从管理人员到管理 AI:当下人人都需要的领导力技能 | Julie Zhuo
From managing people to managing AI: The leadership skills everyone needs now | Julie Zhuo
Introducing the Guest
Lenny Rachitsky: We’re seeing this kind of flattening of orgs. Everyone’s becoming an IC again.
Julie Zhuo: It used to be, okay, I don’t have the skills to do 10 different jobs, but now with AI allows me to do many of those jobs myself. We need to dissolve the boundaries of these traditional roles and call ourselves builders. I’d love for us to get to the world where that’s the title.
Catching Up Briefly
Lenny Rachitsky: I also just saw a stat Google let go of so many of their middle managers.
”The Making of a Manager” Reissue
Julie Zhuo: Management is still really critical. You have a north star, you have a vision, and you’re just trying to figure out how to use the resources that you have to get that thing done. Used to be people, but now it’s basically models and different models have different strengths. You have to assemble the Avengers so that you can use the right tools for the right purposes.
Where Management Meets AI
Lenny Rachitsky: What do you feel is the biggest change in the role in life of a manager these days?
Defining Goals: The Core Skill
Julie Zhuo: It’s always been manager’s job to manage change. I just think the rate of change is accelerating. Today management is really about this idea of be sturdy while being flexible. So I think about this metaphor a lot of the willow tree. It can survive a lot of storms, disasters, et cetera, but it’s also very flexible.
The Trend of Flat Organizations
Lenny Rachitsky: You have such an interesting trajectory from being head of design to now being obsessed with data and analytics.
Julie Zhuo: You want to diagnose with data and treat with design. Data is not a tool that’s going to tell you what you should build. I don’t actually think a lot of the fast growing companies are using data well at this point. Traditionally things just didn’t grow that fast. These companies are totally getting by on just good instincts and good vibes, but what always happens is eventually things stop growing.
Evaluating Competence in Hiring
Lenny Rachitsky: Today my guest is Julie Zhuo. Julie was my first ever guest on this podcast, which I recorded over three years ago, so this is a very special conversation as I’ve shared many times before in other places, Julie’s newsletter The Looking Glass was the inspiration for my newsletter and basically led to everything that I do now. If you’re not familiar with Julie, she was the longtime head of design for the Facebook app used by over three billion people. She’s also the author of the best selling and very important book The Making of a Manager. And most recently she started her own company, Sundial, which is an AI parent analyst used by companies like OpenAI Gamma and Character.AI. Julie is one of the most thoughtful and insightful product leaders that I’ve ever come across and she’s also got one of the most interesting perspectives on product building.
Having worked at a mega large corp like Meta as head of design and now as a founder at a tiny startup that’s all about using data to help you make decisions, it’s really rare for someone to have this spectrum of experiences. In our conversation, we talk about how learning to be a great manager directly translates to learning how to use AI tools extremely well, which specific skills will become more valuable in the next couple of years, her most valuable and timeless advice for new managers, why she’s not hiring product managers at her startup, her simple heuristic for knowing when to use data and when to use intuition in making decisions. There’s something in this episode for everyone. And if you enjoy this podcast, don’t forget to subscribe and follow it in your favorite podcasting app or YouTube. It helps tremendously. And if you become an annual subscriber of my newsletter, you get 15 incredible products for free for an entire year, including Lovable, Replit, Bolt, n8n, Linear, Superhuman, Descript, Whisperflow, Gamma, Perplexity, Warp, Granola, Magic Patterns, ChatPRD and Mobbin.
Julie Zhuo: Thank you, Lenny. I’m so excited to be here. I’ve been looking forward to this all week. I love your podcast. I love where you’ve taken it since our very first conversation and I’m super excited to have a fun and engaging chat.
Breaking Down Functional Silos
Lenny Rachitsky: Can you believe that first episode, the very first episode of this podcast, was over three years ago at this point? Holy shit.
Julie Zhuo: I’m not sure you had that fire in the background back then.
Functions Most Accelerated by AI
Lenny Rachitsky: So funny enough, I don’t know how many people have noticed this Easter egg that I’ve stuck with, in that first studio, I was just watching the episode, I had this funny little mirror. I don’t know if I had in the first episode with a fireplace that was showing up in that mirror because the mirror was showing something stupid. And so I’ve just kind of kept this fireplace across every studio I’ve moved across in my various places.
Julie Zhuo: I even remember we chatted. Video was kind of a newer thing. You’re like, “We’ll record it, but it’s really about the audio.” And now we moved into the video era.
Accelerating Personalized Learning
Lenny Rachitsky: So as you were saying that, I realized my fire was broken, so I just had to turn that on. So we just cut a little piece. Yeah, that fire was my little funny strike for myself and I don’t think anyone’s ever realized this.
Using AI to Test Learning
Julie Zhuo: It’s very cozy. I love it.
