The Hottest Product Manager in Tech Right Now 🔥 OpenAI’s Holly Li Talks About Dots | Office Drama
1h 19m 14s
Holly Lee, a product manager at OpenAI, gained widespread attention after her standout presentation at OpenAI's Dev Day, where she launched Dots—an always-on, persistent AI assistant accessible across devices. Though not on social media initially, her viral moment stemmed from an unscripted, authentic performance that resonated with audiences globally. She credits her success to deep user empathy, a strong belief in her own capabilities, and a commitment to authenticity in both her work and public presence. Holly emphasizes the importance of privacy for users, designing Dots with strict controls—such as read-only access, explicit permissions, and an AI auto-reviewer—to prevent misuse or unauthorized actions. She also highlights the gender imbalance in AI leadership and advocates for more diverse, creative, and inclusive voices in tech. Her journey from a junior PM at Google to founding Figma Make and then joining OpenAI exemplifies her drive to build meaningful, user-centric AI tools. Holly encourages interviewees to use real AI products and approach roles with curiosity, not performance. Her experience with live demos, including a technical setback, shows resilience and humor, reinforcing her belief in transparency and human connection. Ultimately, she sees her role as not just building technology, but fostering trust, creativity, and balance in a rapidly evolving AI landscape.
How does it feel to be the baddest PM out there right now? You went viral and you're not even on social media. I went up there and I was like, "Oh, I got this." And then Doddy goes to you, "Yeah." And you were like, "How's it going? Still checking?" Yeah, so you went back to SF. What happened? I flew with a one-way ticket and I was like, "I'm not leaving SF without a job." I had met an amazing team and convinced them to create a PM role for me. You made your own draw about someone else's company. If I wanted to become a better presenter, what is one tip you want to give to me? Do not be a loser in your own fantasies. Imagine yourself as a f*cking winner. Whoa. I'm not your average AI tech bro. I am hyper aware that on Twitter, everything is like, "How can I get my AI to automate X, Y, and Z for me?" I think a lot of creators love doing things manually. I am with you. I want to walk inside the grocery store and not just have AI order stuff for me. I can't get over giving it access to my inbox, text messages, finances, help data. What would you say about dots to more privacy conscious users? Okay, so if you have one tip for people interviewing for an AI product manager role, what tip would that be? Actually, use all of the products that are out there. You'd be shocked at how many people just haven't tried the basic workflows and had an opinion about what's working for them. None of us know what the future of AI looks like. If a candidate walked in with an iced coffee in their hands, "Oh my God, this is making the way around on TikTok." [Music] Office drama. Welcome to Office Drama, where we define the modern career for Gen Z and millennials in tech, media, and business. I'm your host Clay Shee and today in the office. We have, perhaps, the most talked about product manager in Silicon Valley right now. She just presented at OpenAI's Dev Day and she helped launch dots, which is the AI assistant that is all over my social media right now. She went viral for this. I'm so excited to have you here. Please welcome member of product staff, Holly Lee. Thank you so much for having me. I'm so excited you're here. I know you just got off the plane in San Francisco and you had such a long day. And I think, am I your last stop of the day? No, you're not. I'm not. I'm. She has my work to do. Yeah, well, we'll see what happens after this. Okay, well, my first question to you is, how does it feel to be the baddest PM out there right now? Yeah, I just joined Twitter two days ago. Yes. In order to meet the moment and combat the haters. Well, that was kind of new. Welcome to social media. Yeah. Wow, it's a whirlwind. Do you want out yet? Or you're like, you're okay. I mean, completely. Honestly, I kind of don't know if I will keep posting. Or if it's just like a one and done. No, she's a private person. I'm a private person. In your personal life, have you ever had a crush on someone who didn't know you existed? Oh, absolutely. Like, one was the last time that happened to you. Oh, gosh. Well, I've been in a relationship for quite many years now. So I feel like the correct answer is to say before that. Yeah. But, you know. Yeah, so, you know, if you want to ask the John Malamas, if you've ever had a crush on someone, whether it's that coworker in the microkitchen or that person at that library studying at the same time as you, or that barista at that coffee shop, because my moment is right now with Holly. That's you. You're that girl for me. I first fell in love with you when Mesh Times Marissa, and my friend Mesh Times, posted about you at the config 2025 keynote presentation. Yeah. I love her. And then the title was, like, who is his PM, Baddie? That was so nice of her. That was really crazy. That was crazy. And you went viral. And you don't even, you're not even on social media. Wasn't there to receive it. But I was texted about it. And that was really nice. People from high school were texted. Really? Yeah. Oh, that's so cute. It reconnected me with actually, like, a lot of old friends. Are you aware how much aura you have to the PM and design girls? Oh, my God. That's very nice of you, you know. Did you read your comments? I did scroll through them. You got, you know, Holly Lee. Drop your account here immediately. We're all fans now starting from your presentation. You totally slay. Holly, the goat. I mean, Holly, to share the details of her config fits, because I love them obsessed with Holly after seeing her. Holly's stage presence. 10 out of 10, such a natural. You know what? I think the design community is, like, particularly nice. That's true. I do think, like, that config audiences may be, like, the most supportive, most creative group of people you could ask for. So I got so lucky. What about the AI audience? My comments on Twitter look slightly different. Let's just put it that way. You have so many comments asking you for your ads. Why are you not on the socials? Yeah, I want to keep my personal life and my work life separate. I know that's such a cliche. But I have aspirations to, like, start a creative practice and do more of my, like, art and interior design stuff eventually. And I want to reserve, like, Holly for that. So that's, like, chapter that's coming. Wait, what kind of art? I can nerd out, but I collect, like, folk art. Any type of art that's, like, handmade, kind of quirky, found whatever. And I do, like, little art exhibits and make my own stuff. And anyway, that's, like, my potentially post-Ai hobby. What about interior design? If anyone is looking for someone to, like, style their office or their apartment, that's, like, my number one dream job. I love it so much. So this is, kind of, like, a deep, lore thing. But I don't even know if this is real. Do you know what this is? No. What is that? It says, "Hello, just a grown-up kid figuring things out. I grew up in a small, oh, high-end town." Oh. Oh. Oh my god. This is your website! Wait. Yes, did I? Well, I don't know what this is. I have no idea what that is. Is that somewhere on my website? No. Your website is, says, "Temporary maintenance." They're, like, temporarily not. Yeah. I can't access your website. This is Holly Lee. Oh my god. You know what? I hit my versatile limits. Well, I wanted to see it. Oh my gosh. Okay. This is another website. I guess it's gripping your website. Oh my god. But it says, "Ask me about folk theory, communist propaganda, and Warren Buffett." I'm now pondering leftist edgelord culture, limping 26 miles in reading the Topeka School by Ben Learner. Yeah. Yeah, that is my website. Okay, great. I'll say, yeah. There's, like, 50 percent chance it is her that's totally-- That's totally-- I'm very confused by what this is. Okay. Okay. So my website's down, but that is the exact copy of my website. So it lives on somewhere else, yes. Yeah, I see where it's coming from. We just talked about one thing that we both really love, which is, I asked her if she was a theater kid. And most people, like, 10 out of 10 people usually say, "No." Yeah. Yeah. If you give me a piano, I'm unstoppable, so. Yes. When we all talked about, we have a mutual friend. Yes. A lyric? A lyric. He just left Confluent. But we're working at Airbnb. They went to Penn together. Yeah. And he's my childhood friend. And we do a lot of music together. Yeah. And I guess you guys also connect to that. Yeah. Oh, I'm so excited. All over the world. I know. I said, back to the social media, but you're on LinkedIn. One must be on LinkedIn. Did you accept my connection request? I need to check. I don't check LinkedIn, but I will. For you, I will. Thank you. I'm a little bit nervous and giddy right now, because this is our very first time meeting. So like, you want to get to know each other? Yeah. That's good. You're down? Yeah. Okay. So where did you grow up? I grew up in Athens, Ohio. Small town. Really, really small. Rural, Ohio, corn fields, football teams, not super diverse. But I was there because my parents taught, like, a university nearby. Oh, your parents are academics. Yeah. Oh. And two. College. Pam. What did you study? Study economics. I've got a Wharton School of Economics. Do I regret it? I don't regret studying econ. I do kind of feel like I should have gone to a liberal arts school. Yeah. I don't know if that's a popular opinion these days, but I'm like, damn, should I go into brown or-- I went to liberal arts. Yeah. I'm like, studied red books. Okay. So our viewers are called the Dromolamas. Yeah. Because drama. Office drama. I don't stay for the Dromolamas, your MBTI, astrology sign, and who you're wearing today. INFP. INFP? Is that surprising? That's very surprising. INFP. Sagittarius Capricorn Cusp. Okay. I can see that. Like, literally, like, the day. And then I thrift everything. This is my boyfriend's vintage burberry. They're from boyfriend's closet. Even that we thrifted, I think, in Kyoto. This is dead stock Japanese silk. And then I made this ring, designed this ring with soft goods studios in Oakland. And then this is a Navajo silver cuff I got in New Mexico. Wow. She-- you had that off the cuff. Like, you knew-- I really-- I researched the origins of all objects. That's crazy. Okay. Do you remember your Figma outfit? Oh, absolutely. What was that? I thought you got that. Okay. So, belly flats, boyfriend's jeans, because they were comfy. And then I made the shirt. So I literally-- I went to Goodwill. I got a giant button down butter yellow, because that was a big thing last year. Butter yellow button down. And then I drew on a tie with Sharpie. I wish I could say that that was a completely original idea. I did see it on Pinterest first. I just drew it on, literally, like, you know, one shot. And then I was like, I like that. And I wore it on stage with my little space buttons. Because that's my-- Yes, your space buns were so cute. Yeah. I'm always wearing here in two buns or two pigtails or something. What was your dead day outfit? Oh my gosh. So, a dead day, LOL. I was wearing Painter pants that, like, I legitimately wear when I paint. So, they're paint splattered. And then I wore, like, a very, like, 90s look, like long, white sleeve shirt, short green shirt on top of that that I made. I, like, sewed Jersey with buttons. And then I wore a bright orange, like, puffer vest on top of that. Yeah. Because orange is my favorite color. And I-- fun fact, I just wore that to rehearsal. And the events team was, like, no, we love that outfit so much. That's what you're going to wear on stage. Oh my god. And then I got, you know, people on Twitter didn't like it, but I have a lot to say about that. I want to know a little bit more about your career lore from the beginning of the Times. Afterpen. Afterpen.
