Lenny's Podcast: Product | Career | Growth
00:58:59 · Nikhyl Singhal
[00:58:59] S1: I kind of am pretty all in on Claude these last three months. I was, for a month, pretty aggressive on Codex because I found some of their newer stuff to be, especially with the highest level of reasoning to be quite advanced. I find it hard to switch between tools I try to standardize to be direct. Look, the things that I build, I build a bunch of web properties for my community. So anytime I see something that I can obsolete and code, I try to build code around it. So if you had 100 people, one natural thing you want people to do is to meet each other and 100 people can't meet 99 other [inaudible 00:59:42]. So you have to be thoughtful around who's the best person to meet and how do you match people up? How do you make sure that you match people with people they haven't met before? What are their haves or their wants? That entire thing used to be me sitting down and thinking like, "Oh, Jay-Z would really appreciate meeting Annie, but I don't know if they've met before." Now I write software to do that. I write an agent that goes in and actually does matching. I write an agent to figure out, "Hey, what are all the jobs that my head of products are hiring? How do we make sure that we make those available, but then build a mailing list of folks who I think when they're interested in work can get matched up automatically." So the next generation of recruiting. I think a lot about using AI for content. So when I sit down, I have an AI that takes questions from people and then it's trained on my content and it gives answers. But then I read those questions, evaluate the answers, and I'm like, "Hey, this is a theme I'm hearing." And then I sit down and I write down, when do the LLMs and I disagree, I go through. And so all of these things or anything that I'm doing that I think I can replicate, I try to obsolete myself. When I started my first job, I asked the best engineer that I worked with, who still is one of the best engineers I've ever met, "What's the definition of a great engineer?" And I thought it would be like, "Oh, someone who's got this degree or is well versed in this technology stack." He's like, "Well, the best engineer I know is my dad." And of course, this was before tech in that case. And he's like, "And my dad's definition is still my favorite, which is an engineer is someone who obsoletes themselves from everything they do.
01:01:34 · Nikhyl Singhal, Lenny Rachitsky
That's the definition of a great engineer." And I've taken that to every job that I've worked at. And it's funny because you'll run into people that are like, "Oh no, no, I don't want to obsolete myself. That's my job." And I was like, "Ah, I think if obsolete there'll be a better job for me." What AI has done that has basically put an [inaudible 01:01:57] on that. If you can obsolete by just asking the AIs to do this for you. So why I'm saying that is my stack is what can I do to obsolete anything and everything I do on a daily basis? [01:02:10] S0: That is super cool. I love hearing these stories. What I find as the best tactic here is just solve your own problem. Think about something that's not great in your day-to-day or something you just want to improve and just like... And it sounds so easy. All you do is go to, I don't know, Lovable or install Claude Code, download Codex and just tell it. I want to build a dashboard to control my Sonos and then it just tells you, walks you through everything you need to explore. And so that's a pretty advanced thing to try, but it's just like English. It sounds like I have no know-how to do this. You just tell it. Here's what I want to do, just like a person and it helps you figure it out.
01:02:10 · Lenny Rachitsky, Nikhyl Singhal
[01:02:10] S0: That is super cool. I love hearing these stories. What I find as the best tactic here is just solve your own problem. Think about something that's not great in your day-to-day or something you just want to improve and just like... And it sounds so easy. All you do is go to, I don't know, Lovable or install Claude Code, download Codex and just tell it. I want to build a dashboard to control my Sonos and then it just tells you, walks you through everything you need to explore. And so that's a pretty advanced thing to try, but it's just like English. It sounds like I have no know-how to do this. You just tell it. Here's what I want to do, just like a person and it helps you figure it out. [01:02:53] S1: Well, and I agree with you. And then I think the question is what skill do you have to have? For a while I was like, "Oh, that's a systems engineering skill because you need to be able to train the AIs to go after things." But I'm watching my wife and how she works, and she's not an engineer by any means, and she's getting tons of value. You don't even need to be an engineer. You just have to be opinionated on what you want to see. [01:03:16] S0: And know what you want. Yeah, know what good looks like.
01:03:16 · Lenny Rachitsky, Nikhyl Singhal
[01:03:16] S0: And know what you want. Yeah, know what good looks like. [01:03:18] S1: And then the moment that you get it, it's when the light bulb turns on when it wasn't working before. And now you sit down and there's this moment where it's like, "Hey, I used to have to go and work to do this, but now my agent's doing this. And then what's the next agent I could do to make my life a little bit better?" And that's where you catch the bug. [01:03:37] S0: It's such a powerful thought that your job can be much more joyful as a product manager, as any function, but we're focusing on PMs. Just like if you look at your pie chart today of how happy you are during the day, you can actually be much happier in this future. And I think that's hard for people to really think, just realize, "Oh, wow, okay. I can actually love this job so much more? I did not imagine that."
01:04:03 · Nikhyl Singhal
[01:04:03] S1: Yeah. And I think that it's because you're staring at change. And I think change is such an alarm bell. It's hard to hear the voice that says the thing that you are changing or the things you don't love because change is so hard for us that it's so scary. And I respect that. So I think that if your listeners are able to sort of understand that, hey, there is a world which is better, but I have to go through the tunnel and the tunnel sucks. And the tunnel may mean that I'm going to have to change jobs. My prediction is the vast majority of people that are listening to this will be in a different job in the next five years because they will either choose to move to something because that company will struggle to stay modern just like we're describing. People need to stay modern, companies need to stay modern, or their companies are going to be shedding and returning off staff. And so because change is apparent, I think it's just so deafening that people are very much gun shy to be excited about something that's forced on them. And let's be honest, it's been forced on everyone and it's not something that they chose. So that's why I have a lot of empathy. But when I talk to people, I try to dig through, forget what's happened and why. How do we make the best out of it? And the best is pretty good.
01:05:25 · Lenny Rachitsky
[01:05:25] S0: So say we've convinced a listener, "Okay, I need to make a change. I need to lean into this. I need to take this seriously. I can't just sit back and hope it all works out okay." And then they maybe found a moment of joy. They built something that's super cool and just like, "Hey, check this out." What other advice do you have for folks listening to help them be to survive... I don't know, let's not say survive, but just thrive in this future that is emerging for product people.