YouTube Content

Hey everybody, thank you so much for starting your morning with us. We've got an incredible panel for you. My name is Will Black. I'm a partner here at Foundry Code West, startup lawyer. Work primarily with consumer-facing companies like many of you in the audience. We work with companies from idea stage all the way through the eventual IPO and M&A, everything and anything in between. Working with consumer has been such a privilege for me over the years because we've gotten to work with companies like when we were really excited about websites. And then we— everyone got smartphones, we were really excited about apps. And now we're at the next massive platform shift, AI. And we get to hear from the people that are completely reshaping the consumer experience, the human experience for all of us. And I have the great privilege of working with one of those companies here today, Jenny Wang of Alta. So I'll hand it over to you.

Yes. Thank you.

Thank you so much, Will. Thank you all so much for coming. I mean, Fenwick has been such an amazing partner for us and I'm so excited to introduce our moderator for the event, Emma Hinchliffe, who is the most powerful woman editor at Fortune, is an incredible journalist who I really admire. And then all of the rest of our panelists, Eliza from Corner— Eliza from Corner, Fai from Status, and Claire from Gigi, who are all building amazing companies and are all founders I admire greatly. So, So with that, I'll hand it to you.

Hi, everyone. Good morning. Thank you so much for being here. I'm excited to talk about consumer AI. It'll be really fun. And we have an amazing group of founders here. I'll let them introduce themselves a little more as we dive into it. But I'd love to start just talking about what AI means in consumer today. So you're building a consumer product today. How does that look different than what it would have been like you know, 3 years ago or pre-GenAI? What can you do that you couldn't do, and what is not relevant? What shouldn't you do? Eliza, maybe you can start and tell us a little bit about Corner.

Yes. Hi everybody, I'm Eliza. I'm the founder, one of the founders of Corner. With Corner, our thesis is that the map should be personalized for everyone. So no 2 people should have the same map. You all have different interests, you're doing different things, you're at different stages in your life. So we're really focused on personalizing that your real-world experience and building, like, an algorithm for the real world, basically. Yeah, what building an AI means to us is, I think it's so important to take a hard look at your problem and understand why has this not been solved before and why is it uniquely able to be solved with AI. So for us, we really focus on place understanding, right? Like, so people like Gen Z and young millennials, we don't want to search by Italian food, $2 sign, 4 stars. We want vibes and we want aesthetics and we want to curate our own unique taste and experience in the world. And AI lets us take all of these reviews and text and images and we can put all of it in and output really awesome data around, you know, it's a mid-century modern cafe with like a very poppin' community and the people here wear very stylish clothing. And the types of people who would wanna go there are very different than the types of people who would rather like a quick coffee in and out. Yeah, so that's, that's how we use AI.

That's amazing. Fai, tell us a little bit about how you're using AI and your company.

Hey everyone, my name is Fai. I'm CEO and co-founder of Status, which is an AI-powered social media game where our users live out their dream lives. So they can be a famous pop star, famous actress, soccer player, in the Harry Potter universe. And I think what One thing that's really interesting about how AI is changing consumer behavior, at least for our users who are super young teenagers, they now have the ability to really live out these alternate realities or dream lives because you can really generate and feel like you're in, have escaped reality and are that person or become that better person of yourself that you can translate into real life, but it also can be just a really fun game and improve behaviors that we see right now as well.

And how are you using AI? Like, what does— I mean, when I think of game chasing fame, like that experience, I think of Kim Kardashian Hollywood. So how is what you're doing different than what that experience was like?

Yeah, so instead of it being like a 3D mobile game or Or something like Kim Kardashian Hollywood. We're simulating a new social media feed for you. So instead of when you go on your Instagram or you go on your Twitter, you usually come on as your real-life self. We're like, hey, what if you— we're going to help you experience what social media would be like if you were someone completely different. Or if you lived in a different world, right? I think a lot of our users are really interested in fandoms like Harry Potter or anime, K-pop, et cetera. they are literally able to be a K-pop idol in status. And it looks like a real social media feed. It looks like Twitter. So it's cool.

