100 AI Leaders Explain How to Build AI That Will Win in 2026 — WHAT BUILDERS SHOULD DO NOW

1. 0:00 – 2:00

   Intro

   1. SPSpeaker0:00

      [gentle music] AI has definitely changed a lot for how we build products. The FOMO is very real. I think that's how it got started.

   2. JHJess Holbrook0:08

      The underlying capabilities are getting dramatically better. UX is kind of coming along, but in a lot of ways it's not accelerating as quickly.

   3. ACAyça Cakmakli0:15

      A powerful AI model with poor UX is kind of like a heavy duty power drill with a terrible handle.

   4. PDPasquale D'Silva0:27

      I think things have changed a little bit. The conventional wisdom is you start with the customer need, or you start with a design vision. I think it's more successful to build products starting with the capabilities of the technology.

   5. SPSpeaker0:41

      To answer that question, 100 AI product leaders came together at Swell Summit. The winners of AI race will be determined by [birds chirping] [keyboard clacking] [gentle music] Hi, I'm Summer Kim. I'm one of the lead partner at StratMinds. We are a VC and advisory firm based in San Francisco, focusing purely on applied AI since 2018. Before then, I was dedicating my life studying humans, user experience, for the past 20 years working for big tech like Microsoft, and then I joined Google, worked on communication and collaboration products. I also was at WhatsApp starting their first user experience function, and I joined Roblox really thinking about how people actually exist in a place like Roblox. Now, working with a lot of AI early stage startups focusing on how do we make AI products better than what we have. Three years ago, Anton and I sat down. We thought about how everything starts. It start with actually the care that our founders have, and then care ultimately translate into amazing product. So that's why we start Swell.

2. 2:00 – 3:26

   The New Building Blocks: Chat is not enough

   1. SPSpeaker2:00

      [keyboard clacking] As user researchers, we always try to think about what user really needs. There's also wants, generally come from the users based on their knowledge. What we can do is something that probably user can't even imagine. Tools like ChatGPT and Perplexity gives fast answers, but they still wait for us to ask first. The real challenge is helping people before they even know what to ask. I was so excited for our first speaker. Jess Holbrook has been looking beyond chat and studying deeper patterns, what he calls the new primitives of Gen AI.

   2. JHJess Holbrook2:36

      We're in this, finally this moment of experimentation. Everyone was like, "Chat, chat, chat, chat, chat, chat, chat," and we're breaking out of that. I showed you a picture of ChatGPT. That's the one everyone's familiar with. You can start asking it questions, you can do all these things. But chat is both universal and kind of a dead end for user experiences. We're starting to see experimentation in different ways that the UX and the UI of AI could go. One that we're seeing is a lot more structure. So what I'm showing here is a product called Elicit that creates very structured research reports for you based on academic outputs. It's structuring things. It's telling you what it's thinking about. It's giving you sources, all of this. And this is a product from Runway. If you heard the founders talk, they deeply believe they're, like, creating a new camera, creating a different way to approach creativity all up. So that kind of gets me thinking of, okay, so what are the new interaction

3. 3:26 – 5:46

   The New Interaction Primitives of Gen AI - Jess Holbrook, Head of UX Research @Microsoft AI

   1. JHJess Holbrook3:26

      primitives of this new platform? You know something's a primitive if you would say, "Hey, remember when we couldn't do blank?" What did we even do before, like, pinch zoom when trying to, like, look into an image or something like that? I gotta start with chat. I don't think people have really internalized that we're gonna be able to chat with kind of anything at any scale all the time. One thing I haven't seen any conversation about also is, like, we're about to see Metcalfe's law applied to everything. Number two, we're gonna have semantic resize for everything. So any content you come along, you can make it longer, shorter, more formal, more casual. I'm tired version. I'm in the car and I'm five minutes away from my destination. Give me that version. And it's, you're gonna really be able to adapt any piece of content to your current situational and mental state. Number three is just remix. Everything's going to be remixable from now on. This is maybe, I don't know, old and cliche at this point.

