WTF Summit 2025: Paris Hilton on Social Media Strategy, Netflix CTO on AI Approach | Oct 29, 2025

27:22Okay, and so I'm thinking you're working with pretty picky clients. So where are you seeing AI automate and where are you seeing people say, no, I definitely want a human to do this, but I don't want to handle it? Do people have that level of specificity or demands when they come to Duck Bill? Look, consumers should be picky. Yeah. Especially when it comes to your personal to-do list. The number one thing that you need to understand and I need to understand every day about this personal assistant space is that the holy grail of the personal assistant for everyone that everyone is chasing cannot be solved by AI alone.

27:57How did you realize that? Like that AI couldn't do everything. It's a crazy question to ask because one could think it's obvious. But, you know, I'm seeing a different pitch every day out there. You know, I just saw a TikTok from an influencer who was saying that Perplexity Comet helped them book everything from their flight to their hotel to try out a new restaurant opening without touching their keyboard. So, yeah, you know, where does that, again, where does that line draw on? How did you realize AI couldn't do everything? Yeah. AI is amazing at the things that you can do digitally. And a lot of our lives are online these days.

28:26And so we leverage all the AI solutions that are out there. And by the way, they are becoming increasingly commodity, increasingly affordable to just plug into your system and rip. The unique difference is that for all those things that happen offline, the mom and pop vendors who only respond to phone calls. And by the way, phone calls from phone numbers that they recognize. The camp forms that have to happen in pen and paper or are completely different one camp for the next for your multiple children. The discussions that need to happen with real people on the phone, the negotiations with customer support reps, all of those things are things that happen in the offline world.

29:07And that's where the hard stuff is. We find that if it can happen digitally, there's probably a pretty good point solution out there already for you. And if not, you can do it through AI pretty easily. The magic comes when we give you leverage in the offline world in a way that allows you to get stuff done as cheaply and easily and key high quality as if you were doing it digitally through AI. And are you monetizing based on how much someone does or doesn't have to touch the keyboard when they're trying to get a task done in terms of effectiveness? Yeah, walk us through the monetization and any numbers that would be helpful too.

29:41Yeah, so the average person who comes on our platform either comes on and tries one or two ad hoc tasks and is hooked or signs up for a subscription so that they can have more of an all-you-can-eat feel. This is really a product that creates a habit out of the gate. And the more people use it, the more they love it. And we pride ourselves on that smile curve of increasing usage over time. That is our bread and butter. That is what we wake up to drive every morning. And it's a sign to us that this is truly working. And so when you come online as a subscriber, if you come online as a customer, you start with your first task, second task.

30:18And the more different use cases you try, the longer and more valuable your relationship is with us. Because we can actually start to give you that leverage that a true personal assistant would give in the old world. Yeah. And so when you're thinking about monetizing a new habit, that's where I really want to click into because that to me feels a lot harder to do. People are used to paying for Instacart because it is one clear end-to-end experience. In this case, it's both things you want to do and things you maybe don't know you need help with, but will get help with. Yeah. So yeah, tell me a little bit more about how the cost structures work and how you guys are doing revenue-wise.

30:54So the sneaky reality is that people have been buying their time back since the beginning of time, right? Like to be totally candid, having a spouse who stayed at home and allowed you to have a very fulsome family life was a way that a lot of people, frankly, a lot of men bought their time back for years. And, you know, there was everything from the true personal assistant to having a Peloton in your house that allowed you to get time back. Uber, DoorDash, that's all about convenience, getting your time back and living the life that you want to live. And so buying your time back is actually not a new thing.

31:29Talking to people about using technology for your personal to-do list, that is the new thing. And what we find is that human behavior is notoriously hard to change. The one exception is during a technological revolution. That's when paradigm shifts happen. And my guess is, as a former Uber executive, most people could not imagine 10, 15 years ago opening up their phone and requesting an unmarked car with a stranger to come pick them up on a dark corner to drive them home safely. And today it happens by the millions. That's Uber, that's Lyft, that's ride sharing. And so as we're shifting behavior, we are seeing that happen at lightning speed in this era of AI where hundreds of millions of people are trying these AI agents and paying for search in ways that we never thought possible before, the change is really happening pretty organically.

32:20And it's spreading through word of mouth. The more you hear your friend is using this, the more you see the examples of your college roommate or your sister, faster that behavior changes for the subsequent concentric circles of adoption. I was going to ask about customer acquisition growth, a classic consumer question. Is that getting easier, cheaper, or how are you thinking about investing your dollars? Honestly, one of the reasons we loved this category is because the demand is there. So to date, we have grown almost exclusively by word of mouth and referrals, which is next to free for us. That is our number one sign that we have something.