Data Analysis Challenges in the AI Era
Lenny Rachitsky: That’s the idea. I was actually just looking at the stats. So since that first episode, this podcast has done over 20 million downloads. It’s approaching 30 million downloads.
Julie Zhuo: It’s really incredible. I think it is a testament to just your curiosity and how much you really care about the craft of building great products and sharing that with the world. And I know I listen to your podcast and read your newsletter, my team does. We’re constantly sharing things from all of the amazing speakers that you’ve had, so thank you for doing this.
The New Paradigm of Conversational Analysis
Lenny Rachitsky: My pleasure. I really appreciate that. So the reason we are chatting again three years later is you’re re-releasing your incredible book, The Making of a Manager. I’ve got it right here. You’ve sold a bazillion copies. It’s been on every list that I’ve seen. You’re releasing the paperback version, you’re adding some chapters. I guess first of all, just how do you feel on reflecting back on the success of this book?
Julie Zhuo: It honestly went beyond my expectations, so I’m super happy with it. My big motivation to write it was I think largely because I felt if I had to write this thing, I was likely going to become a better manager. And that was actually a huge part of it because thinking about and writing something I’ve been blogging for a long time. And I know that part of my process is when I really sit down to try and put down everything I feel and write letters to myself, it really helps me. And so that was honestly a huge motivation. I hoped that it would go out there and it would sell some books. I was thinking about that maybe for people who grew up in companies like mine, like Facebook, high-scale Silicon Valley, it might resonate. I couldn’t have expected that it would have much wider reach than that and that’s been really awesome.
And just how many people will tell me things like I thought I was the only one who felt this way, but this book made me realize that, hey, these are very normal feelings. And that’s certainly how I felt, just stumbling through and feeling like an imposter for so many years. And so it really is very gratifying to hear that from readers.
The Relationship Between Designers and Data
Lenny Rachitsky: I feel like it’s the modern-day high output management. That’s the book that’s been mentioned most on this podcast and it feels like this is just a modern version. I feel like that book is actually out of date in a lot of ways, so I can see why people are really drawn to it. And this is a great segue to the first area I want to spend some time on, which is it feels like a lot of the skills you learn as a manager translate to being really good with AI and using AI tools really well. And I want to talk through a few trends that I want to get your take on that relate to this general theme. The first is it feels like just everyone is going to become a manager in the near future because of agents being so integrated into our workflows. There’s this agentic society that we’re coming to and it feels like the same skills of being a manager make you really good working with agents. Just thoughts on that and where you think that’s going to go.
Managers Must First Know Themselves
Julie Zhuo: I 100% believe that and agree with that, which is that management is just about, in my mind, having an outcome. So you want to get something done. That’s the thing. You have a north star, you have a vision, and you’re just trying to figure out how to use the resources that you have to get that thing done. And typically when we talk about management in traditional settings, we talk about the resources being people, and getting the right talent, and making sure that you’ve got the assemble the Avengers, so you’ve got the right mix of skills that you need. The second lever is around, okay, what’s the purpose? Does everyone know what they’re supposed to do with their talents? Do we have a goal? Do we have a purpose? And then the third thing is process, which is how should all of these different people and tools come together?
And these are still the fundamentals of working with agentic systems. You still need a goal. You need to be very clear about what the outcome is and you have to understand the strengths of, used to be people, but now it’s basically models. And different models have different strengths, so it’s like they have different personalities. And so you kind have to get to know it, develop an intuition for it so that you can use the right tools for the right purposes. And I mean, we talk about agents, but we also talk about what are the tools that agents have access to? So you still have to make decisions around that and then there’s of course process, which is how you do it. And now I think with better and better models, perhaps the agents get smarter so they can deal with higher and higher levels of figuring out how to do something, but I think it’s still very important for us to be able to provide the right context, provide the right high level instruction so that we get what we want.
So really, it’s the same principles and I absolutely agree with you that more and more of us are going to have to double down on these skills to be able to use these tools very effectively.
Timeless Management Topics Beyond AI
Lenny Rachitsky: So along those lines, I have your book right here. You have this list of a manager’s job is to build a team that works well together, support members in reaching their career goals and create processes to get work done smoothly and efficiently, which is basically exactly what you just said. Interestingly, that middle bullet is the part you don’t have to worry about anymore with agents. You don’t have to worry about their career development and progress in [inaudible 00:11:29]
Actively Designing Your Career Path
Julie Zhuo: That’s true. That’s true, though some people do joke that if we don’t treat our agents nice, what’s going to happen when AGI comes? And maybe it still might benefit us to be kind.
The Power of Feedback
Lenny Rachitsky: I’m one of those people that says thank you to the Waymo when I leave and just thanks ChatGPT when I’m in voice mode. Just like, “Thank you. That was really helpful.” So along these lines, I know there’s a lot of ways to go here, but just in terms of skills that are important to a manager, which do you think are most valuable to develop in working with agents in AI systems? I think about things like clarity, communication. Just what comes to mind when you think about here’s the things you want to double down on as you’re learning to be manager that will also help you be really good at AI tooling and working with agents?