How did you get your first job at a college? - I interned at McKinsey and then-- - You're a McKinsey girl? - I, that's why you need a lyric. - How do you even need a lyric? - You've got a lyric from McKinsey? - Yeah. - Wait, so you have a soul? - Actually, that, that, it's kind of a cute story. - Okay. - There was like an internship, we went white water rafting, and then I didn't have like a shirt afterwards or something. - Okay. - He like lent me his shirt. - That he was wearing. - So he's, so just naked, a layer gonna bust, you know? I don't remember the specifics, but it was so nice of him. And that's how we became friends, and he was guy. He's a good guy. - He is. - He'll be on office drama, we'll record next week. - Oh my gosh, really? - Yeah. - Okay, I'm tuning in. - We're childhood friends, we're neighborhood friends. We grew up in the same area, and then I always introduced him, like, oh my god, we're like friends from middle school or high school, and then he's like, no Chloe, we're friends from elementary school. - That's nuts. - Yeah, and I'm like, dude, I don't remember you in elementary. - Well, I've been to your neighborhood. I've been to his house, so-- - You have? - Yeah. - Anyway, that's amazing. - I know where you live. - Okay, so after McKenzie. - After McKenzie, I was like, cannot go back to consulting. - Okay, why? - I just didn't want to work 12 hour days, like, making little PowerPoint slides, no offense to consultants. I didn't want it, it wasn't for me. - Yeah. - I also got reprimanded by a partner for wearing yellow socks in the office, and I was like, what? - Can you wear yellow socks? - I had a lot of personality. And I was like, you know, who will let me wear yellow socks? It's probably Google. - That's true. - They love color. They do love color, and I had this like dream of like, you know, running around on a bike and et cetera. This was like the peak tech optimism. So I applied for the APMM program. So marketing at Google, 'cause I didn't have a CS background. And that's how I started out, moved to SF right after graduation to do that. - Wait, I thought you were in New York. So I did SF for two years, and then New York. - And I wanted to like, to a little switch up. - Makes sense, 'cause like, there's no bikes in New York for the Google bikes. - No. - Yeah. - Google in SF for a little bit, honestly, kind of a boring story. - SF or Mountain View? - Mountain View. - Oh. - So like, little 21 year old me was like, getting up at six to optimize her route. - You live in San Francisco. - Yeah. - Community Mountain View. Yo, I hated that. I chose Mountain View, and that was also really bad. I understand that SF Mountain View commute. - SF Mountain View, yeah. And then I wanted to switch jobs within Google. There was a team called Creative Lab, and they were based in New York. And I was like, sign me up. So I pitched them really, really hard to let me join them. And then they said, yes, and I moved out to New York. And that was like a really fun couple years on that team. I started out almost like a creative producer. Like we were writing scripts for like, how to talk about Google Chrome. There was like, Super Bowl ads. And then I was like, I actually want to be closer to product. - Yeah. - And so I like pitched one of the teams within Creative Lab to like, let me kind of join their Skunk Works operation. And we did a bunch of stuff like, we built a browser that only used Gemini. And like we were just like hacking stuff. My kind of spicy take is that this would never be allowed at a tech company now. It was like completely like low ROI, very, very expensive exploration. - It's very R&D. - It's very R&D, yeah. But even then it was like R&D with almost like too much of a marketing focus. - Ah. - Like very much like, how can we convince the world that like Google is very like creative and innovative. And so we'll just have this team create little experiments to put online. - At a certain point I became disillusioned with the fact that we were just really creating PR and not actually building real things. And so I decided to leave. - Were you 20% projecting on that? - Oh yeah. - Okay, got it. - Wait, 20% projecting on, sorry, which part? - Creative Lab. - I just joined them full time. - Oh. - And that's why I moved to New York. - Oh my God. - On product marketing or product? - It was technically within the marketing org. But there were like, I don't know, five teams. - Right, okay. - And like four teams did product stuff. And like one team did advertising. It was a really weird setup. - So you didn't have to re-interview? - I didn't have to re-interview. - That's like, 'cause it would suck to re-interview. - Yeah. - So what is Creative Lab and what is Magic Editor? - Creative Lab was like this secret squad that like has the 16th floor of the Chelsea building in like Google. And they just like do cool stuff. And it's like a giant creative loft. It literally looks like out of a movie. It was a dream to have worked there. - Wow. - Magic Editor, like this was like around the era of like, 2020-2021. - Yeah, I was starting to become like quite new. - It was really nearly a thing. - We had worked with research and research had had like a new image generation model. Like this is back in the day when like, creating like a pixelated picture of a puppy was considered like a big deal. You know what I mean? It's like crazy how far we've come. Research had this. We like built a like internal tool where you could upload like 10 photos of yourself and train a model on you and generate photos of you that like looked like this or looked like that. We worked with the pixel team to like create a special version of like Google Photos that had generated AI in it for the first time. And you could erase certain things, but you could also like generate a new sky in the background of this picture. You could move over a person a little bit or remove the like trail of a wedding dress. And that was like one of the first big like AI. - That's so exciting. - So Google New York or Google Mountain View? - Google New York, for sure. - But then you went back to SF. - Yeah, how long you're in New York? - All in all, maybe two and a half or three years. - You only had a shorts didn't in New York. - Yeah, I know I might go back. - Okay, so you went back to SF for adept. - Yes, tell me why, tell me what happened. - So when I was 26, a bunch of stuff in my life happened. You know, some family stuff and also feeling like I had outgrown my job and I did all of the things you do when you have like a crisis. I cut off all my hair, ran a marathon, I applied to grad school and then decided I needed to quit my job. - Quarter life crisis. - Yeah, I still remember February of that year, I flew with a one-way ticket to SF, one-way. And then I stayed on friends' coulches and I was like, I'm not leaving SF without a job. And it has to be a PM job, no longer PMM. And I wanted to be at an AI company. And I just like, I spent I think three weeks literally hitting people up on LinkedIn coffee chats. And by the end of it, I had met an amazing team at adept and convinced them to create like a PM role for me. And they said yes. - Wow, you made your own job at someone else's company. - Yeah, I mean, I was got so, so, so lucky. They were amazing. - Are you type B or type B? - I'm super type B. - Okay, 'cause I can't tell. - I mask it very well. I think I'm type B. I mean, if you ever look at my house, it's like crazy messy. - Okay, that's where the FP is. - Yeah, exactly, yeah. - So you were at adept. You were PMing for enterprise B2B AI agents. - Kind of. - Yeah, I mean, we basically didn't have a product. - Oh, okay. - It was a research lab. It was a research lab that had developed like computer use models and we were trying to sell companies an agent that could use the computer to do a bunch of things that like otherwise would have taken someone a really long time to click through. Forms and like logistics companies and stuff. And I'll be completely honest with you. We were struggling to turn a model and package it into something that we could sell. - I mean, that was a hard time and a hard year. - It was a hard year, yeah. - But then 10 months later, you got poached. - Yeah, the research team at adept one over to Amazon. And we tried to figure out what to do with adept afterwards. We explored a bunch of really cool ideas. I think eventually I was like, I think I need to go to something where there is more product market fit. And I used to freelance as a designer all through college. And I was like, actually, Figma is my dream company. - Figma is a dream company. I want to wear cool clothes and be around cool people. And I hit them up. - Wow, I love that. - You were the founding PM of Figma Make. - Yes, that's lit, dude. - Yeah. - I got reached out to join, I think maybe the FIG Jam team. - Oh, cool. - On product. - Yeah. - But I said I couldn't. I didn't interview at config 24 when they were announcing Figma Draw. So I got to interview the draw team. - Okay. - And so there's a bunch of engineers. - Yeah. - And then they're like, please join. And I was like, no, I can't. 'Cause my work ended me with, yeah. And they're like, who is it? And I'm like, I don't want to ruin people's lives and careers. And then one of them was like, don't worry, I have no power. I'm the most junior one. I was like, okay, I'll tell you. But that person was leaving. So I was like, okay, I can tell you guys. - Honestly, that's a good boundary to have. - Yeah. It's like tough to work with an enemy. - Yeah. There's not a lot of people on my enemies list. But I try not to have that kind of energy around me. But Figma would have been beautiful. - Yeah. - I love the Figma team. They're so cute. Everyone's so good. - Yeah, and everyone's so thoughtful. I actually, I truly think Figma is the nicest company I've ever worked at. - Aww, that's so great. - Okay, so then what happened with the OpenAI? You were at such a great place. You had a great team, the nicest people. You're saying all these praises and then you switch teams. - Yeah, brother. - Figma always has a very, very special place in my heart. I was there for a year. A lot happened that year. We launched FigmaMake. We IPO'd, we built out a huge AI team. A lot of things shifted. And I think a little part of me was like, I just kind of want to keep this momentum up. I do think working at a really big lab is gonna be incredibly challenging, but it's kind of now or never. And I think I was just like, I was really spontaneous about it. Truly, I was just like, I woke up on day and I was like, yeah, let's just go for it. And so, that's how that happened. - Are you spiritual? - Yes. Okay, so you'll feel it in your life. It's time. - Yeah. - The day is today. - Yeah. - Yesterday didn't know that, but today I know it. - Yeah. - I'm incredibly intuitive. And I was just like, I actually don't think I could stay. The body wants a different challenge, so we'll go for it. - And when he told your team, did they cry? - I was super sad. Yeah. - Did they cry? - I don't know if literal tears were shed. - A lot of drinks were had together, but I'm still incredibly close to a lot of people on the make team. - Make was your baby? - Yeah. - You birthed it. - We birthed it, yeah. - You was from your idea, right? 'Cause you were, okay, I read somewhere random, like you were just on cloth and prompting stuff. And you were like, oh, this would be great.