   2. SPSpeaker4:19

      [laughs]

   3. JHJess Holbrook4:19

      But this is, uh, Harry Potter by Balenciaga. We're gonna have style transfer, like, across the multiverse. Everything will be able to be crossed with everything, like, effortlessly. I originally made this talk before Sora came out, and so now it's, like, Sora, one of the core mechanics is this remix. Number four, format translation. Transfer among all these formats with very little loss of fidelity is gonna be enormous. One of my favorite examples of this right now is this new company called Obo, and you kind of tell it what you want to learn, and it says, "Great. Here's a podcast. Here's a lecture. Here's a deep dive. Here's some key takeaways. You want to now play a game? Like, what do you want to do?" You almost ha- just have, like, this total format freedom. So my last one here, I call it attention is all you need. Agents, agents, agents, agents, agents is, is all the stuff right now. It's like all these things existing at wildly different timescales and just multitasking forever at ever-expanding scales. We got a lot of data that is bad, and that's not how people work, and actually that can be the antithesis to great work. There's a lot of like, "Don't worry. We're doing lots of good things in the background. Go about your work. We'll come get you." I don't think we're designing very good lobbies right now, for you don't know what to do once you've started that agent. And so we kind of need to figure out what you do with this downtime and how you monitor all of these agentic experiences that are going on right now. So this is what I'm seeing. I think this is what's gonna underpin everything we're gonna build, so I think we should consider them and, and start to build them in today. Thanks a lot.

   4. SPSpeaker5:43

      [clapping] Woo-hoo.

   5. SPSpeaker5:43

      [keyboard clacking]

4. 5:46 – 6:45

   Make AI Alive: Craft a Magical Experience

   1. SPSpeaker5:46

      Pascal, is it time? Is it lunchtime? Tell me something sweet. My animal. Apple map. Woof, woof.

   2. SPSpeaker5:58

      Anton, our partner, had a great idea of why don't we invite Spark?

   3. SPSpeaker6:06

      Oh, hey buddy. What's your name?

   4. SPSpeaker6:09

      Spark is a really fun character. It's a dog, a magic dog, living in a quantum portal in a box. So I thought that that was really cool. We realized this was the best way to showcase that UX isn't about your latest tech. It's about making people feel something real. So we invited local kids and student to meet Spark, and watching them laugh, play, and connect showed us what it means to make AI really feel alive.

   5. SPSpeaker6:37

      Oh, this is a bridge.

   6. ASAndy Szybalski6:38

      This is the magic bridge. So Spark is the first non-human founder to go through probably

5. 6:45 – 9:16

   Spark, The First Non-Human Resident at the Hacker House - Pasquale D'Silva *Chief Creative Officer(correcting the CEO caption) @Illusion of Life

   1. ASAndy Szybalski6:45

      any residency, I think. Uh, and he's going through one in San Francisco called HFZero. Do you know about HFZero?

   2. SPSpeaker6:51

      Yeah. We should talk about that. Yes, yes, yes.

   3. ASAndy Szybalski6:54

      [laughs] They let him into the house. World first. He went out to San Francisco, he raised a million and a half bucks, and he- we promoted him to co-founder.

   4. SPSpeaker7:01

      But how do you actually pitch it?

   5. ASAndy Szybalski7:02

      You make the pitch not a pitch.

   6. SPSpeaker7:04

      Oh, I see.

   7. ASAndy Szybalski7:05

      Yeah. Or you have to make it really memorable.

   8. SPSpeaker7:07

      Yeah. How did you get up here?

   9. ASAndy Szybalski7:09

      I knew I wanted to be an animator for as long as I can remember remembering anything.

   10. SPSpeaker7:14

       Mm-hmm.

   11. ASAndy Szybalski7:15

       I started my first job in animation when I was, like, 14 years old.

   12. SPSpeaker7:17

       Okay.

   13. ASAndy Szybalski7:17

       I thought whatever that thing is that the folks were doing behind the scenes-

   14. SPSpeaker7:21

       Mm-hmm

   15. ASAndy Szybalski7:22

       ... on the Disney movies was the coolest thing ever.

   16. SPSpeaker7:25

       Mm-hmm.

   17. ASAndy Szybalski7:25

       And that has to be something that I do. So we like thinking about what is gonna make Spark's story more interesting. When we make his story more interesting, he becomes a more interesting character.

   18. SPSpeaker7:35

       Yeah, right.

   19. ASAndy Szybalski7:35

       More people like him. And we thought, what is another great environment Spark could be in? Doing a keynote or speaking at a conference would be an incredible thing, so I discussed it with Spark. He tweets it.