How to Build Feedback Relationships
Julie Zhuo: The first is defining the goal and defining the outcome and being really, really crystal clear on what does success look like. If you ask a company to do this, we’ll know that this is challenging for humans. I think a lot of times when you talk about why is alignment so difficult at a big company, it often comes down to this question, which is different people may have different pictures of what success looks like. And even if I describe in human words, Lenny, I want to build this product and it’s going to be amazing, or this podcast episode, which you asked me, want lots of people to hear it and take away things, that’s very general. How do we get even more specific so that we know without question whether we’ve hit it or not. And this is actually a really, really difficult problem.
It’s a difficult question for us because, again, we tend to think very high level. So figuring out how to boil it down so that an agent can really understand what success and failure looks like is a lot of the game. And I think this also relates to things like, well, that’s why we have to write evals and that’s why they’re so important, because they’re helping us understand what is the objective criteria. And these days I work in data and my company is all about trying to automate data analysis. And the forever question goes the whole point of data and the whole point of metrics and KPIs is we’re trying to put a little bit more of an objective measure or get as crystal clear as possible about what success looks like. And I think it’s really an art more than it is like a science, but that’s the first thing. I think if you’re really unclear about what success looks like, the prompt, you’re probably not going to get the most amazing work. I think that’s true for managing teams and it’s very much true for managing AIs.
Tips for Giving Difficult Feedback
Lenny Rachitsky: Okay, so let me actually flip this on you and talk about another trend that we’re seeing, which is this kind of flattening of orgs, managers being let go. Everyone’s becoming an IC again. I just had the CEO of Airtable on the podcast and his whole shtick was that CEOs have to become ICs again. He’s coding more than he’s ever coded again and his feeling is you have to know what’s possible by being in the weeds in order to figure out what your product should be. I also just saw a stat that Google let go of so many of their middle managers of smaller teams. It’s just like this flattening trend. So do we even need managers, I guess is one question in the future, and then just thoughts on how this will play out?
Win-Win Thinking in Management
Julie Zhuo: So I think the real promise and magic of AI that we’re seeing in the workplace is that it leads us to each individual is far more empowered. So it used to be, okay, I don’t have the skills to do 10 different jobs, so I need to supplement by hiring people to do these jobs. I need someone who’s really good at design, I need someone who’s really good at coding, I need someone who’s really good at data analysis, and then I’ll assemble that team. But now with AI and my companion, it’s like, wait a second, AI allows me to do many of those jobs myself. Now, I’m not going to do them at what’s called the PhD or the highest 1%, 10% level, but if I was at the zero or 10th percentile, it can certainly get me even today very quickly up to the 60th, 70th in terms of what the state of the art is.
And I think that that unlocks so many doors. And so the main thing that I felt so excited about, and this is something I tell my team all the time, is we need to dissolve the boundaries of these traditional roles. So in the past, again, we would have a traditional team, engineers, product manager, designer, researcher, data scientist. And I think now the teams can look more like, well, it’s just two people. And they could be any of these traditional disciplines, but the key thing is they can now use AI to help themselves do a lot of the things that the other folks used to be able to do. So in some ways we can drop all of these different role distinctions and call ourselves builders. I think that’s sort of the most general purpose way of thinking about what we can all be. We can all be builders. We can all be builders and I’d love for us to get to the world where that’s like the title.
What Else Is Worth Sharing
Lenny Rachitsky: That’s funny. That’s the term I’ve been actually using more and more. I used to orient this podcast as a newsletter around product managers and then I started using just product to be a little more broad. And now I’m actually using that term builder and I love that term because it’s exactly what you’re saying. And this is very much a theme that comes up often in these conversations more and more, just the lines are blurring. I’m curious at your company, how does that look? What are you doing differently? What are you seeing on the ground within your company that maybe would be different from a few years ago?
Julie Zhuo: So we have eliminated more roles. For example, we thought we would need a bunch of product managers. It’s turned out that actually if you don’t have a product manager, I know this might be going against a little bit of the ethos of where Lenny started, but I find that sometimes when you have a designer or a product manager, and let’s say I’m an engineer, then when I have a problem, like I need to figure out the product definition, my default will be, well, I’ve got these people and that’s kind of their job description, so I’m just going to delegate that to them. And I think that in doing so, again, we want to be polite, we want to respect everyone’s lanes. I think that’s a missed opportunity for that, if I’m the engineer, to be like, “Wait a second, I should probably focus a lot, too.”I need to understand and have an opinion about what to build or what the user experience is.