cool and do-do and Figma or something. - Well, this was Dylan's idea. - Oh, okay. - Yeah, yeah, yeah. I think Dylan was like really thinking ahead and he was like, we have to do some kind of like generative with code design tool. - And then I was like, Holly, you're the one for it. - I mean, to be completely raw and transparent, I had just joined Figma. - Yeah. - I was a nobody PM. I had been there like five weeks. I'm being honest, I think like he gave it to me just because there was no one else around to do it. I was like, I will fucking do that coach. Put me in. - Put me in coach, I'm ready. - Yeah. - You're ready for the plot. - And then it was like a crazy sprint and then we like built it in like a couple months. - That's really fast. - Yeah. - So you took that leap of faith. - Yeah. - Your body was like, it's time for your change. - Yes. - And you applied to OpenAI? - Yeah. - Or just a lot of other companies. - I applied to OpenAI, I knew like I had a mentor who worked at OpenAI and she spoke very highly of it. So I was like, okay. She's chat. - How did you get that mentor? - Actually through a book club of all things. - It's gonna have hobbies. - The artists' way. - Oh, of course, yes. - You do your artist daily or whatever. - Yeah, your daily journal, your artist dates. - Honestly, that changed my life the artist way. - I think it did change my life. - It did change a lot of people's lives. - You are now a member of the product staff. - Yeah. - What does that even mean? - I'm just a product manager. - Oh, okay. I've seen a lot of people, a member of technical staff, a member of a partner and I'm like, isn't that a PM or is that not a PM? - Yeah, it is just a PM. - Oh, okay, okay. It's just like the group title. - Yeah. - Without reviewing anything confidential, was it like interviewing for OpenAI? What kind of people were they looking for? Was it hard, was it scary, were you nervous? Were you sweating? Were you like, I got this down Pat, you'll know nothing. - You'll don't know about me, Holly Lee. I popped off at the keynote at Config. Did you? Thank you for saying all that. - I was nervous. - Okay. - I was nervous and I will say, I think the interview process was like, it honestly maybe surprised me a little bit how by the book it was. It was very like, if you go and Google right now, like product manager interview questions, it was like that. - It was like those. - I was like actually a little surprised that there wasn't curve ball, weird AI question, et cetera, et cetera, but I did prepare for all of that. - It was like a lot of sense. - And a lot of execution, yeah, whatever. - Yeah, exactly. And there was like a bunch and they lined them all up like on the same day. And what kind of people do they like recruiting? And what kind of people get offers? Like what kind of people make it to the culture of OpenAI? - I will say, I think looking at the PMs, they're actually like truly quite diverse. Like they've come from like lots of different places. I do think probably the one thing that they have in common is like hunger or just like people who are like very dogged about pushing something forward. I will say like one of the best PMs I've ever worked with is my boss, I'm not like just gassing him up here, but like he literally, I have never seen someone be so completely determined. Even though like stuff is really, really hard, he'll be like, no, we're meeting now, we're figuring it out now. He won't leave until it's done. I do think most people say that about themselves at most companies, but like most places will just be like, oh, we'll figure it out tomorrow. I think OpenAI is really a place where you like figure it out. - Yeah, like you're really obsessed with solving that. - Yeah. And you're just kind of like there is no tomorrow. Like why can't we just do it today? - That's something, like there's a quality I thought I had in me and I think I do. And I'm like, I'm realizing that the people there have it like on mega mode. - If you have one tip for people interviewing for an AI product manager role, what tip would that be? - 100% use everything. Like actually use all of the products that are out there. That seems really obvious, but I've talked to a lot of people who are like interviewing at OpenAI and like haven't used Codex before. - What? And they're doing for a product role? - Yeah, or like they, you know, I get it. There's actually, there's so much to try out there. There's a lot to wrap your head around. But you'd be shocked at I feel like how many people just haven't tried the basic workflows and like had an opinion about what's working for them. We're all just like playing around and like seeing what sticks to. So like in that way, every single one of you is just as qualified to do the job, you know? - If a candidate walked in with an iced coffee in their hands. - Oh my God, this is making you way around on TikTok. - Yeah, is that an ik for you or is that fine? - I feel like there's like a professional answer and then there's my actual answer. - An iced coffee and yellow socks. - I think it would depend on a lot of other factors. But okay, I think largely for a PM interview. - I would judge you more if it was an iced coffee. If it was a warm coffee for some reason, I feel like you are more serious. So like I think a yes, I would judge you. And actually the other day I was on a call with the chief product officer of Slack and I had my smoothie from the breakfast bar and I literally put it underneath the table by my feet 'cause I saw that TikTok in person or on video. - Okay, video. - And I was like, just in case. - Just in case. - I'm gonna put my smoothie on the floor so she doesn't see it. - So they don't think of you a type of way? - Yeah. - All right. Anyway, so we're in a no iced coffee household right now. - No iced coffee. - What do you think people are getting wrong in the interview process these days? - This is my like very honest take. I think people are too unsure of themselves and that comes across as like a type of stiffness or like over indexing I think on the answers that they have prepared for and they like expect that you want from them. - I think it feels rehearsed or something. - It feels rehearsed or like, I know this is a privileged thing to say as someone who has the job but just like even the smile feels fake or you can kind of sense that they, they almost like want the job so badly because of what they think it means instead of like actually being curious about the role and like relaxing into it and like wanting to spar with you a little bit on like the actual ideas themselves. - I understand. - So you know when someone's being performative but to an inauthentic degree. - Yeah. - That's like, that's an Ick level of, of course you have to perform all interviews. You got it. You got to show your best. - Yeah. - But there's a level at which you're like, yeah bro, that's not a real answer. - Yeah, or it's like I need to be able to understand like what you actually believe. I know everyone says this but I've seen it a lot even just like in the kind of like networking dinners and like stuff like that. Like you can really tell when someone has rehearsed a line that they like read on Substack or on Twitter or something. - Oh, they're parading or something. - Yeah. - But let's talk about what you're building right now at OpenAI. So I mean, I'm pregnant right now but you're already a mother of dots. First of her name. - First of her name. - What was giving birth like? - There were a lot of people in the room. - Dots I think is not like, that's a personal time when you start giving birth. - First of her time. - I don't know if I'll have that many people at my birth. - Yeah, I mean I think this is the first birth on record where there were like 80 people birthing one thing at the same time. That's the same thing. - Yeah, I mean it's the craziest launch I've ever worked on just in terms of like the sheer amount of stuff that we built that we thought through that we designed. The sheer amount of like collaboration and just crazy talent from all across OpenAI. Like I truly think almost every single part of that company was part of this launch. It was so cool. - Oh that's awesome. - Yeah. - To the people, what is dots? - Dots are and always on persistent assistant for you. It's help that you can have on your phone, on your laptop and it's the same help all the time. You can call it and like literally just like talk to your assistant and it does work around the clock and it's proactive and it like learns about you over time. And already my dot is indispensable to me. I called it on the car ride over here. I called it to book my flight to LA. - That's like that post. - I called it, oh my gosh yeah, I do post about that. I literally talk to my dot for hours a day in a phone booth just to like talk about our product roadmap and like work through really, really hard problems. It's my second brain. - Wait, so you joined OpenAI seven months ago? - In April, so five months. - Okay. And then you were assigned to do this immediately? - No, I was on a product called Workspace Agents. So like thinking about multiplayer agents for teams and then it basically blended really well into this because we've just been thinking about like how to get your agents into places where you actually meet them like Slack and email and then one thing led to another and then we built this. - I feel like this is a recurring pattern where you get pulled into the biggest particle launch of the year at the company and you keep doing keynotes on it and you keep going viral for it. You got one a year. - I'm curious about the name and don'ts. - How did you land on that? - Dots are just like a really familiar sweet name for something that like you want to be able to talk to every single day. I like that they work plural and singular and you can rename them. But like Dots I think is just like it perfectly encapsulates something that like you want by your side and that like you trust a lot. - So it feels like right now is a great AI assistant race. - Yeah. - What is the difference between Dots and all the other AI assistants that got launched? Like we got Brockbot, we got Muse, we got Instinct, we got OpenClaw, what is the Dots positioning? - The way I think about Dots is you already have all of the kind of connectors, the plugins that like you know and love inside of chat to be tier codex. So like it's very little setup. Like it's already kind of like the workspace that you love and now there's just this assistant there that like already has all that context. And then I think the really big things are just that it's reachable in all of these different surfaces. So it's actually been a really big game changer for me to just be able to call my dot. And then I can call my dot, it can kick off work just like as I'm on the phone call. And then when I'm back at my laptop, I can also just do work in the cloud or even using the stuff that's on my computer. I do think that like we have an incredibly like holistic, simple offering, which is just like a low set up assistant that you can talk to also on the phone. And like honestly for most people, that's where I just start with. There's like certainly power use cases for like developers people who are.