That's super fun. Gigi, tell us about what you're building.

So I'm Clara Gold. So I'm French. So you're going to see that my English actually sucks. But anyways, so I'm building Gigi, who's an AI basically who knows everything about everyone and who wants to connect them better. And so I'm really thinking of what's the next AI social platform. And I think AI is completely disrupting the way we're going to think about social interactions in so many ways. So, A, now we have, thanks to LLM and what you were saying actually, a complete understanding of human beings that wasn't possible before. And before it was declarative, now it can be entirely non-declarative. And we can understand very weak signals that were not possible before, like the vibe. Before you couldn't get the vibe. And I don't know how you use ChatGPT. He's like my mentor, my boyfriend, my whatever. And I've never felt that understood in my entire life, to be very transparent. And I think AI really helped us make sense out of ourselves. And if AI can do this with everyone, then you have a new unlock, which is like, can we create an intermediary that can get all the context from everyone and then do things? Because the second unlock is just agentic behaviors. Before, we were doing the stuff ourselves, and now we can just give all this intent and all this context and have AI just go and do things for us. The third unlock for me, and I think this is the biggest one when you're building consumers, there's no cold start anymore. Building consumer social products was so harsh because you had a huge cold start and you had no graph whatsoever. And I think the biggest unlock that we saw is You can already create a lot of things about the people because the data is already there. And now we have the ability to just structure insane amount of data and put it out there for people. So my 2 cents on the matter of how AI is going to disrupt is like, A, we have no clue and we're just iterating like crazy and trying to see how it could work. But the second thing is now we know that there are 800 million people talking every day to ChatGPT. This is the biggest unlock. And to ChatGPT, meaning someone— and I say someone on purpose because it feels like fucking human— someone who really understands you and can then go doing things for you. So the way— and just to wrap it up, with Gigi, I think my biggest obsession right now is how can we make sense out of our digital lives in a very, very granular way? And not only like, you're the founder and you're whatever, but what's my vibe? And I'm very non-politically correct. You're going to see that. And so I want people to know that because this is part of my personality and this is probably how I'm going to— Connect to a different level with people. How can I already create a new graph of people? Because, you know, Facebook invented 20 years ago people you may know. But now I can pretty much predict, like, who are the people you should know? And I should have actually prepared, you know, the event and just send you everyone your profile and who are the top 3 people that if you don't talk to them today, you're going to miss the biggest opportunity of your life. No pressure, no FOMO. But so that's pretty much what it unlocks. And just agentic search, I think for me, this is the biggest thing, like just bringing all the mess in our heads, intents. Now you don't have to write the content, like AI is going to write it for you and you're going to digest the content through AI contextualized for you. So like static content is going to die and we're just going to get in a world with dynamic content. So we're going to go from a world with, you know, from dynamic feeds to dynamic content. What does it mean?

I have no clue.

I want to chat 20 seconds to an AI and then she goes and does things for me, introduces me to people, creates content for me, sends me places. I want to tell— and I'm probably— you're probably working on that, but like, hey, I want to go out tonight. You know exactly my vibe. You know exactly the vibes of my friends. Tell us where we should go. Anyways, I shut up, but this is pretty much the end look. But we have no idea how it's going to perform.

That's very exciting. Oh my gosh. Well, let's get into that more. Thank you, Clara. And Jenny, thank you again for bringing us all together. Tell us about how AI is changing the shopping world, the styling world, the fashion world, and how you're doing that at Alta.

Yeah, absolutely. Hi everyone. I'm the founder of Alta. We're building an AI styling/shopping companion. And the way I think about AI is, so what Alta does, it is Alta will generate shopping recommendations as well as outfit recommendations for you based on the weather, the calendar, anything you have going on in your life, upcoming trips. And then it will also put the recommendations on an avatar of you. So it's really trying to reimagine the future of shopping. And I have tried actually building a version of this in the past a few years ago, and the technology just wasn't there.

I was trying to train neural networks and Fashion is such a complex dataset.