And so we found that if we actually make teams smaller and we even in the past, pre-AI, just have fewer of these, it allowed everyone to be like, “Wait, we don’t have product manager on the team, so communication’s up to me. Figuring out how we get greatest value to users is something that is now strictly in my charter. And so that’s why I’m such a big fan of we can make teams smaller and we can eliminate these lines. Sure. Again, I’m not trying to say everyone has to do everything. We still can respect the fact that you might be much better at this particular skill than me, but it’s less about the role and it’s more about the specific context that we’re in.
And I find that whenever you have teams and you empower them to be able to take more action on their specific context rather than having these higher level of rules or policies or this is how it’s supposed to be, then you get better work. You get faster work and you get happier employees because people feel like they actually can have the power to create the thing that they want.
How to Execute What You Disagree With
Lenny Rachitsky: That’s really interesting, just that constraint of not having a PM makes the engineer realize they’re not going to wait for someone else to do it. They have to figure it out. The obvious trick there is they have to be good at this. It’s a very different job from engineering to be really good at articulating here’s the problem we’re going to solve, here’s why it’s important that we’re solving, here’s how we’re going to prioritize everything we’re thinking about, here’s how we get alignment. Is there something you do differently when you’re hiring these engineers, knowing you’re going to probably not hire PM? And just that feels really hard to hire for someone that’s really good at all these things.
Julie Zhuo: It is true and I’m not trying to say again that everyone needs to be good at everything. I don’t think that’s very realistic. I do think, for example, if we were going to create a team and we’re going to have a couple engineers and none of them are very good at thinking through product requirements or what the user angle is, we probably do need to supplement the team with somebody with that skill set. And that might be a designer, or that might be another engineer who’s really good at that, or that might be a traditional product manager, or even sometimes a data analyst who’s really good at it. So that skill is still important and the team still needs to have that skill, otherwise it’s probably not going to produce the best outcome. But I like to think of it as what are the skills that are needed for this and can we now find a couple people?
But it doesn’t mean we just automatically go to that script of need a PM, need a designer, need three engineers, need that. Another example for us is even thinking about front end, back end engineering. And it used to be like some people are front engineers. So if you have a project and it’s got some front end, some back end, the shortcut is like I need one of these and one of these and that’s how it’s going to go. But if you say, look, you’re an engineer, you’re a builder, this has a little bit of front end, but you know what? You can probably figure that out. Use AI to help you figure it out. Get obviously someone who’s a specialist to review the code or to give you some high level guidance on things, but just do it. And ever since we started to implement that as well, we see again a little bit of you have to invest a little bit in the beginning. So people are not as comfortable.
They have to learn, so initially things take longer, takes a little bit of extra time, versus if you did slot in a front-end specialist and this is a front end project. It probably would’ve gone a little bit shorter, but in the long run that investment really pays off because now you have a lot more people who are, again, a little more well rounded and can take on many more pieces just on their own. And then in specific scenarios this is super front and heavy. Sure, let’s still bring in somebody who is more specialized in that particular skill.
The AI Corner
Lenny Rachitsky: I love that you’ve had the experience of working at a mega large company at Meta and now you’re building your own startup that’s small and in the middle of this trend of just staying very small and staying really lean and just everyone doing more things. It’s so cool that you’re experiencing that. So a couple of questions there, just which functions are you seeing most accelerated with all these AI tools? Is it engineering? Is it something else? And then are there tools that have been most helpful to you? Just AI tools for folks who’d be like, “I should check it out.” I’m guessing Cursor, but curious if there’s anything else.
The Contrarian Corner
Julie Zhuo: Yeah, certainly engineering is one that, I mean, most of our company is engineers, so that’s the one that we focused on a bunch. I certainly do see more people also prototyping things. We have two designers, but we also see engineers. We have a team that’s called product science, which is this interesting blend of you can think about it as like a forward deployed person who has a lot of data analysis background and is kind of playing a customer success role and also kind of playing a product role. And you see them starting to take on building more prototypes or getting into some of the engineering. And so it’s really lovely to see that blend of everyone can do a little bit of everything else and we’re all encouraging each other. The other thing that recently we’ve also been trying to do a lot more is just obviously we say, “Hey, engineer, now you can do analysis.”
And their first thing is like, “Oh, I don’t really know analysis.” This is where ChatGPT comes in. And it’s like traditionally we would say, “Well, I have to learn that from a human. I have to ask this person and now I’m going to take a bunch of their time because I want them to explain everything to me.” And in fact, I think these days ChatGPT or these other AI tools are better teachers. I find that we tend to maybe not use them quite as much just for the purposes of accelerating our education or even going through something. Sometimes what I’ll do is I’ll find a curriculum online. And if you take a course, it’ll be like this 12 week curriculum and I’ll just feed it into ChatGPT and I’ll say, “Help me customize a program for me using the ways that I like to learn.”