creating more like content on their laptops, but for most people, just like you can call it. And it's really smart. What are the best ways to use dot like the main use cases for it? My like really simple answer is that sometimes I just like to talk to it and like basically brain meld with my assistant. So I like talk through my day, I talk through like whatever big problems I have. So my biggest like use case is literally just having like a second set of like eyes and a second set of brain energy to like think through really, really tough problems with me. But then I also have it like monitor a bunch of stuff in my life. Like when x happens, go do z or like notify me when this kind of thing happens. In my case, it's just like a lot of user feedback or like a lot of stuff happening across the org. But like I just talk to a bunch of business folks today and like, you know, your dot could be monitoring sales or like helping you flag when things are happening across your company. I could go on and on, but like it's an incredible second set of hands to actually just help me stay on top of so much like communication that I have to do. I ask it to draft a lot of things for me. And then I basically just review them. And then I'm able to actually like help out and access like so many more people than I would have before user experience question. Why does it ring when you call it? Yeah. I mean, that's a great question. I think it's like both in part that like it honestly does. I think make it feel a little bit more delightful and slightly human. But then on our end, I think we also just like literally need some startup time for the voice model to work. Got it. Okay. So might change later, huh? Yes. I see. What is the thought process behind putting dots behind the paid tier, not the free tier? Right now, we are still very much like a learning phase for dots. I think we need to really figure out like what's working. How kind of the proactivity needs to be. I do think the intention is to want to give this to as many people as possible. And we just have a bunch of stuff that we need to work out first. Wait, where do you get your user feedback? Right now, we have alpha users like external users. We're also obviously like always looking at Reddit and Twitter. And then we have a ton of internal feedback that we just like go through inside of our Slack channels. Okay. Okay. Where do I get feedback? You can text me directly. Oh, yeah. You got the direct line. Yeah. I'm a little confused how I use Chatcha BT right now because there's Chatcha BT, Chatcha BT, work, codex, dots, and spaces. Yes. What do I use and when and how? My simplified answer is that I think you should be able to use your dot for almost all of the things that you would usually do. So your dot can just orchestrate your codex. So whatever you were using codex before, if you were building things, if you were doing like harder technical tasks, you can just tell your dot to do and your dot will oversee that. If you were going to Chatcha BT for like questions and answers and that kind of conversationality, you should be able to have the same thing with dot. You can just ask dot spaces are a new product that we launched also a dev day, which is just meant to be almost like a canvas in a space where you and your dot can work together. It's actually been really powerful because conversation sometimes can be kind of limiting like your brain can only carry so many threads in your head at once. But we found that by allowing your dot to also just like write to a page, it can create like a chart or like even just like rows of things. Like it's actually so crazy how much humans just like need to see it to do list. I do. I'm so visual. I can't think in my head. I got to think on a paper. Yeah, page. And it's also really cool because it unlocks a bunch of new ways for you to manage your dot. Like you can literally like I'll have my to do list written out and then I'll do a green check mark emoji on all of the things I want it to do right now. And then I'll just say like look at the list. And then you know, I'll do like a question mark emoji for the ones where I'm like go investigate those and like I could probably be 10 times more creative or you could be. But like that's just the tip of the iceberg. I'm really actually super excited about all of the ways that we're going to use visuals and graphics to like manage our agents. Got it. I have a question on all the connections. Why no connection to discord? That's great feedback. And I would love to understand why do you want your dot and discord? I don't use Slack for work. I use discord. Okay. It's free. Yeah. Yeah. But I used to work at discord. So why can't we marry? Great piece of feedback. I'll take that one back to the team. Okay. We'll get work with my proposal. Yeah. But I essentially I have a lot of my clients and yeah, creator friends and we all use discord. Okay. For our and word work team. Yeah. This is discord because like voice channels are so easy. Yeah. And then we have communities or like creator communities are on discord. So a lot of my viewers are there. Okay. And when I was like streaming before. Yeah. Discord was more for the streaming community. And then I just kept having it. And now we have podcasts posted on there in the business. And we got bots are there. And then we have discussions on it. Yeah. For work, we have like a private discord server to manage all of our projects. Kind of like in it. It's very much like you know, Slack. But yeah, paying for software. Totally. I mean, that's a really great insight. That's really great. I'll bring that back. Okay. Yeah. Yeah. Let me know if you need to connect. Let's talk about dev day. Yeah. I unfortunately wasn't able to make it because I cannot travel right now. And I was really sad. That's a great reason. I was so sad. It's a big reason. It's a big reason. It could be, you know, fun. Rome knows we'd see it's like. But I know you presented right after Sam did. Yeah. Sam Altman. And that's like a really big role to take on the very first speaker. So my question to you is what time do you sleep the night before and wake up the morning of? I slept it. 9 p.m. P.m. Wait, that's because so early. Yeah. Because we'd prepared so much that weekend before. The night before I was chill as a cucumber. I was like. We're not the office until like 1 a.m. No, we did rehearsal. I went back to the office to go see everybody in like high fives. Good luck tomorrow. Good vibes. I wanted to see them all. So you were all rested. And then I went to bed and I woke about five to get there for sound checks and everything was great. We woke about five. Yeah. A.m. Yeah. I want to sleep at 5.30 a.m. Today. Oh my gosh. Oh my gosh. I'm a morning early riser as they say. I'm not like you. I know. Almost a little walk right past each other as I get up to get my warm coffee. That you'll hide that I'll hide out of the table. But yeah, I was I was up and I saw them like setting up the stage and everything. Dang. Okay. What was it like behind stage when you're about to go up? It was actually like really fun. I mean, they like mic you up. They do hair and makeup for like two hours before and two hours. That's too long. I know. I we were like back there for a lot. And I mean, I'll be real with you. I'm like pacing. Right. I'm pacing back and forth. I'm like practicing my intro. I'm like. How do you do with people? Do you have a teleprompter? There are teleprompters in the back. Yeah. Honestly vibes are really good. I was feeling pretty great about it. That's good. And you walked on and you saw all of the eyes on you. Yeah. Like I got this. I was pumped. I was ready. I love public speaking. Wait, where does it come from? Beater. Chloe Feater. You ever take improv? Yes, I have taken improv. Yes, and yes, and I have taken improv. I went up there and I was like, oh, I got this. Hello, everybody. Yeah. And he did your live demo. And in my life. And then Daddy goes to you. Yeah. And you're like, uh, hello. Yep. And how did that feel on the spot? I actually had prepared for this contingency plan of voice knowledge. So you knew. Yeah. Okay. Just things are changing all the time. I was like, yeah, there's a chance that this won't work. And so, you know, I was like, Daddy's having a slow morning. No problem. I can just, you know, type to daddy. And then the demo went on. Second half of the demo didn't happen because of a different issue. But, you know, live demos. That is what it is. And that one I kind of had to kind of make up on the spot. You were so smooth. I didn't feel that way, but thank you. Did your hard drop or like, yeah, so there's a whole second half of the demo that no one ever got to see. And I knew in the first half that it wasn't going to happen because I saw that the codex thread didn't start because of a technical issue. And so the entire time I'm doing the demo, I know that the entire second half isn't going to work. So I was like, saying lines as I was like, well, what am I going to do? Like, literally, what am I going to do? We had like three minutes of prepared demo that's just, I have to make something up. So when I got there, I do think there's like a part of the demo where I'm like, um, this is live, everyone. Yeah. And I just kind of like, I don't even remember. I haven't watched it and I refuse to rewatch it. But I think I made up some kind of conclusion. And then he was like, okay, you were a fine girl. That moment went viral on X. What was hearing about that? Like, I, um, I went on Twitter for the first time, like three hours afterwards. And it was crazy because I made this very bold decision to wear a bright orange vest. And so I opened Twitter and it was like orange, orange, orange, orange all the way through because it's hard to miss the favorite color. The video clip of, of your orange vest everywhere. Yeah. And you decided to tweet back. Yeah. It was a bummer. And I felt like we had to own the moment. Okay. I mean, like, now I think we have openly told the public there was an outage during the demo. Unfortunate. Like, we rolled out 20 something products as the demo was going live. So something broke. But I kind of wanted to make fun of the moment and like pay homage to that like 2007, like iPhone demo. Yeah. Where they were like, it's a Wi-Fi thing. So I just retweeted and I was like, damn Wi-Fi. What can you do about it? I got it. That was my first tweet ever tweeted. And people loved you for it. It got over a million views and you're at over 10k followers. Yeah. Yeah. That was unreal. I heard that there was like a particular guy clowning on you for your outfit that day. Yes. There was. And then, and then he got ratioed so hard that he had a delay. Yes, it was. That's so embarrassing for him. Actually, shout out to my friend Ben. Ben. But yeah. Shut up. Ben. Shut up. Ben. But yeah, the guy was like, who does she think she is? Like Scooby-Doo, which honestly, good dig.