Like, even when you look at my shirt today, like, there's like lace detailing, there's a bow here, like part of this is sheer, like all these sorts of nuances that you need sort of multimodal transformer-based models in order to understand correctly. And so I think a big unlock for like why AI for us is just the fact that a lot of what we're building right now could not have been done 3 months, 6 months, 2 years ago. And A lot of our current pipeline is we have to nail every part in order to try to understand how to best surface recommendations for someone. E-commerce is such a huge industry. Being able to nail personalization discovery is a massive opportunity, but it's a lot of AI models that need to execute correctly in order for that to happen, whether that's on the computer vision tagging part of fashion or on the personalized recommendation system that we serve to all of our users. And similar to what Eliza said, For us, if you search, I'm going to a date night in Milan, and then another person searches, I'm going to a date night at Casa Cipriani, based on your vibes, your aesthetic, what we know about you, the recommendation is very different. Versus if you go on nordstrom.com right now and you search for outfit for Wimbledon, you're going to get a book that says Wimbledon. I tried doing that recently. And the search result is the same for everyone. So I'm excited about the technology for us. I think the technology is still getting better and there's a long way to go. We measure ourselves by how good our recommendations are, how far the Alta algorithm matches against our internal eval sets. I think we're still not there yet to the quality and the accuracy that I want to, but that's the exciting thing is that everyone is building towards an AI that can predict exactly what you want, what you want to buy, serve it to you at that at the exact right time. And to Clara's point, agents are a huge part of that future.

Yeah.

Thank you so much, Jenny. To that point, I'm curious, how have consumer expectations changed already, and how do you expect they'll continue to change in the future? What is just table stakes for building in consumer AI? And Fai, you mentioned how young your user base is. So for users who may not even remember anything else, and soon there'll be more of those users who have grown up with this, What are they expecting?

Yeah, it's kind of crazy how fast I would say young Gen Z and even like Gen Alpha have adopted kind of like AI chatbots into their world, whether it's like ChatGPT or Character AI or what we're doing with Status. I think the craziest thing, which I think it's scary to a lot of people and it's scary to a lot of older people, but also I think is very powerful, is like AI as a friend. I think our users, one thing that is really crazy or why they're so obsessed with Status is they have all these interests, right? They have, maybe they're big into pop music or they're into anime or whatever it is. Before, you would just listen to a song from your favorite artist, you would maybe go to their concert. But now you can actually feel what it would be like if I was a pop star, if I was famous. And I think really what AI has unlocked is the ability of AI as friends, but also— and I think that's something they're going to expect. Maybe not like your best friend, but also like a confidant. someone who you can talk to without judgment. I think that's probably the craziest unlock that is very powerful for those younger users.

Yeah, fascinating. Eliza, what about in the world of— you're obviously engaging with the physical world. What are consumers expecting? I mean, of course, many people use something like Google Maps every day. Are they going to expect this level of personalization when they open Google Maps, or do you have an opportunity to do something different?

When you think about existing search engines, they're very much beholden to existing stakeholders and existing revenue models. So Google Maps can't truly not show you a huge swath of restaurants because they are like the Yellow Pages of the physical world. Just like Google Search, when you search for Wimbledon outfits, their entire search engine is predicated on a revenue model. But as startups, we can completely disrupt this. Like, I can show you whatever. I want to show you, you know, and no one cares. And I think that's the beauty of, like, generative AI, especially for place recommendations. For places specifically, a huge unlock for us was, like, we can't actually create new places. Like, a place is a place, and in a 1-mile radius, there are only so many vibey date night spots in an $80, $100 price range with good cocktails. Like, it's actually about us using the right nomenclature and the right language to describe and provide context around why this place is being recommended to you. And that's obviously the beauty of LLMs, right? So we can say, you know, Fai and Jenny, this is their favorite Italian spot. They've been here recently. They have photos here. It's great for walk-ins. You know, it's got this really aesthetic mood lighting. It has this cocktail that you love. Yeah, so you can't invent places, but you can reinvent how you speak about places and speak about them in a way that really resonates with the user that's, like, doing the search at that moment.