I am a person who really needs examples. I need a lot of explain like I’m five. Give me an analogy. And I know some other people on my team are like, “These examples don’t make any sense.” We’re different types of learners and so the idea of a tool that personalizes learning for each of us really helps us, I think, accelerate and just learn these skills much faster than before. So yes, the tools are great. We can use Cursor, it helps us, it autocompletes, it writes a bunch of things, but the acceleration of learning I think is another maybe underutilized tool in all of our arsenals just because I know whenever I talk to people, we forget. We don’t think that, wait, yes, we could be doing that and just sitting down and probably in 30 minutes or an hour learn so much faster than what we used to be able to do before.
Highly Recommended Books
Lenny Rachitsky: That’s such an interesting point. There’s these tools that are in the just in time, helping you move faster, but you also need to learn how to do something, some foundational lessons. What’s an area that your team did that? What did they work on learning?
Julie Zhuo: So I’ll give you an example. I was just talking to an engineer this morning and he’s written a bunch of these algorithms. So one of the things our company does is we’re trying to automate data analysis, so one of the things we have to do is obviously understand the best practices. If there’s a type of question …
… To do is obviously understand the best practices. If there’s a type of question, “Hey, what features are really the ones that people pay for?” We need to kind of figure out what is the right analysis to do. And so the engineer was saying to me, “Julie, I feel like I really understand the how. I know the algorithms, I know we do root cause analysis, how we do that. But what I don’t really understand is why or when this would be most useful. In what context in a company would this company come up?” Because he’s an engineer, he hasn’t done that job of being a PM or an executive that asks these types of questions. And that was like the perfect thing where yeah, traditionally you might’ve asked someone, but this is more general purpose. There’s so much resources in the world on the internet about it. This is like the perfect type of question where if you just talk to ChatGPT, it’s probably going to give you a much better answer and allow you to go deeper.
And a secondary thing we’ve been learning too is this idea of, almost like as a… Using ChatGPT it’s for to test your learning. So explains a bunch of things. And so what I often like to do is like, “Okay, I read this, so this…” I try to explain back what I heard. “So does this mean… Is that right way to think about it, that this is kind of like this analogy?” And ChatGPT will critique me. “Yes, that is right,” or “No, you didn’t quite get that right. In fact…” And it always tries to say it nicely. This is a funny part. It’ll be like, “That’s close, and then eventually it’s like, “You were completely wrong.” Just in the style. But it helps so much because it’s interactive and so we can really test whether we really understand the concept by trying to retell it back in our own way.
Recent Favorite Shows and Movies
Lenny Rachitsky: It’s incredible just how many ways all this AI breakthrough is helping us advance and do more and learn more and become better. I know there’s some downsides, but this is incredible. So many ways of getting better and faster. I want to spend a little more time on this data analysis stuff. So again, you have such an interesting trajectory from working at a big company to starting your own small company. From being head of design to now being obsessed with data and analytics. So let me spend a little time there. What do AI companies that have kind of figured out how to use AI for data analysis and data work, doing differently, what are people missing and sleeping on in terms of getting better at working with data? And let me just add this point. It feels like we’re almost working through , here’s all the blockers to a team moving forward. There’s like waiting for the PM to write the PRD and then there’s waiting for the data scientists to give you answers analysis. So this is another really cool unblock that every team member will have.
Julie Zhuo: So your first question was how are a bunch of AI companies using data? So the funny thing, my funny answer to this is, I don’t actually think a lot of the fast-growing companies are using data well at this point. And the main reason why is because traditionally things just didn’t grow that fast. And so if you got to a hundred million users, your company has probably been around for a while, and if your company has been around for a while, you’ve had time to set up things like logging and you’ve hired a growth team at that point and you’ve hired a data team and they’ve done a bunch of work to log an instrument and then transform the data. And we’ve talked about what is the observability for our business. And you just usually had years to build and develop that, because of the rate of growth.
And so today we see companies that are growing insane and there’s still about 10 people or two people or however many people, but they’ve got hundreds of millions in ARR and hundreds of millions of users. And you know what? They don’t actually have all of that infrastructure, that logging, to be able to truly do data analysis. So I would say that these companies are totally getting by on just good instincts and good vibes and we see that. You don’t really need data analysis to sometimes make something that works. But I think what data helps us do is in my mind it sort of is helping us reflect back what is really reality. And so of course if ARR growing, awesome, great, keep doing what you’re doing. But what always happens is eventually things stop growing. Growth does not happen forever. And usually when growth stops, everyone has this question of, “What’s going on? Why did it happen?”.
And then you start to be able to see the power of, if you’ve instrumented everything very well and you have a very good observability model for your business, it’s much easier to start to get into the root cause, it’s easier to even predict whether growth will slow down at a certain point, it’s easier to catch these trends earlier. If you don’t have good observability over how your business runs and what the company’s key levers are, then you will be scrambling, and at that point, that’s usually when people start investing a ton in data. So I wouldn’t say that a lot of these hot companies are quite there yet, but what I also think is a trend is that every time there’s a new technological shift, we actually have to change the way that we think about… Analysis has to answer the questions that we have, and if technology changes or context changes, we need new methodologies of analysis.