- Good dig, Scooby-Dee does have a killer vest. - Yeah. - And then my friend Ben came out to defend my honor and there was a public apology on Twitter, which was like crazy. I was like, he had a public apology. - Well, yeah, he didn't, he didn't mention me, but he was like guys, I deleted it. It didn't mean to be mean, or whatever. It was just a comment. And I'm like, mind you, this is my first 24 hours on Twitter and I'm seeing the whole life cycle of like mean jab, apology taken down. I'm like, wow, this is crazy, this is how it works. So how do you feel about social media? - I am still scared of Twitter. - I would be scared too. - I would love your advice. I'm like, how does frickin' you do that? - I'm not on that platform. - Okay. - Yeah. I just started because I was looking for your account. Well, like I've had it for a while, but I don't use it at all. - But I will, I guess, now that you're on it. - Well, see, we'll see if I stay on it. - How many PMs are on dots? - A bunch. It's a super like collaborative effort. Literally so many PMs have pitched in. - So how did you become the person to present on dots? Was it like Battle Royale? Was it like Rock Paper Scissors? Was it a half potato? - I was very surprised. There was none of that. I worked really closely on the product on a bunch of the behavior stuff. And then one Saturday I just got a message from them being like, hey, have fun. You're going up and that was it? - What? - Yeah. - That was it? - Yeah, they were like, are you in or are you out? And I sent them a saluting emoji. - Yeah. - And that was it. - That sounds about right. Well, you're known for your insane charisma or a on stage. If I wanted to become a better presenter on stage, what is one tip you can want to give to me? - Here's a saying and then there's a tip. I think the saying is like, do not be a loser in your own fantasies. - Whoa. - I got that off TikTok and like online. But basically like when you imagine yourself on stage, you have to imagine yourself just like winning. - Okay. - The best like cares matter. Everyone's falling in love with you. It's so great. Like imagine yourself as a fucking winner, right? - The second thing that like I think actually does help is like you have to remember that you want to tell people the story. And like it's not actually you saying the words or whatever. It's like remember that the whole point of this is to convince them. It's to tell them the story and like bring them in. And when you position it less, it's like I have to do this presentation and more like I have a story. I'm going to tell you the story. It actually becomes a lot easier and like way more fun. - Aww. - And so I was genuinely not nervous. I was like, oh, I genuinely want to tell people the story of blossom music and how dots help you. - That's an artistic take to public speaking, which can feel like mechanical. - Yeah. - Well, I'm so glad that you were able to be on stage. Okay, so I wasn't there. So I didn't like see all the speakers, but generally I would assume that there were lots of men speaking and representing. And that's just the AI tech space man. That's just what it is. A very few women speaking. And also especially in the AI space, like whenever I see AI content, 99% of the commenters, I just know what they look like. So my question to you is, you know, you're working the space so closely. What is it like being one of the very few female voices of AI? - Well, we need to change that. - Yeah. - You know? - Yeah. - I'll just start with like, we gotta be blunt. We gotta change that. I think like we need more women. We need more like creative people. We need more people who have like amazing taste and like point of view to work in this space. - Yeah, it's true. It was like a very male dominated audience. And I am very, very lucky to have been up there with Tajel and our amazing research leads. And I think we're hoping to get more female voices in the room. And I guess the way I see it is if me being there means that like another woman feels more comfortable joining our team. - Oh yeah, for sure. - I'm like, yeah, let's do that. - When are you gonna rock your space buns again on stage? - Honestly, yeah. The next time I go on stage, I promise you Chloe, I will do space buns. - Yo, when I see a fascinista girl, Deva up there, PM Daddy inspires me. - I'm inspired. - That's why I wanted to go a little bigger with the outfit. Almost like to make a point to be like, - I loved it. - I'm not your average AI tech bro. - How boring is it to clown on an outfit? Like, is that the best you got? - Yeah. - I don't like that. - Yeah. - But you questioned up there, I love the colors. - I'm in like this in really interesting space. I used to be a tech product manager and then did a lot of tech content. And then right now I'm more of like a lifestyle, more generalized. I wouldn't say people see me as a tech creator anymore. - Yeah. - But privately I am. I'm still a lot of conversations with my tech friends and I'm in a lot of beta testing groups. - Yeah. - And like I get, I get to see a lot of product demos, a lot of NDAs are signed. So I'm in the space very, very, very deeply in them. Like I share my feedback privately, I ask my questions and then I started off with drama. I'm generally an AI optimist, but I am so afraid of sharing my information with a lot of AI products. I just can't get over it. And I know that in the more I share, obviously the more powerful it is for me. And I'm trying to solve for going to bed earlier so I can be healthy and live a life and watch anime. Do you watch anime? - I don't. - All right. That's the one that we have today. - C-dramas, K-dramas. - Oh my God, did you watch pursuit of J? - I watched, oh my God. Did I watch pursuit of J? - Yes. - Sounding, huh? - So don't get me started. I was obsessed with pursuit of J. I literally sometimes played the soundtrack. - Oh yeah. - Just to like get me in like a very Chinese. - Yeah, you're just listening to that today. - Headspace, yeah. I loved pursuit of J. I actually just finished watching eternal spring. I think it is called eternal spring. - Yeah, yeah, it's like the other one anyway. - Yeah, we are talking about that. - Yeah, I love C-drama. - But yeah, I just want to have more time to watch my C-dramas, my K-dramas, my anime, me, my read my monholym. - Yeah. - And that gives me hope and an excitement in life. - Yeah. - But yeah, I can't get over giving it access to things like my inbox, my text messages, my finances, my health data. And so what would you say about dots to more privacy conscious users? - Okay, so a couple different things there. There's privacy and then there's, I guess, controls around what your dot is able to do with all of those things. The data itself, you have control over how your data gets used. You can actually just go into the data controls in your settings and opt not for the model to improve based on- - So it was like your use or something? - It just means that we don't use it. - Okay. - So that's just like a toggle that you can toggle off today. And then in terms of what dot can do with all of that, it's absolutely true that the more you connect to it, the more it can connect dots and be helpful to you. I certainly think that we want to take as much off your plate as possible. So a couple of things. Number one, just like baseline, the model that we've used for dots is our most aligned model. Then there's the second layer of the product behavior. We've intentionally made dots really conservative. So for example, we actually had dots proactively draft things. So they would be like, hey Chloe, I noticed you haven't responded to this email. I just wrote a response and then you can decide whether or not to send it. And we decided that even that was too much. So we've toned it down even further so that you will even ask you whether or not you want it to draft a response. So that's the product behavior level. Then we've added even more control. So number one, just like to be really explicit, you can decide what your dot has access to. So you can always turn off, I don't want it to have access to Gmail or like, I don't want it to use XYZ connector. And then even when it like is looking at your kind of like information, like scanning your emails, it's always read only. So like it can't write to things unless you give it explicit permissions. Is that the permission for reading write privileges? Yeah, so right now, actually everything just defaults to like only read. And then because we're taking such a kind of conservative stance, but we've also added this thing called like user rules. So you can actually just specify, never do X without asking me. Or on the other side, you can always do X without asking me or like once I tell you you're allowed to. We've heard from people that like, if I want to say buy toilet paper from Target, you don't need to keep asking me if I'm allowed to buy it, like you can just buy it. So we've like put that in the user's hands that they can just decide. And then the last pieces, we actually have like an AI auto reviewer looking at all of these different interactions and it can also additionally flag when something that like you've asked your doc to do has some kind of unintended consequence and it will actually catch it. Yeah, my greatest AI fear in that respect is if it sends an email to someone else without my permission or if I explicitly say it and still does it or it gets prompt injected. And I guess this is how you design against it. But I feel like there's so many reports of like hallucinations and saying I already said that rule. Yeah. So how do you guys go beyond that? Because like I was also in another product demo where they're like, well, obviously, AI won't do anything that you can tell it not to do. Like it won't and I'm like, but that's not how it works. It's not guaranteed. Yeah. So how do you go further in protecting the user controls? Yeah. I mean, these are all really, really great points. There is always this inherent risk as you pointed out. And I think we're taking it super, super seriously that that auto review process is done in conjunction with research. It's like one of our biggest areas of basically how can we train the model to be able to detect when even if the user gave permission a long time ago for this thing, like, can we actually sense that, you know, that permission still stands, et cetera? And so it's one of the biggest areas where we're trying to improve right now just to make it as secure as possible. You know, the prompt injection risk of someone else sending an email and giving instructions. Yeah. How do you guys combat that? Our auto reviewer is basically an AI model that takes a pass at everything and like looks at it holistically and says, is this fishy and like, would Chloe have wanted this to happen? And it's incredibly like accurate at basically detecting issues like this. Are there things that the everyday user should understand about the risk of sharing your information, like this type of inboxer? The basics that a user should understand is that if you give dot access to, for example, Gmail, and then you revoke that access. So, you know, like, you can't look at my Gmail anymore. That doesn't change the fact that up until that point,