Yeah, that's so interesting. Clara, the example you gave about, you know, being at an event like this and coming away with a recommendation for like, these are the 3 people you missed and you absolutely must talk to, like, that's super interesting. Like, how do you make that happen? Like, what— obviously you're using AI, but to get more detailed about it, like, what are the inputs? What do you know about the person to give them that insight that could potentially change their lives?

We know everything. No, that's interesting. And just to piggyback on what you just said, like, I think what's going to change from a user perspective is like we're becoming Really, really demanding about technology. Like, I don't know how. Like, last time you looked something up on Google and you're like, what the fuck is that? This is so irrelevant for me. But like a year ago, it was like normal and the biggest company in the world. And now we're like, you know what? Maybe Google is gonna die. Don't think this is gonna happen. But like, you see my point, right? And so it's a great unlock. It's a great opportunity. But building AI product also means that you can't miss your target. Like, you have to be. Relevant for the users. Otherwise, they're just going to discard you in a second. We're becoming kings in our own world, right? We want the right answer instantly. And sometimes I just get frustrated with ChatGPT that, you know, I just like, this is not what I wanted, but it's already like 100x better than what we had like 6 months ago. And so expectations are rising. And so we need to build product accordingly. And so to your point about Gigi, what I got really obsessed Was, and that's interesting because at first I was like, I want to talk to an AI who's going to connect me with the right people, right? But then you know I started chatting with an AI from scratch and I was like, this is boring. I don't want like it's like a first date. It's like what's your favorite color and who are you? Like who I am is mostly already online because you are a compound of what you did or what you're doing and what you say, right? And it's already online. Like there are 800 million people on LinkedIn, millions of people on Twitter. You have a Substack. You have an Instagram. Sometimes it's You know, it's public. There's so many things about everyone in the world. And I think the beauty of AI is what you were saying, like grasping the vibe and the weak signals. And so, you know, I've been— we launched like 2 weeks ago this version of the product. So it's very early. But a lot of people just texted me telling me like, how the fuck did you know that? And everyone reading their profile is amazed with the level of granularity. Because if you just commented something one day on a thread, we're going to grasp your tone of voice. If you reshared an article, whatever, we can understand your interest. And so that's interesting because we— I think the biggest hypocrisy in the world right now is like we're spending 95% of our time, like of our life, online, and we keep calling this the real world. So breaking news, this is not the real world. The real world is the one where you're on your phone and your computer, right?

Yeah.

And on this one, we're leaving a lot of, you know, signals about who you are. The past 15 years, we've been using the signals to basically push products onto you and content onto you, right? That's why, like, TikTok is amazing because, man, you're 3 squirrels and you understand I love cats. That's amazing. I think the beauty of, of what AI really unlocks is we had the ability to index content like Google with PageRank, right? What we've created with JiJi, we created something Tell no one. That is called a social rank that's running in the backend because now we have the ability to understand social status about people based on the way you talk, where you're working, where did you study, who you're interacting with, who you're connected with. And so we know that we're social animals and we're driven by status. And so you want to interact with more or less people who are the same ranking as you. But then I can understand people you're going to vibe with, people who are adjacent operators in the field that you're working in, people that are going to help you, people you're going to be able to help. people who are sharing the same interests as you. So, the beauty, we have now the ability to grasp all this information, to embed it, and vector spaces are like when everyone tells me, how do you know if we're going to connect? I'm just like, you're close in the vector theta space, and I don't know.

The embedding is close.