   20. SPSpeaker7:46

       Hey everyone, it's me, Spark Lincoln-Penny. [laughs] Yay. I can't wait to come to Hawaii and hang out with all my-

   21. ASAndy Szybalski7:57

       Universe, it'd be cool if a magic dog was the first one to do a speech. It's magic. It's like a hypnosis in a sense. That quality is within the work that, that we do. It's why we like magic so much. Like, how do you get someone who's had no experience as far as they can tell with these technologies, to get on board with the technologies without being scared of all the other nerd stuff that is happening? What we should best do with the influence we have, we have a very potent, magical dog that people love, and when someone loves someone, they listen to them. So what should they listen to? What messages? What do the other people wanna know more about? I think there's, like, a, a very natural crossover there. We wanna bring Spark to more people.

   22. SPSpeaker8:38

       Mm-hmm.

   23. ASAndy Szybalski8:39

       We bring him to more people, there's the potential to spread a lot of good, and so, like, how do you push it through that, that prism? I think, uh, is something that would be really good to explore.

   24. SPSpeaker8:47

       If Spark were to watch us talking about this whole experience, what do you think that Spark would say? One word? Okay, it's fine.

   25. ASAndy Szybalski8:56

       [laughs]

   26. SPSpeaker8:56

       You can do it two words.

   27. ASAndy Szybalski8:58

       There's three things that he would actually say, which are from his core principles.

   28. SPSpeaker9:02

       Yeah.

   29. ASAndy Szybalski9:03

       Creativity, collaboration, and kindness.

   30. SPSpeaker9:06

       I love it.
6. 9:16 – 13:01

   Don't Chase Users, Build on What the Model Can Do - Andy Szybalski @Cove, ex-Uber, Google

   1. ASAndy Szybalski9:18

      [keyboard typing]

   2. PDPasquale D'Silva9:20

      Now, if you had told me this two years ago, I would say you were insane. But today, users know their models, right? Like, people have opinions out there like they would about co-

   3. SPSpeaker9:32

      For your talk, what was message you try to really convey?

   4. PDPasquale D'Silva9:36

      If there's one message, it would be that as designers, we all need to know our material that we're working with, right? LLMs are really a new material. Actually, not even just one. Like, each model is almost like its own material with its own capabilities and its own properties and strengths and weaknesses. The best way I've found to build products out of this new material is just to play with it and see what it's capable of. And so I kinda shared some of my tricks, uh, that I've developed over the last two, three years building products with AI.

   5. SPSpeaker10:05

      Is there anything that you think that is relevant now may not be, or what do you think about this, like, the trick that you're developing?

   6. PDPasquale D'Silva10:10

      I think it's more successful to build products starting with the capabilities of the technology, which is not always how it used to be. The conventional wisdom is you start with the customer need, or you start with a design vision. Nowadays, it's actually the answer to the why now question is so important. I think being the first to identify a new potential for this, a new capability for these LLMs is really powerful.

   7. SPSpeaker10:35

      So you start with the capabilities, and what about next steps?

   8. PDPasquale D'Silva10:38

      It's not quite a linear thing. It's a back and forth. It's a push and pull between what do people need and what's the technology capable of.

   9. SPSpeaker10:44

      Right.

   10. PDPasquale D'Silva10:45

       So for example, with Cove, one of the things we think a lot, about a lot is how do we create these sort of exothermic reactions? We talk about how do we help users never get stuck in their problem, right? And part of that is about the challenge of a blank page. Like, how do I get started? But part of it is also when somebody is on a particular path to solve a problem, how do we help them go deeper but also go wider and consider other alternatives? And so we experiment a lot with different prompts to be like, how do we get the AI to act more like a true thought partner? How do we elicit the user's underlying needs? So they might ask, "What's a good venue for a kid's birthday party?" But what they really mean is, "Help me plan my kid's birthday party." And so I-

   11. SPSpeaker11:27

       Got it

   12. PDPasquale D'Silva11:27

       ... they need to know kid's interests.

   13. SPSpeaker11:28

       Mm.

   14. PDPasquale D'Silva11:28

       What theme would be good? What are fun activities? How many people should I invite? What should I do for food, right? Often what people ask for is just the tip of the iceberg of their actual goal. You have to crush the actual thing they're asking for in order to earn the right to help them with the rest of it.

   15. SPSpeaker11:42

       How, how are you doing that?