So for example, when mobile came to the forefront, looking at sessions or sessions per day or time spent on mobile, or length of sessions became something that was important for us to understand, are people getting value in this new medium? I think that’s the same with what we have today. Conversational analytics is totally different. Used to be, let’s say in the Google world, I knew you were interested in shopping if you click the shopping tab, I know you’re interested in maps if you click the maps tab, we can measure clicks. Today it’s just all conversation, and so it’s actually harder for us to tease apart what is the user intent.
If I worked on any of these LLM, I would say one of probably the biggest questions is, hey, what use cases are growing or what use cases are shrinking? And that’s much harder to tell today because it’s not just clicks on tabs or pages. It’s like we have to probably use an LLM or a machine learning model to bucket user intent. We probably have to ask questions like, is the flow going really well in conversations? Like, if I just ask one question and I don’t go back and forth, did the user get value? It’s always trying to get back to, we’re trying to figure out if this was a good experience, but now it’s like we need to actually invent new methodologies to help us analyze that.
Recent Favorite Products
Lenny Rachitsky: Yeah, I think the question is always like with conversation, do you want it to be a long conversation, do you want it to be a short conversation? What’s the right answer, what’s better?
AI Parenting Assistants
Julie Zhuo: Yes.
My Personal Life Motto
Lenny Rachitsky: I had a ChatGPT on the podcast, Nick Turley, and turns out one of the ways they found the most common use cases early on was watching TikTok comments and things going viral on TikTok after they launched. How about that?
Julie Zhuo: Yep, yep
Educating Children in the AI Era
Lenny Rachitsky: Okay, so I want to come back to this really interesting, unusual path that you took from being a head of design at Facebook, you’re an inspiration to so many designers, now you spend your time on a data startup obsessed with data. I don’t know, classically designers aren’t the biggest fans of experiments and data and making decisions based on data. When you look at designers and you hear designers kind of push back on like, “No, we don’t want to be super data driven, we know better than… We have a sense of what’s beautiful and great and intuition,” all these things, what do you think designers are missing when they feel that and say that when they’re afraid of writing experiments and data and kind of want to push that out?
Julie Zhuo: There’s one phrase that my co-founder and I would always discuss with amongst ourselves very early on in which we shared with a lot of the companies that we work with, which is, what you really want is you want to diagnose with data and treat with design. So data is not a tool that’s going to tell you what you should build or what the solution is or how we’re going to cure the fact that you don’t have really great retention. It’s just not. But it can tell you if you have a problem and where that problem or opportunity might be. But you still need to go back and undergo a very creative process to figure out what’s the best way to solve that. So that’s the first thing I would say, is this framework of, data helps you figure out what’s actually happening, what do people like, what are they engaging with, what what not.
It just gives you a story that better reflects reality. Because again, we all have stories. We’re like, “Oh, my company’s amazing, people love us,” blah, blah. That’s the story I want to believe, but reality may be a different picture. And so what data is trying to do is capture reality. And by the way, I don’t think of data just as it’s an AB test and it’s quantitative things we can measure. To me data is also, well, what did people put onto TikTok and which things went viral, and what are they saying in the Twitter verse or X verse I guess is what it’s called now.
And if you do a customer interview, that’s still all data, it’s just that that is a little harder to distill and quantify. Although now with AI, we have better tools for synthesizing. So that’s all data in my mind and it’s just all trying to help us understand what is really happening, what is the phenomenon that’s happening in reality and how do we understand it? You still have to go and invent and create and dream, and there’s no formula and there’s no science that will tell you exactly how you’re going to make a hit. You can experiment, which allows you to try more things maybe and more rigorously understand what that does in the short term. It’s all very contextualized. A-B tests don’t tell you what will happen in the very long run, and again, it’s all still data, you still have to synthesize and figure out what to do.
So that’s the thing, I’ll say. Diagnose with data and treat with design. The second thing I will usually tell designers about, is I find that sometimes, and maybe it’s the, let’s call it the false precision of numbers that we kind of fall into, right? Because it’s like, okay, we got these numbers and the numbers go up. It’s like no, the fact that you still have to choose which things you look at, is an art, not a science. And your interpretation of if the number went up 5%, is that good, is that not good, is also an interpretation and is an art, not a science. It’s just that sometimes I think we can give ourselves this feeling, and I get it, sometimes there’s this instinct to want to control things and we want everything to be buttoned up, and we want to know that if we did ABC, everything will be great, our career’s going to be awesome, our product’s going to rocket ship.
And I think designers are rightly often pushing back and saying, “No, the reality is this stuff is ambiguous and there’s uncertainty and we can never know for sure.” And I think all that is quite true. So the other thing I would say that I really support is you just actually can’t make a really great product by thinking you can A-B test your way into it. So I fundamentally believe that, but I don’t think we should throw the baby out with the bathwater. I think there’s actually… You know? It’s not either or, it’s not like data or design.