The embedding is close. That's interesting because when we launched The products, I think the real magic came from Sequoia. Like, Jess Lee is the partner on Gigi at Sequoia. And we scraped 100,000 people, right? All the tech industry, like the cool kids. And I can promise you without cheating, her first match was her husband. And Tim's first match was Jess. And so this is when Jess said, it actually fucking works. And they're like, it does fucking work, right? And so I think now we just have the ability to process a lot of data. Everything that you've put online is going to help you not just push content onto you, but push the right people onto you. And I think now we're going to create a world where serendipity is not going to be something that is luck, but it's going to be productized. And everything great that happens in your life— love, your career, your friendship— is about meeting the right people at the right time. And now I have the confidence, and I'm not saying that we cracked it, but I'm probably going to spend the next years putting all my soul and my heart into cracking that is, we're going to live in a world where you're going to have a social, whatever, connector, assistant, whatever you want to call it, friend, probably. It's going to be this mixed blurred line. It's going to tell you who you have to meet, and where you have to go, and what you have to wear. And that's great. It's going to make us like superhumans.

Okay.

That's fascinating, and we could spend like another hour just on that.

That's right.

We've got a lot of other stuff to dive into. Claire, you were also talking before about kind of, you know, looking at, okay, AI has kind of cracked the productivity space, the B2B space, and now we're on to consumer. Jenny, you're kind of at the intersection of those 2 worlds, of course dealing with consumer, but also the giant e-commerce industry. What differences are you seeing between the best uses of AI in both cases, and what maybe we haven't seen that much of yet in the consumer space because it's not as relevant in the productivity world?

Yeah, that's a great question. I think with enterprises, especially in retail and fashion— LVMH is one of our investments, So we see a lot of this firsthand. The pace at which an established brand moves is very different than the pace at which a startup moves. To Eliza's point, we can throw up a new feature. We can talk about a new feature on Monday. My team will build it. It's live on Friday. We can just A/B test these things extremely quickly. I think that gives us the flexibility where a lot of larger companies are announcing features like virtual try-on and more intelligent shopping search and recommendations. But I think for us, it's like these are hypotheses that we can decide to build and test the next day. We're constantly A/B testing models and different recommendation architectures in order to see what are users really resonating with. I think that's hard to do when you're an enterprise-stage company. It's also when you have users who are not necessarily logged in and you're missing metadata and demographic about them. Then also what I thought you said was super interesting, Claire, is how the bar is getting raised and we see this every day. A user might have been satisfied with the same exact query 6 months ago and now they're so upset because at one point they told the AI, I will not wear heels more than 2.5 inches tall. Then I'm in there zooming in on the shoe that they disliked and they were mad at Altiva and I was like, oh my gosh, it's so close. How did they expect us to do that? But I think that is the reality, is just that you're not just competing with people in your specific industry or space. Like, yes, like, you could benchmark us against, like, the searches on another retailer, but in general, like, the state of consumer— like, there, people are using ChatGPT, people are using other products. Like, if another product performs better than you, like, how are you going to think about the product that you're currently using? And also, I think people are so creative now, and they're willing to tell the AI everything. So maybe, like, 12 months ago, like, common searches were like, what do I wear to a presentation at work? And now it's like, What do I wear to get revenge on my ex? Or like, we just see some really funny things like dress me up like a hot Minion. What do I wear to get my groceries at Costco? I was like, wow, Costco is killing it if people are routinely searching what to wear to Costco. So I think it's both like increased expectation for what you expect AI to do, but it's also like people are willing to share everything with the AI and that is also So you guys are building in New York.

What's the benefit of building an AI company here? Why be in New York? I don't know, who feels really strongly about New York? You, of course, care about places and the world around us.

Yeah, so we moved our entire company to New York from SF 8 months ago. My co-founder and I spent 1 month in New York, met 35 users, used the product probably twice a day, and we Went home, packed up in 2 weeks, and moved here. And I think SF is an incredible place for, like, startup boot camp. Every single person you meet is someone you can hire, or someone who can give you money, or someone who can introduce you to someone who you could hire or give you money. And in that sense, it's like you're constantly on. And it really took me from private equity to consumer founder. And that wouldn't have happened in New York, I don't think. But the beauty of New York is, I think, once you kind of get to a certain point in your SF boot camp journey, building in consumer, like, there are normal people here. There are people from all over the world here, and they have different interests, and they do different things, and they drink a lot, and they do a lot of drugs, and, like, they go on a lot of dates. And in SF, you know, it's like YC founders, Ivy League founders who are dropouts, and they go to sleep by 9 PM, and they're, like, maximizing. They worship Bryan Johnson. And that's just—

That's not true.