   16. PDPasquale D'Silva11:44

       It's common across, like, whenever you're solving a difficult problem It's not a linear process, right? I think the chatbots that we have today are very linear. And in fact, that's not how real problem-solving works. Anything sufficiently complex, you're going to explore multiple branches, you'll diverge, you'll have a bunch of different ideas, you'll kind of prune, you'll probably rule out some ideas.

   17. ACAyça Cakmakli12:02

       Yeah.

   18. PDPasquale D'Silva12:03

       You'll explore multiple paths, and then you'll narrow down and come to a solution. And often you'll do this over a long period of time. I mean, there are gonna be a lot of winners, right?

   19. ACAyça Cakmakli12:10

       Yeah.

   20. PDPasquale D'Silva12:10

       There's gonna be winners that create great models. There's gonna be winners that create great developer tools. There's gonna be winners that win because they are going very, very deep on a particular vertical-

   21. ACAyça Cakmakli12:22

       Mm-hmm

   22. PDPasquale D'Silva12:22

       ... because they really understand law firms or the insurance business or whatever. But yes, I think that there is going to be a category of winners who find the right experience for delivering general problem-solving intelligence that have yet to be found yet. It's a problem that we've not cracked yet as an industry, so I think there's a lot of greenfield there.

   23. JLJenny Lo12:47

       We say ship to learn, right? Uh, but shipping speed may not equal learning speed. We see companies shipping and iterating a lot, but not necessarily progressing their product. Hi, my name is Jenny Lo,

7. 13:01 – 15:09

   How to Design AI the Right Way - Jenny Lo @Global AI Platform, Ex-Grammarly, Uber

   1. JLJenny Lo13:01

      and I do product strategy and user research. I've worked across many different companies, helping to identify user needs and translate them into product roadmap and features. For any type of product to be successful, it really needs to have clear problem identification as to the exact value that the user is going to have. Take the example of Grammarly. There is a lot of trust in the brand equity, in the product itself, loved by many of its users. And so the inclusion of Gen AI is definitely, one, is because it's very relevant for the business and to be able to really demonstrate that type of capability. But how do we do so in a way that does not lose trust? So that meant how do we actually use the technology in a way that is much more thoughtful, that is much more valuable? I think it remained back to try to look at its core as to users are trying to improve their communication. Uh, we know that we do a very good work after someone writes documents. Then you can come to Grammarly. Now, how can we do that piece of work better? Then start to move into composition. Even before you have written something, we could actually help you think through the process of what you could write. And I think that was also a very big shift for Grammarly, uh, at that time too. One of the biggest pieces of work with emergence of Gen AI was that key top opportunity areas or top 10 problems that the customers had, and then we mapped out the capability of AI. Like, where do we have in which AI could solve against these types of problems? That really helped the business to look at it in a different way. Those are some key moments that we could actually start looking to improve. But oftentimes we're seeing the reverse, is like, "I hear there's AI. I want that in my company." Like, figure out how. The reverse mentality, when you could ask, "Here are all the different types of problems, which are the ones that are from the, uh, AI standpoint of technology that can really best serve?" I think that's a much more healthier conversation, a better reduction of cycles of iteration and find what works.

   2. ACAyça Cakmakli15:07

      My name is Ayça Cakmakli. I'm

8. 15:09 – 17:01

   Don't Build Confusing Gadget - Ayça Cakmakli @Google

   1. ACAyça Cakmakli15:09

      a UX lead at Google. And while I was prepping this talk... We're at a time where AI foundation models are becoming commodities like electricity. Companies are gonna pretty much have access to very similar models and algorithms, and it's gonna come down to are you a company/developer who is creating a confusing gadget, or are you a developer company who's creating an indispensable product or experience? People don't adopt technology. They, they adopt tools that solve problems. And I view good UX as the ergonomics of artificial intelligence. For decades, we've perfected the physical ergonomics of tools like chairs and power drills to make sure that they're safe, comfortable, and efficient. And at the day one of figuring out the ergonomics for AI, it's still very rudimentary.

   2. PDPasquale D'Silva16:07

      Right.

   3. ACAyça Cakmakli16:07

      So our role in UX is really to design the interface between the human and the technology.

   4. JLJenny Lo16:15

      One of the key difference, I think, in the AI era is that the interface is also changing. Now everything is a chat interface, so every types of intent and use is not through a button click. Now it's through a prompt that gets recorded. A lot of research approach that al- I also try to include now is study of user prompts. That's a very key piece of making sure we understand what are the things that people are requesting. That means how do you study conversations? Also being able to capture whether the outcome of that conversation is satisfactory or not. So I think that's also something really, uh, a key opportunity area for how research methods might change.