32:56And it is such a signal of positive product market fit and the fact that we can scale this thing in a really robust way. As we start to get out into paid advertising, paid marketing, we actually have a really healthy LTV to work against because when you build that habit, when you do have people who are paying for a monthly subscription, and then additional transactions above and beyond the subscription that they're making on the platform. So buying flights, booking appointments, all on your platform. You have a lot of margin to play with. I was going to say, so you can kind of, is it crazy to assume that one day someone books a United flight through Duckbill and you get a revenue share agreement going or something like that?

33:38Absolutely. Okay. Right? Like, Everybody is looking for ways to plug into these new surfaces, both the LLM search engines and AI agents and Duckbill alike. And so when you buy a flight with us, we're not guiding you based on any kind of commission structure. But if you do and we're able to monetize that, that just subsidizes the subscription rate and makes the product more available to more people. And then you mentioned sort of other AI companies in the space. Like I said, Perplexity Comet has been trying to do work here. Talk to me about the landscape, you know, becoming more autonomous, more agent forward.

34:14It feels like everyone in Silicon Valley is attacking this problem. Yep. I want to know what you're doing differently, but I also want to know what you've had to change about the business in the past year to respond to this increased pressure. Yeah. So this came out of the inspiration of my years at Uber looking at autonomous vehicles. I had a stump speech that I used to give back in the 2014-2015 era where I said, my kids will never get their driver's license because autonomous vehicles are going to be ubiquitous. And here we are 11, 12 years later, and there are more Waymo cars on the street than ever.

34:47And autonomous technology helps us parallel park and, you know, drive safely every day. It's there. But the reality is there are still human hands required to be on the wheel for the vast majority of miles driven in the U.S. today and around the world. And that's because the real world is messy. And the platform that Uber built, combining technology plus humans in the loop, was wildly powerful because it not only allowed real people to access this product before autonomous technology was ubiquitous, but it was able to collect data that informed getting the product better and better over time, whether that meant integrating autonomy or not.

35:25That is the exact thesis that we're pursuing at Duckbill. So unlike the pure AI alternatives, we combine AI plus humans in the loop. And therefore, not only are we able to do the things that AI does on its own, book a dinner reservation at someplace that has a digital online booking platform, but we're also able to loop humans in for kind of ad hoc phone calls for which we don't have automated calls, confirming that the restaurant can handle someone in your party with a peanut allergy. Got it. And in the future, we can actually direct that resource to ensuring that when you're on your way to that peanut-free dinner and your car breaks down, someone can show up and wait with your vehicle while the tow truck comes so you can get ahead to your dinner because it's your best friend's birthday and that's really important.

36:10The ability to combine AI plus humans is what allows us to tackle that long tail of your to-do list. And we are the only ones who have it. Okay. And we are two years ahead. And it came through blood, sweat, and tears of figuring out how to combine AI plus humans in the loop. It has been an incredibly hard problem to solve, to your question of how we've pivoted over the last few years. Really getting that right has been the problem that we wake up every morning to solve. And over the last several months, we've realized we have it. And that is our secret sauce. I mean, it must be a great feeling.

36:46Can you share how many customers you guys have today? How many of these doers or contractors you're working with? So let me put it this way. At this point, we have millions and millions and millions of real world tasks that we have done on our platform. Okay. And that has allowed us not only to get really good at what we do, but to continue to make the system better and better by the day. So all of those millions of data points we're pouring back into the system to not only make our AI models better and our triaging better, but to also ensure that when the human does get the baton, they're handling it as seamlessly and high quality as possible.

37:23There we go. Well, I will end with a question about the summit and what you're looking forward to. What are you hoping to sidebar with someone about? So I think right now we are all looking forward to this future and really wondering what it's going to look like. And the concern that I have is that the future is being created today by a lot of people who have a vision of the future that, frankly, I don't want to live in. I wake up every morning and think about the version of the future that I want to be a part of and I want my kids to be a part of. And what I'm really excited about is sitting alongside a lot of other people who have alternative visions of what the future could look like and are actively building toward that every day and to figure out how we can work together to make that a reality.