It’s like these are just tools for us to use, and I would say every amazing designer that I’ve ever met is absolutely obsessed with trying to get a better understanding of reality. They want to know what users really think, they want to know what they’re really doing. If they could read every user’s minds, that’s the thing we would all want as a designer is like, if I could just know what everyone is thinking, feeling every time they used it, my life would be a lot easier, because then I would be able to build better and better things. And so that’s what it’s trying to help us do. It isn’t perfect, no metric is going to tell you whatever we hope that it can in terms of the true certainty and precision, but it doesn’t mean we can’t use it to better our product development.
Final Closing Remarks
Lenny Rachitsky: I was going to say exactly what you just said, which is every great designer that I’ve worked with was obsessed with data in the most leaning into the data, versus designers that are just like, “Nah, I think I’m good, I have a sense of what’s right, and why would we let that tell us what to do?” And to your point, it’s not going to tell you what to do, it’ll tell you where opportunities arise. Let me take us back to the management chat and maybe just let me ask something broad. What do you feel is the biggest change in the role and day-to-day work and life of a manager these days with the rise of AI?
Julie Zhuo: I think that managing change. It’s always been manager’s job to manage change, and there’s always the chaos of what’s going on. I just think the rate of change is accelerating, and we’ve seen that over the last couple of decades. And so I find that there’s just a great deal more uncertainty that people have about things, like where is AI going to be in two years from now? I don’t know. Who really knows? And so are we going to have AGI in five years? That kind of changes a lot about the landscape. Not to mention, I think there’s quite a lot of fear that many organizations are feeling. It’s like if my career has always been in design and now these tools are getting better and better at what I’m doing, then holy shit, what happens to my career and my future? And do I need to pivot? Do I need to learn different things?
And so it’s this change, it’s this feeling of uncertainty. And I think a lot of times managers have to deal with that in addition to what you were saying before, which is they also have to learn these new skills, which is managing AI and managing these more powerful tools in their arsenal of trying to get things done. So that is very different, I think, than maybe 10, 20, 30 years ago. And so I think that the skills that become more important are obviously communication, feedback, compassion, but just being able to work with humans and to have them understand that yes, we are in a state of change. I think every leader has to do this now, every startup founder that I know, every CEO, is how do you land this message that things are changing and we need to be very open to change?
If we go and stick to our old ways, we’re probably going to get left behind, our product’s going to get left behind, even our way of doing things is going to be left behind. So we need to change. We need to change our product and we need to change the way that we work, as we all talked about in terms of smaller teams, more nimble, blah, blah, blah. But at the same time it’s like, how do we do that in a way that doesn’t just freak everyone out? And it’s like, “Ah, it’s chaos. Everything’s changing.”.
So I think about this metaphor a lot, of the willow tree, which is the willow tree is a very sturdy tree. It can survive a lot of storms, disasters, et cetera, but it’s also very flexible. The branches are very, very flexible, and that’s in some ways what allows it to be very sturdy. So I think today, management is really about this idea of be sturdy while being flexible, and that is a very hard thing to pull off, but I think that’s at least when I even go into… I’m like, “Be like the willow tree, Julie. Just imagine the willow tree and try and channel that as the kind of feeling of what it is that we’re trying to do together.”
Lenny Rachitsky: This reminds me of a couple things from other guests. I had Marc Benioff on the podcast and I asked him, “Just how do you deal with all this change? It’s like agents now, it was, I don’t know, there’s AGI coming as you said, just like, “How do you survive through this?” And his advice is just, he’s like, “I’m always just like, ‘Good. This is great. This is what we want. This is exciting. We have so much opportunity, it’s just not boring. We can always reinvent.’” And he’s always embracing with “This is good.” And just I’ll never forget the way he responded to that.
Julie Zhuo: I think if you don’t think it’s good, it’s kind of a painful way to live. It’ll be very, very difficult over these next. So I do think that all things be equal, lean into it. If you can wake up every day and see it as opportunity and excitement rather than fear, again, they’re all flip sides of the same coin, but I think if we can lean more into what could it be, while recognizing that the other side does exist and it’s still there. And I think if managers who try to pretend like it isn’t there, it’s all good, no one’s upset, et cetera, there’s something also missing about just addressing and being able to be like, yeah, it’s hard. Change is hard. We’re probably going to get upset. We’re going to have some chaos. This is going to happen, but we will work through it because we’re going to be flexible and we’re going to be able to put our eyes on the big picture of what is possible, which is exciting.
Lenny Rachitsky: There’s another quote that and came up as you were talking. I forget who it was exactly, maybe Kevin Wheal, maybe Mike Krieger. They said that this is the most normal things will be, ever. Like, it will only get weirder. And I think giving people that sense of like, okay, just enjoy this normal, because this is going to be only weirder, is we’ll at least give people an expectation, real expectations where things might be going.