9. 17:01 – 21:21

   The Real Moat for Next-Gen AI Products

   1. SPSpeaker17:02

      When you're here, it's such a beautiful place. It in- inspires us to talk about things that we don't normally get to talk about.

   2. SPSpeaker17:07

      Which AI conference you get to and the first thing you meet is a double rainbow?

   3. SPSpeaker17:13

      Yeah, totally. We had a fairly in-depth philosophical discussion about what AI means for the next generation, what kind of products are actually proper or appropriate products for humanity, all the way to actually being immersed into using AI to make media or interacting with a live AI. Spark was by far the most magical moment of day one for me because I got to see it interact with different age groups of people.

   4. SPSpeaker17:40

      I remember closing the conference yesterday with this thought popped in my mind. We were talking about, you know, personalization, AI, multimodal, all this subject, agentic stuff. But then if you think about, like, first, one of the earliest form of personalization is when your mom ... cooks you your favorite dish. That's, you know, care. Care goes into-

   5. SPSpeaker17:59

      Yeah, yeah

   6. SPSpeaker18:00

      ... personalization.

   7. SPSpeaker18:01

      And you know, when you say personalization, it sound, it feels like you're taking a lot of data away from me, or I have to go through a lot of settings. But the interaction with Spark was interesting in the sense that there's not much. It's just the initial hello and couple of lines being exchanged just naturally because I'm trying to get to know this particular creature. That leads to hyper-personalization. That was really cool.

   8. SPSpeaker18:21

      Yeah, we just-

   9. SPSpeaker18:22

      Yeah

   10. SPSpeaker18:22

       ... need to be seen. And, like, what I notice about magic products or AI products, it al- often makes me feel like I'm seen and heard, and that's important.

   11. SPSpeaker18:32

       Yeah, totally. I think we need to have the beginner's mind that you practice as a, as a Zen practitioner. If you try to know everything and try to control everything, sometimes you, you miss or lose even bigger piece of the pie that you could have attained. But in terms of direction, I think there are some very profound thinking we need to put into this. With a lot of progression we have made from the old web to the new web, old app to the new app, when it comes to human privacy and implications of the technology on the society and so forth, we learned a lot what worked, what didn't. And AI has a huge amplifying power, and I don't think we, we want to get it wrong too much this time. We wanna do it right because this time even the wrong will be amplified. It's okay to not know everything and not control everything because as you saw from the kids' interactions during our special session, how they interacted with this magical technology-enabled dog was different from how grown-ups did. And I don't think we, even the grown-ups could know all the answers. And having some room so that they can explore meaningfully and potentially even teach us how to actually do this right-

   12. SPSpeaker19:39

       No, Spark told me I'm not an adult, so-

   13. SPSpeaker19:41

       [laughs]

   14. SPSpeaker19:41

       ... it's okay. I'm good.

   15. SPSpeaker19:43

       I'll, I'll give you some room so that you can help us figure out how things are gonna be.

   16. SPSpeaker19:46

       I have, I have to grow up-

   17. SPSpeaker19:47

       Yeah

   18. SPSpeaker19:47

       ... or, or not.

   19. SPSpeaker19:48

       Oh, I don't think he suggests that you need to grow.

   20. SPSpeaker19:52

       This year really felt different. More questions and more perspectives. People are thinking deeply about AI and UX now, using it more. We're experiencing it firsthand. And that's when one question kept surfacing: What about our kids? At the end of the day, how do we think about the next generation? 'Cause I have two kids. I'm a working mom with the two boys, they're 5 and 11. They're gonna be living in a completely different world. Their building and thinking and even studying is gonna be very different. They need to think about AI as their thought partners or friends or whatever they have. It is equalizing a lot of things, so you don't have to live in San Francisco to have this access. And then you can also start a company early on 'cause you have all the tools. They're available than ever before. I don't have all the answers, but one thing's clear. We're riding waves of constant change. Models get better, capability expand. But what lasts is the experience, a sense of magic, trust, ease, and the feeling of that AI show up at the right moment, sometimes even before you ask. We call this Swell for a reason. The waves will keep coming. We just have to keep learning how to ride them. We believe the, uh, winners of AI race will be determined by great, really great user experience. [outro music]

Episode duration: 21:31