38:04Driver's licenses or not, I am very here for that. There we go. Megan Joyce, thank you so much for joining us on TITV. Excited to hear more about Duck Bill in years to come. Thank you so much for having me. Thank you. and now let's head back to the main stage where Paris Hilton the CEO of 1111 Media joined us last night to talk about the trajectory of her business here's what she had to say I feel like I've always been very ahead of my time and you know I was doing what they call an influencer today before there was even a word for it and I think it just because everything has really just been so authentic, like everything I do in my life and what I choose to do is things that make me happy, things that bring sparkle to the world, and also knowing what I wanted to do and realizing that at a young age, just being very close with my grandfather and always having lunch and dinner with him, and he was my business mentor, and I always really looked up to him and saw what my grand grandfather and he did with the Hilton Hotel brand.

39:10And that's when I just thought of like the power of a personal brand and that I could do it as myself. And another important thing to me was to be known as Paris, not just the Hilton Hotel granddaughter. And one day when I was having dinner with my grandpa, he said to me, you know, I used to be known as Baron Hilton, but now I'm known as Paris Hilton's grandpa. That must have felt good. It was, yeah, it was something I'll never forget. Such a special moment for me. And I'm just so proud of all the hard work and everything that I've done and always leaning into what I love, what I believe in, and what is going to make the world a more fun and slithing place.

39:52So walk us through how you use social media today. I've heard through the grapevine that you're very involved in your social media, understanding it, analyzing it. There are probably a lot of lessons for all the founders and brands and the audience in that. I am obsessed with just everything to do with it. Like I love going on TikTok, on Instagram. I love reading all the comments from everyone. I have a very close relationship with my fans. They call themselves my Little Hiltons and they are all around the world. and we just have this very, very close relationship. And I love that I'm able to contact them all around the world.

40:33So if I'm ever doing either a performance or a different, you know, like a perfume signing or something for my book or whatever it is, I'll always make sure to invite them. I'll even get them a hotel at the Hilton and they'll be staying there and then come with me to all the meet and greets and the parties. So it's really fun just to give them these really special, magical moments. and another thing I love to do is I search like hashtag Paris Hilton a lot and I just see so many people leaving so many amazing messages like after they've read my memoir or they just saw me do an interview or something and they just have these incredible stories where I'll just sit for hours and just like watch and cry and just be so touched by everything they say and then I always do at them back or I DM them or leave a voice note or send them a gift.

41:25So I really love that I'm able to have like a serious like contact with people all around the world. And I think that's exciting where you can just really utilize it in such a way. And I just do it in a way that it really comes from my heart. And I think that's what really resonates with people because they know it's not, you know, just a social media manager or somebody doing it. They actually know it's me. And even though I'm the busiest person on the planet, being a mom, all of that, I'm such a night owl. And I think with my ADHD, like my mind is just going in a million directions all at once.

41:58But it's hard to sleep at night. But the one thing that I do at night is that just connecting with my fans all around the world. All right. Now I am sitting with Merit Hummer, a partner at Bain Capital Ventures. Merit, welcome to TITV. Thanks for having me, Natasha. I want to start with the lay of the land. I feel like summer was absolutely insane, and now we're heading into end of the year. How busy are you? Is everyone trying to cram deals before the holidays start? The short answer is yes. So it is a frenzy right now. It feels like every day we're hearing about another company that went zero to 10 this year that's raising.

42:35So the task right now is figuring out how to separate signal from noise. Oh, my God. Yes. And you're in growth stage investing. So you actually have real numbers and real data to look at. So tell me a little bit about, you know, what you focus on. So you're focusing on B2B enterprise software. That phrase has meant so many things. In the AI era, what does that look like? And yeah, tell me a little bit about how you are delegizing companies in this moment, too. Yeah, so we focus on both what we call inception stage investing at Bank Capital Ventures, which is we have a team dedicated to seeds and series A's.

43:10And then we also have a team dedicated to what we call inflection stage investing, which covers the series B through D range. And that's where I focus and lead deals. And I think you're absolutely right that on both sides, we're going through a time of change. And I think that all the paradigms we used to think about for enterprise software investing in the past are really being overhauled right now. and ways of thinking about enterprise SaaS and what great looked like no longer apply in some cases. Actually, today we had an offsite and we revisited a framework we had from three years ago to think about what an excellent company profile is.

43:50And we realized that it's become obsolete. Tell me more. So is it like you're seeing, you know, okay, 100 million in ARR is the baseline right now and the starting valuation for a Series A is$500 million? Like, where do the numbers change? Yeah, I think the reality is AI has created so much market pull for certain new tools that are coming out that the expectations around growth have gone up. And the bar for how quickly companies need to grow to command a lot of investor attention has just risen. So, you know, a year ago, two years ago, it might have been a really great year of a company in inflection stage grew from$2 to$6 million of ARR.