the Gmail access that's in its history, that's in its context, is still going to be there. So it will still be able to remember maybe an old email that you've given it, even if at that point it can no longer read future emails. That's just kind of like a, this kind of how like talking to a person is, you know, like if we stop talking, you'll still remember all of the stuff that we talked about until this point. If God does something that is unexpected, how can I see what happened, undo it, stop it, review it, retro it, a couple things like number one dots have this like what we call the thread, right? So like you'll be able to see all of the things that it's done very, very visibly in the app. But we also have this like little activity bar. If you like click on the icon, you can see a little side panel where it's like where, you know, what phone number it has and where it is in Slack and stuff. And in there you'll see like an activity section. And you can also see like long running tasks that are still maybe working in the background that you want to check in on. And those will be active there. Yeah, those logs. If you want your dot to stop anything at any point, you just say stop or like you can steer it with your words. But I think there is a very kind of real AI reality, which is like if it's already taken in action, sometimes those things are hard to revoke. I've had an instance where I sent an email with my own volition like I told it to send it. And then I realized, oh, I actually like shouldn't have sent that email. And I just I told it to like undo. And you know, sometimes it can with it's within the 30 seconds of Gmail undo. But it's just like any human. Some things are harder to control Z. What's a feature that you guys have held back on from building or had to change because it just wasn't safe enough? Maybe a good example of a feature where we were like, oh, we need to work on this more before giving it to folks is actually like today. If you put your dot and slack, it only responds to you. Okay. So like imagine, you know, one day we have it in discord, like it would only respond to you. But initially, I think we, we really felt like it was more natural. We wanted it to be able to also respond to other people. So like, you know, your brother could ask Chloe dot for something and Chloe dot could unblock him and like work with him. And then, you know, that was one of those things where we were like, okay, there actually are a lot of things to think through here. And so we like kept it really, really conservative to start. So your dots only respond to their original owner for now. Okay, so no other people can't add my dot. If your brother acts your dot, it just won't respond to him. Oh, ghost. It'll say it'll literally say like, sorry, I can't respond to you. And goodbye. Bye. It's because I'm a content creator and a public figure. So if anything happens and my name goes out there with that email and I get screen shoted, it's like over for me. Yeah. Yeah. So I'm so particular and I have so much sensitive information in all of these aspects of my life. Totally. Obviously, really wanted to read and then help me. I do connect it to my calendar. Yeah. I feel like what's the worst that can happen. You delete an event. Okay. I can just recreate the event. And I think this is why like the the user rules I mentioned earlier, I think will really make sense because everyone has different thresholds for what they're nervous about. And like for you, you could set really, really explicit user rules that like every email draft needs to go through you or every email draft to these specific people or people of this type always need to go through you. Like you can set those rules in complete natural language. So you're like absolutely in control of how your dot behaves. I'm really excited about AI systems because I'm not like the most technical person or do I really want to be? I don't really have the appetite. So just having like something that is way more human and can be my one source to go to and be able to delegate on my behalf. I don't have to choose models. It's really, really nice. So I really hope that I can be on the content side, someone who can share out a AI education in a good way. Yeah. But I think that it's very difficult to be in this space as a creator because obviously in Silicon Valley in tech, like everyone's like Gungho. A lot of people are Gungho about AI, which is like fine. It's great. It's great. We love it. Yeah. But it is kind of an echo chamber and a bubble. And so when I do go to these product demos in in creator groups, we're all sitting there like this is cool, but unnecessary. Yeah. Yeah. A lot of the demos are like it can buy. It can be your personal chef and buy Instacart. And I think a lot of creators love doing things manually. Yeah. Yeah. It's like the art of going through the process. They want to get their own groceries. They want to talk to people themselves. They want to go through the artistic process. Yeah. They want to choose where they sit on this the flight. Yeah. Yeah. Yeah. So I think that there's product. And of course you're going to explore what you're going to what you can do. And then there's like the telling the rest of the world part. And then we're in that space. Yeah. And sometimes a little bit strange. Totally. What are your thoughts? Because you're on the other side of the spectrum. I and you do marketing and you think on the keynotes. Yes. And this is actually something I feel really passionate about. Because like I'm from rural Ohio. Oh yeah. Yeah. I remember that. I'm from rural Ohio. I'm from the fields. I used to be in marketing. Yeah. I like am an artist in my free time. I am like the person who cares the most about like preserving, preserving, spending hours and hours like touching objects and like creating things. And you know, paradox, a paradox is like I also work in AI and like I am hyper aware that on Twitter, everything is like, how can I get my AI to automate X, Y and Z for me? I am with you. I'm like, I want to walk inside the grocery store and not just have like AI order stuff for me. Like I want all of that. The reason why I think dots can still resonate with people is because I actually feel like my biggest use case for dots. And I know I'm like, you know, I've said this a few times. It's like I literally just talk to it. And sometimes it makes me just feel like I have, it's not even it just makes me feel like I actually, I feel like do have more control over my day because all of us have so many different things going on. Sometimes just the psychological impact of like knowing that you've like put something down and that there is someone who's like, hey, I'm flagging that, you know, you just like contradicted that thing that you said yesterday or you still need to make time for this other thing that you said you were going to do today. I just kind of like just feel like a more sane human to be able to pour out into something. And anyway, I've done a lot of reflection on this. But I guess my point is that I think dots and just this whole era of like AI assistance and whatever now and we end up on is going to be a lot more than like efficiency. I think that like efficiency is almost never the point. It is just like helping you regain control. And for some people, maybe that means handling their door dash and their code and helping them, you know, draft things faster. And I think there are a ton of work use cases. But I think for at least maybe people like you and maybe people like me, the thing that's resonated most is actually just like regaining agency over your life and just like feeling like the hard. Yeah. So I feel like a lot of the marketing and demos are how to make a human superhuman essentially like gets the next lover or whatever. But I'm not like that interested because when I'm out of my 100 percent, I feel like I can do anything. I don't need really need help. I can do everything. Yeah. But once I became pregnant and I'm going to go through postpartum, I give birth postpartum. Apparently, I'll have a lot of brain fog. I think you still have brain fog right now. But apparently I will just be at my 10 percent. And that's just the reality. And I won't be able to know what you're saying or or cash things because whatever. But I still have to be the command center of my entire household and all the work that I do and somehow catch everything. So I'm actually more interested in helping what it is like to be at 10% weight because you have to because you got to step away from light. You got to take a break because you physically just can't anymore. And so something else can help you while you're out. That's what I want to help with because I can't do it. Yeah. I mean, there's something really empowering just about giving everyone this assistance when they need it. And I think yeah, that example of like postpartum, I'm like really, really interested to see how dot can lighten that load for you and like make you feel more confident if you can solve it for me. Yeah. Great. I have very little time left. I would love to work with you. And I think dot would be able to help you like just like know that a bunch of stuff is handled. And just like it'll check in with you and be like, hey, I'm going to do X, Y and Z. Do you want me to do these things? I love that. Here's a report on like how your last video is doing. Yeah. If you want me to help draft like X, Y and Z, I can do that for you. Yeah. Yeah. I just use cases for us to watch more training dramas. Yes. That's the ultimate goal. Well, it's funny. You mentioned that because like literally that got me through this entire launch. Oh, really? I mean, we worked really, really hard. Yeah. And I would come home and no matter what, no matter how late it is, I would watch one episode and then go to bed. Yeah, those are hell along. And they're long. And it's always on a cliffhanger. Yeah. So you stopped yourself after one episode. No, there were many nights where I was like, okay, three or four. And there was at least one night where I've turned into eight. You were watching on the side while you're doing work. No, I can't do that. That's too much. I got to give Dongling home my total attention. Sometimes I would watch it while editing because like when it's like processing, I'm and I'll just keep watching it. I think your profile is like stuck. That's a great example. I watch it like 10 seconds at a time. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Do you watch AI dramas? Reluctantly, only because I get little ads for them sometimes. And I have been sucked in like at least once or twice. I watch a lot of that like AI history things. Have you seen those? No. Where they're just like, I'm a Viking in 1073. Okay, so like visualization. Yeah. Yeah. Yeah. I actually really like those. My parents really like watching AI dramas. Okay. It's like on TikTok or like on their own websites. And they're like, it's really funny because sometimes their face changes and it's inconsistent. Yeah. And like the the protagonist space.