That's very SF, and that's so unique, but that is just not the average consumer. And I think if you're at that point where you're ready to build something that truly scales to a large demographic, it's going to be hard to really build for your customer in SF.

Do you guys agree?

Yeah, I totally agree. I also think, like, people don't really talk, but New York does have, like, a really great consumer pedigree. You know, Tumblr, Foursquare, a lot of big consumer apps were built here. And yeah, you need to talk to normal people. You need to talk to people who don't know, you know, who don't know who Bryan Johnson is and don't know what YC is and things like that.

Oh my gosh, one day my co-founder, when we just moved, he was wearing his Claude hat. And we were with a user, and she was like, oh, you like Claude Monet? And we're like, we're not in SF anymore.

It's good to talk to people who don't know who Bryan Johnson is, even if you're not building a startup. That's just like healthy. Yeah, Clara, it looked like you had something to add.

So I'm still based in Paris, and I'm moving to the US. And I have very mixed feelings, actually. And I'm not entirely sure that I'm going to move to New York, because I do believe that SF is where the weird people are. And when you have a new social protocol to build and a new social parity game, you want to be weird. You want to be with the people who think entirely not in this world, because this world is going to change. Like, in 6 months from now, we're going to all talk to an AI who's either going to help us roleplay or just just be the master brain of human beings and just do and go things for us. And so the most brilliant people right now are not in New York, they're in SF. There's still a lot of really smart people in New York, but the best engineers, AI engineers, are in SF. And the weird dudes are in SF. And you want to be really weird. Remember Facebook early days? You could have a feed and see Everything that your friends were doing, like adding someone, changing their status, whatever. It was so fucking weird, right? It was like a stalking tool. And I believe that to create this new social protocol, we're gonna get into simulations, like weird stuff. And here, you wanna be real. You wanna go to a restaurant. You wanna be dressed up. You don't wanna do that. You wanna get big in your brain, do mushrooms and fucking ayahuasca, and just like go deep in what the world could actually be. So I do believe, and this is where I agree with you, You should build in SF, and you should scale in New York, in my opinion. And you can still talk to a lot of non-weird people from time to time. That's my 2 cents, but I still haven't made my decision. If anyone has something to help me decide, but— You want to be weird. Weird shit is going to happen, and that's interesting because we shipped this first version 2 weeks ago, and I chatted with a nerdy guy on Discord and Facebook, and he was like, make it weirder.

This is too—

like, this is not weird enough to take on the world.

So yeah.

Okay, contrarian take for the room of New Yorkers.

Um, sorry.

Yeah, we're, we're almost out of time, I think. Jenny, you brought us all together, so anything you want to add? Anything we haven't talked about that you think is just super important in the consumer AI space?

Oh, that's such a great question. Let me think. Well, I mean, I'll chime in quickly on the New York topic, which is I'm an engineer by training. And similar to Eliza, I feel like I did your time in startup boot camp, writing code, scaling businesses, watching how other founders did it. And I think it's possible to build— I'm really proud of the engineering team we've built in New York. Everyone on our team is amazing, and people that we've recruited. And I think it's more about finding people who can withstand understand the noise of everything going on and just be incredibly focused. And I remember actually one of the questions we talked about on this panel is like, when do you decide like if you want to pivot or not? And I think with a lot of consumer companies I've seen just like pivot so many times. Like you could have had something really exciting, but you just didn't spend long enough on it. And I think if you are building something in this room and it's a problem that you're obsessed with, you just need to keep on going on that problem and just keep on digging and digging until Until— and you'll find the insights, or your users will tell you what they want. And I think that has been such a great unlock for us.

And then you have to pivot sometimes. Don't be too stubborn.

Yeah, yeah.

Well, thank you all so much, and thank you everyone for being here.