44:29Now it's not unusual to hear about a company that's growing from zero to 10 on its way to 30 or 40 or 50. And that now defines sort of the top decile opportunities out there that we see. Yeah. I mean, I saw that you recently backed a company, Pylon, that ended up raising their Series B in seven days. How do you even start doing due diligence in seven days? Walk me through what that process looks like as specifically as you can. Yes. Well, I'm happy to do it. And actually, you know, the founder, the CEO and co-founder, Marty, likes to document his fundraise processes on X. We're back in building public.

45:06Yes. We're back in building public. And he's a great marketer. So if anyone wants the details, he also has talked about this publicly. But in short, the truth is we've done a lot of work before those seven days. And so it's not that from a starting, we start on day one and by day seven, we've built conviction. Our approach is to do a lot of homework before we even meet the CEO, before we even meet the team, so that we go into that meeting with a lot of conviction. That's what we did in Marty's case and Pylon's case, and that's why we were able to move really quickly and issue a term sheet. With Pylon and with just generally actually thinking about that pre-work, I'm always fascinated by this concept.

45:44I've written a lot about preempting and just like, you know, meeting someone and giving them a term sheet without any questions asked. How does that dynamic play out at a place like Bain Capital Ventures. You guys are not a new firm by any chance. You have a great track record. Do you have to also be that aggressive or are you able to sort of have a conversation with the entrepreneur as crazy as that sounds? Yeah, no, we absolutely, we will not compromise the integrity of our decision-making process. And so we do think we can operate, we do operate on accelerated timelines, but we don't skip any steps along the way.

46:17So in that compact seven days, we had multiple partners on our team meet with Marty. We did an investment committee. We circulated multiple rounds of materials for everyone to review. We received data from the company. We digested that. So we followed all of our steps. We just did it in a really compressed timeframe. And so that's how we operate. And we ultimately still make group-based decisions. And so no one on our team has authority to just issue a term sheet without the collective support of our team. Okay. And is it the kind of thing where you need everyone to say yes or can it be a mixed bag?

46:47It can be a little bit mixed bag. We can, you know, it's, I would say it's, it's, if you think of, you know, so independent operator to fully consensus based on one to 10 scale, we're sort of a seven or eight. So we don't need full consensus, but we need collective support and blessing from the team. Okay. Feels, feels right in this moment. And you mentioned sort of the compressed timelines. 996 is another, you know, topic I've been thinking a lot about in this moment. I mean, are you working 996? And also, how is that translating to the ways that you're seeing companies grow in this moment? Because I'm sure even if you're not, your founders definitely are.

47:23Yeah, absolutely. Or at least I think the ones who feel most energized by this moment are and who feel the pressure to keep up with the competition in their respective categories. So absolutely, I think it's a real thing. I think we see it across our portfolio. I think our investment team feels it, and we operate that way as well. And so I wouldn't go so far as to say it's strictly 996, but we will do what is necessary to pursue the best opportunities we see in the market. And if that happens to be 996 one week, then so be it. Are they having trouble recruiting people to do this work style, or is it becoming the new normal for AI companies?

48:01I think how I characterize it is I think within certain roles at certain types of companies, it is normalized. But I think there is space for people who work differently, you know, people who are in different roles, maybe remote workers, people who have families, you know, I think there is space for people to work differently. But with the backdrop of the reality, the expectations are very high right now. But I think the best recruiting strategy for Pylon, for anyone else for that matter, is to get the best people in the door and adapt to what they need in their lives. Yeah. I mean, and like you said, it's extremely fast moving.

48:37It's extremely competitive. I mean, I wanted to ask about the big tech companies in this moment. OpenAI, honestly, being one of them, you know, they're trying to do everything from being, you know, your Salesforce alternative to maybe your calendar one day and your Google Docs all in one company. When you're thinking, you know, about how to advise your founders building software startups, are you looking at OpenAI as a distribution channel? Are you looking at them as a platform risk? How do you tell your startups to deal with some of the bigger businesses? Yeah, I think one thing that's important for founders to keep in mind is you probably don't want to build a tool or a product that is right in their sweet spot or is such a large opportunity like coding automation, for example, that OpenAI or Anthropic or the other labs are very likely to be leaders and be fiercely competitive in that market.