and the villains faced swap and they really like finding the moments. Oh my gosh, that's kind of cute. Do you know that one Chinese movie that was the one with the cow or the machine? Yeah, it was really large. Yeah, and it was like so bad it was all manually made by the mom and her son. Yeah, they're playing theaters here like this month. I need to watch that. Oh, I knew I knew I knew I so it's a cow. Yeah, they made a lot of money. Yeah, it was so bad. It was so bad that people watched it that there was a hilarious. It was like a thousand dollar budget. They like gross like gazillion dollars. Yeah, I made like the characters or drew it herself. Honestly, dream, that's folk art right there. That's folk art. You need to collect that. Yeah, I need that. That's so funny. Anyways, it's all for the dramas. Yeah. Which I understand deeply. Yeah. But as someone on the product side who's done marketing and me who is on the creator side, how do you think we should bridge so that the story of AI is a little bit more, I don't know what it is. I just think that AI marketing is like really challenging. Yeah. And on the creator side, we're all hesitant. We know we're just going to get hate comments. We know we're going to back lash. Yeah. And so we're very, very sensitive about the messaging, about even attending these events, about posting about it. And I try my best to have a very balanced perspective as like an a optimist of the art. But I am terrified at times, and I'm just like, I hold my breath a little bit every time I post about it. Oh, I can imagine. Yeah. Yeah. I think it's really real. My honest thoughts are that like we ought to be doing like way more to talk to the rest of the world. And like focus less on like the 1% of Twitter that's going to care about like what's going on. And like specifically like make sure that the people who need it most, I know this sounds cliche, but this is like actually true. Like teachers, students, mothers, et cetera, like actually get the benefits. Dots can actually help a lot of people, save money, start businesses, scale their ideas. Like there's actually so much benefit to it. And one thing is like I really don't want it to be seen as just like another Silicon Valley phenomenon. And I would love to actually have it reach the people who need it helpful for the everyday person. Yeah. Because this is maybe our most accessible product in terms of interface. Like you can just call this thing. You can just talk to it. I was just with the user who like didn't type anything to it at all. They just like gave it a ring first thing and then started talking about projects. And then immediately we're like, oh my god, I can just talk to it. And it will just like start working with me that way. Just imagine all of the people for whom like that's a huge unblock. I really want to do that. I want to focus my content on helping the everyday use case instead of doing like these are really obscure. Yeah. Or flows to hyper optimize for mimaxing your life in ways. I'm just like, uh, okay, well, I guess like that's cool. Yeah. Anyways. Like that. Those are the type of use. Because I'm just like, okay, yeah, we should talk about something else now. So I really want to focus my content around there. I just don't think there's enough of that type of storytelling. Yeah. So hopefully we can work together. Yeah. Yeah. Yes. Yes. Yes. Marriage. Like, I think that you are living in such a, you have like two personalities. Yeah. And I think that that's really awesome to ground yourself in the real world in the past and the history and the origins and the human experience. But then you're all set the cutting edge and you're literally like at the end of day announcing for perhaps the biggest AI company in the world. And so what is your own personal fears with AI development and what guardrails do you want to keep in your position of power? We have these conversations all the time. It's important for us to just like stay in the room with research. And it's been really, really heartening to see actually how much of just like our influence of like, hey, we think the product should behave this way. Like this is what we're hearing from users. Like actually does affect how research like trains these models. And so just being part of that conversation, it's like our personal responsibility to like make sure that whatever the users really like need is like what gets communicated that way. This is a really big responsibility to like help figure out the next couple versions of these like product shapes and like what is most impactful. And I think even having this conversation with you today and being like, how do we get this in the hands of more teachers and students and folks who like actually have real applications for something that could lift their burden. I'm personally very, very passionate about like getting it to them and like seeing how they use it. And I would love for that to be the feedback loop. How do your parents feel? They're educators. Yeah, I mean, my mom just sent me a photo the other day where she used chat to be tea for the first time to generate an image of herself in the 80s, like a wig and like a giant neon pink suit. So I think my mom is very far from using personal agents, but actually was really interesting. And I haven't talked to her since this launched. I think the thing for my mother is that she's not super tech savvy. And I actually think this is one of those areas where like the conversational aspect of this, like she can just speak to the agent in Chinese, like it can just like talk back to her and like do a bunch of stuff so she doesn't have to like learn how to use chat to be tea. My sister is like a working mom of two boys under three. Oh my God. And like two boys. And she's a doctor. Oh my God. And like also like, you know, helps run a household. And I'm like, I really want to get her dot to like help figure out like the boys schooling. So she doesn't have to like remember when I paid the bill and like, you know, help her manage all of the things and the dog and et cetera, so that she can spend more time watching Netflix, like low key, truly like that's what she wants to do. Yeah. So we all did. And she spends so much time in paperwork and admin. And I'm like, I think there's a huge way that we could just help relieve a lot of that. I think it's very valid. I'm very lucky to be here to finally meet someone on the product team at OpenAI so that we have this connection and this bridge. So we can build together and share, I can share feedback. That's a part of the product management process. I really miss. So as a creator, I feel like my responsibility in the tech and AI space is to put my ears on the floor and like, hear what that everyday person is going through and bring that back and funnel that back into the rooms that might not have access to those people because, you know, I've been in those slack beta testing channels and I've been in one where there's like 200 people and there's literally two women and it's me and it's other woman. Yeah. And I'm like, oh, scrolling the list. I'm like, wow, this is fun. So the way that I look at partnerships is I very much want to have a long term understanding of like, I'm also here to give you feedback back so that we can build responsibly and build for the right use cases and then just share that story so that people who are going through the challenges can actually have a good solution for their own lives and take back their own agency and power and freedom, understand kind of the risks involved and be able to make informed decisions, not to further taste to technology too much and still build their foundations and still focus on human connection like that's, that's what I'm about and girl. I think you and I were a good pair, we're a good pair. So I'm very thrilled to finally meet you like before we finally wrap. I do have one last section and this section is called the performance review. Oh my god. Have you had it yet? Open AI? Yeah. Every six months? Oh my gosh. That's so fast. Yeah. Well, legally, our performance review does not affect your compensation, your title, but it does affect your aura and that's really important in 2026 but you already have really high aura. So Holly, are you ready for your performance review? I'm ready. Question number one, would you rather be right or be promoted? Right. Ooh. That's a Libra. Would you be down to make $1 million but you'd have to teach an online course about public speaking on stage? Yes. And then you get the poster battle on LinkedIn and X and feel like, hi, I'm teaching an online course on stage. Take my master class. For a million bucks. Yeah, she would do it. Would you ever write a post starting leaving my, I don't know, 500K tech job to start a dot, dot, dot, if it would promote your next venture? No. Let's just say, let's just say your next venture is opening an art boutique. Fine. Yes. Okay, she would do it for the right reasons. Should filming a day in my life working in tech be banned or permitted in the office? Banned. Ice coffee during interviews, banned or no? Banned. Working with main character energy right now, or are you in your NPC era? A hundred percent main character. Do you think you're living the right era of life right now or were you meant to be alive 200 years ago? Definitely not 200 years ago. Okay, yeah. I think this one suits me. Meetings before 90am, human rights violation or personal problem. Personal problem. Staying at the office until midnight, human rights violation or personal problem. Girl, you're asking the wrong person because I'm doing both of those things. Out of 100% for a promotion. What percentage is skill and what percentage is visibility? 40% skill, 60% is ability. That's reasonable. Yeah. Is Gen Z cooked or trainable in the workplace? I think Gen Z's fine. I think we're fine. Do I get to say we? I don't know. I think Gen Z's has its values in the right place. Like the most critical questioning has the most boundaries, like I think it's good. Are we currently in a literary crisis or do you think there's just too much social media? Oh my gosh. I don't think we're in a literary crisis but maybe I'm in a social media bubble where everyone around me is reading a lot. What have you been humbled by recently? Twitter. Oh right. Yes. My outfits, apparently. No. No, your outfits are amazing swag. Marketing or product? Product. Live demo or p-recorded? Live demo, baby. Okay, so figma make and dots are hanging off of a cliff. Who do you save? Dots. Why? I'm working on them now and I think dots are going to have like major potential. Okay, so figma make dots and jungling her are hanging on a cliff. jungling her. What do you say? jungling her. All right, dots, you're dead in the water. Do you think this video is going to make
to get to China. - I don't think so. - Can we do it? - We'd have to post on Xiao Hongshu, you know what I mean? - Can we do something? - Promote this, like two Chinese techie girls. - So what are we gonna do if he's in the room? Like what do we do? Do we just look at him? - I just wanted to see this video. - Okay. - Take a pay cut to be a manager or stay as a super icy. - I would take the pay cut. - And you wanna be a manager? - I think managers have the potential to change, like, meaningfully change people's lives. - Okay, so you like people. - I like people. - I take it, you don't. (laughs) - I like some people. If you found out that your underperforming work enemy is moonlighting two full-time tech jobs, do you narc? - I don't narc. - You don't narc? - You don't know what they've got going on in their life? - But it's your work enemy, like you literally hate them. They make your life miserable. - I am. - You're complaining to your partner every single day about them. - I have experienced this and I would still not narc. - Oh, you're not a spiteful person. - Is any PR no matter how bad it is, good PR? - No, absolutely not. - Okay. - Absolutely not. - New York, San Francisco, or LA? - San Francisco. - What is scarier? The rats in New York or your live demo going wrong? - The rats. - Oh my god, this freaking rats. - And finally, what are you trying to figure out this year? - Trying to figure out what kind of like creative career I want to have. - A creative career? - Yeah. - Outside of work. - Yeah. - What are your three options? - Art, buying, consulting, lifestyle business. I don't know, I just, I want to do something where I spend a lot of time looking at and thinking about art. - I love that. I can't wait to hear more and see your creative career journey except that's what I'm all about. - Yeah. - The corporate creative type thing. - I know. - That's what the viewers are all about. - I know, I need all of your advice. - Yes. - Oh my god, thank you so much. So