? I mean, it's, I don't think that he's right in saying that this stuff is baked, but the challenge is great, right? Because if you think about The consumer apps that have broken through, let's say since, I don't know, Instagram, uh, TikTok is, is TikTok the only big one? What's your thought on this?

A Uh, I think that there are in theory that, you know, there are more apps that have broken out in my mind. I do think that he's right. There's a big barrier to entry in some of these cases, especially when you talk about video, because video is, you know, it's expensive to move around is expensive in terms of user time, uh, in terms of how people consume it. It's cereal as opposed to parallel. So, you know, the way you have to engineer the stream and things like that, just to make sure that you really hold attention is more intense. So I think he's got some good points, uh, in terms of TikTok and video in particular. That said, I think that, again, if you look at that core of what people are doing in consumer, what are you helping them do? And if you're helping them do something that they need to do anyway, and you're helping them do it better and with new innovation, I think that that is really greenfield, uh, because I think people are always looking for ways to be more efficient. They like trying things new, right? We are creatures that love novelty, and so I think that those types of things have a real chance of, of, of breaking through, and if there's something that helps you handle your everyday tasks, and something that we really like to focus on here at Sunshine, I think the odds that it becomes part of your routine and a critical part of your routine is, is, is there.

AI assessment note: “I do think that he's right. There's a big barrier to entry”

Q Right. And so I'm putting all these different areas that you're working in together, and I'm looking at, okay, so it's photos, contacts, events. To me, it seems like it's almost breaking down what Facebook used to do well, and trying to build it back in a different, maybe more of a utility versus a social network. What do you think about that thought?

A You know, as I said, for us, we're very focused on what we think people need every day, what we think will delight them, what we think will make their relationship stronger. And I do think that there were things and are things that, that Facebook does really well for us. We're not so much following that template as opposed to just looking at people's everyday tasks. And overall, what's getting in the way? And I think we can all agree today, photo sharing is really broken. The number of times you might say, oh, let me take that on my phone too, or, you know, don't forget to send me that photo, uh, when you're parting, you know, at the end of, of time together. You know, those types of things, and the odds that you get the photo is probably pretty low. Um, maybe for some people it's higher, but I think that it's overall pretty low, and so we wanted to say, look, if, A photo really belongs to the time and a place and the people who were there. It should be easy, especially among friends and people who are connected, even through other people, to seamlessly share those photos. And at the end, you know, Right now, AI is working on things like global facial recognition, but for a lot of us, we spend a lot of our time pinching and zooming and trying to make sure that we have the, the, the, the shot that has the right expression, the eyes open, all of those types of things. And those a…

AI assessment note: “We're not so much following that template as opposed to just looking at people's everyday tasks.”

Q down before this conversation, I wrote, you know, you had just tweeted about Gmail's 20th anniversary. You mentioned it here. And I wrote a note. Could that happen today? Like, where is that today within Google? We don't see the same experimental projects coming out of the company that we used to, at least as far as I can tell. So what makes you think the boldness is still there?

A I have to say, even when we started Gmail, which Gmail launched in 2004, but started in 2002, took a lot of effort. I remember when they first had like, we're going to build an email client. We were all like, no, like we're building a search engine. Like what part of email is interesting. It was actually hard to get it off the ground and it's had to start kind of a skunk works project that was kicked off. Uh, by Larry. Uh, and you know, over time we started to see the value in it and we started to see what search and some of the profoundness of the size of storage and the, our cost of storage could really bring to bear on email. And that was all very exciting. Uh, and so, but in the beginning it took, it took a lot of energy, right? To, to start something new, even in a little company, um, you know, breaking out of that status quo takes a lot of energy. Um, and it takes courage, but I think Google has both that energy and that courage. And I think, as I said, this, this space is something that they've been, been working in and preparing. And this moment is something they've been preparing for, for a very long time. Uh, you know, everything from a lot of the underlying advances of the underlying insights. Um, there was a great article last week, uh, the transformers talking about the eight people who've really shaped modern AI, all of whom worked at, at Google. Uh, you know, thi…

AI assessment note: “I think Google has both that energy and that courage. And I think, as I”