49:26So one thing we think about is what are vertical applications that are very sizable and can produce big outcomes, but maybe narrow enough or specific enough that OpenAI is not going to pursue them independently. So something like manufacturing software and all the data that comes out of the manufacturing world. We have a large investment in MaintainX, which is a vertical software platform for the manufacturing world. And they have a lot of very nuanced workflows. They have distribution. They have data in their platform that would be very hard for an open AI to replicate. So we're excited about companies like that that have a very clear moat around why open AI can't just go and do that.

50:02Okay. Yeah. I mean, I remember the wrapper conversation being a big conversation in 2024. I mean, as we close out 2025, you think about these vertical AI startups. Beyond data, what are the top things you're looking for when evaluating a startup that shows maybe open AI, Anthropik, even Google, Meta, Amazon won't try and build this? Because if you ask me a year ago, we wouldn't have thought coding either. And so I would just kind of be curious how you even start to think about diligencing those. Yeah, it's a good question. I think the answer can look different for different companies. So one thing we would think about is, is there a distribution advantage this company has in some way that would enable them to proliferate within a market faster than even OpenAI could?

50:44Do they have access to proprietary data? For example, something like an open evidence that has a proprietary relationship with the New England Journal of Medicine. You know, there are companies that have access to data that not even OpenAI has. Are there, is there, does a founder just have specific know-how or expertise in a market that would enable them to build a uniquely insightful product that generalists at OpenAI might not have? So those are examples of questions we ask ourselves when we meet companies to figure out if they can actually compete alongside an OpenAI. Got it. Got it. And then obviously also balancing the fact that they hopefully have quite a bit of traction by the time they're seeing you.

51:22Has anything changed about the metrics or numbers you're looking for when first meeting a company? I mean, I think about AR meaning so many things in this moment. Yes. What metrics are you like first tapping into and double clicking on? Yeah. So I would make a gross generalization here and say, you know, an average company today that we see relative to a few years ago is growing faster, but has lower gross margins and lower quality revenue. So it may not be as retentive. It may be less predictable. Like I mentioned, it's lower gross margin. And so that's a tradeoff. So we see companies growing really explosively, but the quality of revenue is lower.

51:57And so we're trying to adapt our framework to think about how do we identify the best companies in that new world? Whereas, you know, three years ago, I think a consensus had formed around how to think about traditional enterprise software startups. Is net revenue retention enough of an answer for now, or is there a different sort of thing you're looking for? It's a good question. I think, you know, we think about the answer with a more qualitative answer, actually. So there are two layers to it. I think what we're seeing is a lot of these companies are growing really rapidly, but that unusually compared to, I think, past years is not necessarily an indication of product market fit.

52:34So we have seen a lot of companies with very aggressive growth ramps. But when you speak to customers, you actually hear a lot of mixed feedback on the product. It may not actually be transforming the workflow. It may not even be production ready. There may still be material gaps in the product that prevent it from really driving value. And so we think about two layers of adoption. You want to find companies that have exec sponsorship and are able to win those large contracts and become market leaders quickly. But just as importantly, actually, maybe more importantly, is to have the support of end users.

53:06and the product love by the people using the products every day. And that piece, that latter piece, is something that we see missing from some of the fast-growing companies today. Wow. Well, I am not envious of the life of a growth stage investor. It feels like a very complicated but exciting time. So thank you for talking to me about it. I would love to end with a conference question. I'm going to be interviewing partners from Index and Kleiner Perkins tomorrow. I'm just curious, what's a conversation you wish venture capitalists were having louder? And maybe I'll pepper it on stage. Okay, let me think about that.

53:35But I think you should ask them what they plan to do with their 2021 portfolios. Because I think that's a conversation that people aren't having out loud today. There's a lot of enthusiasm about the moment that we're currently in. But just a few years ago, we were wallowing in the aftermath of the 2021 era. And I think there's not a lot of discussion happening about what will happen to the hordes of zombies, as they call them. Well, thank you for that reminder. Merit Hummer, thank you so much for joining us on TITV. This was great. Thank you for having me, Natasha.

54:14And that is a wrap on our special edition of TITV. TITV is on Monday through Friday at 10 a.m. Pacific, 1 p.m. Eastern. Thank you to Amazon Web Services, our presenting sponsor for this production. I'll call you a week back tomorrow, but I'm Natasha Mascarenas. Thank you so much for your viewership. Until next time.

54:35Thank you.

55:08You