Holly, you have received an exceed expectations. - Thank you, thank you. The Asian kid and me loves to hear that. You know that. - Congratulations. - Thank you. - Tell your parents. - I will. - So, you know, Holly had a really long day, she literally just flew in to LA from San Francisco. You've been going meeting after meeting being a girl boss in the corporate tech AI space. Holly, if people wanted to follow you, where can they find you? - Well, now I am officially on Twitter, so you may find me there. - Are you open to people called messaging you on social media or email? - Yes. - You are? - Yes. - Because you like people. - If they are genuine and want to chat about creative stuff or career stuff, I'm down. Where can they message you? - DM me on LinkedIn, Twitter. - If people want to reach out to you, what do they have to do to increase their chances of you responding back? Because you're a really popular girl right now. - I will try my best to respond to everyone. It might just not be a phone call. Even if you have like three questions, I'll boop some answers back. - Very nice. - I think that you'll have a higher chance of that than having a life. - So you're someone who texts back. - Yeah. - Ooh. - Yeah, for sure. - Especially if you're like a student, you're like a young woman trying to figure stuff out. - Absolutely. - Yeah. - So if you're not a student and you're not a woman, you're cooked. - Yeah. - Hello, Q, maybe. - Which you come back for a future office trauma episode? - Yes. - Oh my God, yes, Holly. Yes, this is not our last date. She said yes to a second date. - And what episode of your life or what chapter of your life would you want to come back for? - Maybe if we complete this arc of like tech to creative stuff, that could be a fun little conversation that I can ask you questions about it. - We have a lot of friends in the guest list that have gone through that arc of like doing tech, becoming a dancer, doing tech, becoming a singer, doing tech, I'm starting their own pottery business or their own like side project to be an entrepreneur but also keep their feminine side, you know, stuff like that. - Yeah. - Yeah. - Okay. - Well, I need to tap into that. - Stay tuned for the future. Thank you so much for watching and we'll see you next time. - Thank you. Thanks Chloe. - Yay. - Shout out to our drama llamas on the Patreon. Hope you guys like the content. The behind the scenes content we've been posting on there, the little mini blogs of my interview days. Anyways, let's talk about this episode. This interview came together so last minute, like literally Tuesday was open day to day. I didn't get confirmation from hauling her team until like Wednesday night and then I stayed up that night. Putting together the interview guide, she flew in on Thursday, she had a full day of work and then she came over to my studio around five. We started shooting at 5.30 pm so very sorry for all the lighting changes. I just really like natural lighting. I don't often shoot in the evening so we weren't quite prepared for that but hopefully it wasn't too distracting. Also, I'm just so grateful and excited that I got this chance to interview perhaps. The most talked about PM in AI and tech right now. I feel like that's such a crazy thing to stay out loud. I have never been able to get a scoop like this before and I think that's because we, you and I, have built off a drama to a point where all of you are watching, y'all are supporting and people are reading the comments and they see the brands and the companies and the people who don't really know me and they see that you're real people. And not just like, I don't know, evil tech bros or bots. There's a real conversation going on in the comment section and it builds credibility and trust with the other people who are down to send over their product or tech leaders to be on office drama in a non-stuffy, non-tech broey format. And you guys make our guests feel really safe and excited to come on podcasts. I'm not gonna lie. I have been asked to be on so many tech business AI even creator podcasts and I say no to a lot of them because if you look at their comments, it's kind of gross. Like, there are people who are full of hatred, negativity and judgment, they judge all their guests. It does not feel safe. I would not wanna go on a lot of these podcasts even if they do better than off-stroma from a metrics standpoint. So I deeply cherish this community that you and I have built together. Thank you so much for building this little corner of the internet with me. Also, before and after this interview with Holly, we found out that we have so much in common. Okay, we're really different because she is an INFP and I'm ENTJ-like, we're very different in that respect. But we have our mutual friend, Aleric, who is my childhood friend. And they know each other because of Penn and McKinsey and they do a lot of music together because she's into musicals and she's a theater girl and she likes singing. I was like, do you know a showtunes piano bar in San Francisco called Martinis? And she was like, yes, and I'm like, did you ever go to their Monday night 9 PM showtunes piano bar nights? It's like an open mic night with Joe the pianist and she was like, yes, and I'm like, what? Literally no one knows what I'm talking about every single time I bring it up because it's like one of those, if you know you know places and I was like, did you ever sing up there and she was like, yes, I sang popular from wicked and I was like, what, that's crazy, she knows the pianist. I've been there for almost every Monday night in my 20s for three years. It was just like my save haven. It was a place I really love in SF and there's not really a place like that I have in LA. So it's such a special place in my heart and you know, we bonded. We really bonded through that moment. Also, Aleric was in my wedding musical if you guys know my personal life and you've seen my longs before. You would know that I, at my wedding last year, had a surprise wedding musical and Aleric was in it. He sent that musical video or maybe short form to Holly. So Holly was like, wait a minute, you're the girl who had the wedding that Aleric sang and he sent it to me last year and I'm like, ah, yeah! So all the dots were connecting was a crazy moment, anyways. I love her. I think we're soulmates moving on. As much as I love content creation and being in Los Angeles, I've always been a tech girl through and through and being able to connect the bridge content in Silicon Valley has been such a fun space to be in. I think my deep dark 30 secret is that. I don't know why I said it like that. But I really believe in this format where we can deliver tech career, business insights in a girly pop casual kind of way. It doesn't make you feel like you're hustling and bustling. I try to deliver these conversations that are like deep, meaningful, educational, but in a way that isn't just for the tech, tech people. It's like good for the rest of the world. I feel like I have two wolves in me. One is that sweet, happy, go lucky content creator that doesn't take life too seriously. And the other one is a very type A E and TJ corporate person. Like that's who I am deep down inside, I guess. Like I'm so serious when it comes to work. And I feel like we're finally in a place with all this trauma that we can bring on people who have both. Like we're not all just stuffy business people trying to min-max and optimize and build for their shareholders. We've got life identity perspective philosophy and we just want to do a good job in our own careers. I feel like we are so early in our podcasting content journey. But it's okay, we're gonna build this up. It takes years to build a strong podcast audience and a strong base. Consistency is king. And you know, I'm gonna keep pushing my best. And maybe one day we'll make enough money and builds up strong enough team where I'm not staying up until 5.30 AM, trying to put together an interview, especially not while I'm in my third trimester of pregnancy. Anyways, what do you think was the most fascinating part about this conversation? I feel like it was a little bit different. It's more current news. It's not as much general career as the previous episodes because we got to see a lot deeper on the product side of things. I am so fascinated by cutting edge technologies. I have so many questions. I never got to talk about this in my short form content. Also, a lot of companies don't really talk about their product in the way that we were able to with Holly. So we got really lucky here. Let me know what parts of it you really loved, what parts of it you think we should do more of and how much you love Holly because I love her too. That's all for now and I'll see you next week. Bye!
Podcast Summary
Key Points:
Holly Lee went viral after presenting at OpenAI's Dev Day without being on social media, sparking global attention and reconnection with old friends.
She created her own PM role at Adept by convincing a team to hire her, demonstrating initiative and strategic networking.
Holly emphasizes authenticity in interviews, urging candidates to use real AI tools and avoid rehearsed, performative answers.
Dots, her current AI assistant product, is designed as a persistent, proactive, and accessible assistant available across devices, with strong focus on privacy controls.
She advocates for women and diverse voices in AI, noting the need for more inclusive representation in product leadership and design.
Holly highlights the importance of user control, with Dots offering read-only access, explicit permissions, and an AI auto-reviewer to detect unintended actions.
Her presentation at OpenAI was marked by a technical demo outage, but she handled the moment with humor and transparency, gaining public admiration.
She promotes self-belief and confidence in public speaking, advising people to imagine themselves as winners in their own fantasies.
Summary:
Holly Lee, a product manager at OpenAI, gained widespread attention after her standout presentation at OpenAI's Dev Day, where she launched Dots—an always-on, persistent AI assistant accessible across devices. Though not on social media initially, her viral moment stemmed from an unscripted, authentic performance that resonated with audiences globally. She credits her success to deep user empathy, a strong belief in her own capabilities, and a commitment to authenticity in both her work and public presence.
Holly emphasizes the importance of privacy for users, designing Dots with strict controls—such as read-only access, explicit permissions, and an AI auto-reviewer—to prevent misuse or unauthorized actions. She also highlights the gender imbalance in AI leadership and advocates for more diverse, creative, and inclusive voices in tech. Her journey from a junior PM at Google to founding Figma Make and then joining OpenAI exemplifies her drive to build meaningful, user-centric AI tools.
Holly encourages interviewees to use real AI products and approach roles with curiosity, not performance. Her experience with live demos, including a technical setback, shows resilience and humor, reinforcing her belief in transparency and human connection. Ultimately, she sees her role as not just building technology, but fostering trust, creativity, and balance in a rapidly evolving AI landscape.
FAQs
Dots offers a persistent, always-on assistant accessible across devices with minimal setup. It integrates with existing tools like ChatGPT and Codex, and its key advantage is proactive, context-aware help without requiring complex configuration.
Dots takes privacy seriously by defaulting to read-only access and allowing users to control what data it can access. Users can disable specific connectors, and an AI auto-reviewer checks for unintended consequences like unauthorized actions or prompt injection.
Her presentation went viral for its authenticity and energy, especially after a live demo technical issue where she humorously addressed the failure, showcasing resilience and charisma. It highlighted her role as a prominent female voice in the AI space.
She advises imagining yourself as a winner on stage, not a loser, and emphasizes telling a story to connect with the audience rather than just delivering facts, which makes presentations more engaging and natural.
She felt a need for a new challenge and was inspired by OpenAI’s culture of innovation and urgency. After a year at Figma, she joined OpenAI to be part of a high-impact, fast-moving AI product launch.
She recommends using all available AI tools and products firsthand to gain real-world experience and form informed opinions, as this demonstrates genuine engagement and practical understanding of the technology.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.