Julie Zhuo: Yes, yes.
Lenny Rachitsky: What a time to be alive.
Julie Zhuo: What a time.
Lenny Rachitsky: Okay, let me zoom out even further and chat about… I want to ask you just outside of AI, management in many ways is unchanged. It’s still a lot of the same work, managing people, helping them be successful, producing great work. What are just some of the, I’d say most timeless, most important lessons that you think managers, especially new managers still don’t totally understand, need to hear more? What are just some that come to mind? And then we’ll see where this goes.
Julie Zhuo: The first thing that comes to mind is the importance of managing yourself and understanding yourself. This was chapter five of my book. It’s called Managing Yourself. In fact, when I wrote it, I kind of wanted it to be chapter one, and then my publisher was like, “Well, maybe you should get into some of the tactical…” People don’t necessarily think managing other people or manage a team starts with them, but I really do fundamentally believe this, because I think all of us, of course, like any human being, we have things that we’re strong at, we have things that we’re weak at. And I am a very big believer that every strength is its own weakness, and every weakness is a strength.
There’s no such thing as you’re going to somehow get every dimension to be 100%. In fact, I think one of the most interesting concepts or frameworks for myself, and also even, this is also kind of like a data framework concept, is this concept of dimensionality. So what dimensionality means is you’re a human being, but we can kind of look at you in infinite dimensions. There is, for example, how good is Lenny at throwing an ax? That’s one dimension
Lenny Rachitsky: Pretty good.
Julie Zhuo: How good is Lenny at being a podcast moderator? Fantastic.
Lenny Rachitsky: So-so. Okay, thank you.
Julie Zhuo: How good is Lenny at doing a zero to one type of project in the AI space? Right? So again, just can think about these as infinite dimensions. And the reality is each of our profiles is very unique, it’s like a fingerprint. So for you it’s like these are all these areas that you’re really great at, much better, like in the top 1%. And then there’s some areas where in the top 10%, then there’s some areas where you’re kind of average, and then there’s some dimensions in which you’re worse than average compared to other people. And that’s just true for all of us. And what I like about that is therefore if you take that as the model, you realize that none of these dimensions are you entirely. So I can make a comment like “Lenny, your ax throwing really could use some improvement.”
And ideally you’re not like, “Julie is saying I’m a bad person, my identity is at risk,” right? Because it’s just one dimension of who you are. But what happens sometimes is that we can get very attached to certain dimensions because we start to think that that’s who we are. And I think managers can do that, and clearly individuals on their teams. And when that happens, it starts to get very difficult to have, I think more objective conversations about, okay, what can you get better at? What can get worse at? And so I say all this because I think this framework for me at least, and many people that I’ve talked to, has helped them realize that somebody can give you feedback or you can be maybe not great at certain dimensions, you can have room to improve, and that’s not who you are because you are all of these infinite dimensions in one, and none of them is representative of your true worth as an individual.
I’m a big believer that we are all beautiful and worthy, and sure we have all of these skills and we want to improve those skills, but it does not speak to whether we are worthy or not by saying whether we are strong or weak in these skills. And so I think if you can take that and really internalize that, then you can look at yourself a little bit more objectively as a manager, and you can realize that there are areas where you’re going to be really strong, there are areas where you have biases, and often they are one and the same. So I’ll give an example. People have often told me, I would get this in my performance reviews from managers in the past, like, “Hey Julie, you’re really thoughtful. So when you think about something, you have a way to think about it, you’ve clearly thought about it in depth and you’ve got these frameworks and all this. That’s a great thing.”. And then on the flip side, I’ll get feedback like, “Well, Julie, you don’t really say a lot in a dynamic discussion. You’re kind of quiet and you don’t really think that quickly on your feet.” And what you realize is these are kind of, again two… Because I don’t do that and I’m not just off the cuff, that’s what allows me to oftentimes be very, very thoughtful, or at least, okay, when I was younger, it’s very clear that that particular weakness also very much is speaking to a particular strength, which is I am the kind of person that doesn’t always have a snap judgment. I have to really think about it and internalize it and sometimes get to how I feel, and then I can share it and present it in the world.
And so just knowing that about me is supremely helpful. Now doesn’t mean of course that I can never get better at this thing, but what I often think about is mastery is where we realize that both of these we can get better at, and what we want to do is just figure out in the context, what makes sense to be. So I got this feedback and I’m like, “Cool, one of the things I need to work on, is figuring out how to be more open in person, how to speak a little bit more clearly in person, maybe say things like, ‘I don’t know exactly how I feel about it yet, but this is what I’m thinking right now,’” if there’s still clear tactics that will allow me to be a more effective team member and to do a better job in the context of what I’m trying to do with my team.
So I’ve tried to build those skills, but the meta skill is now being able to step back and say, okay, in certain context it is really important that we move fast and we are decisive and we just do something. And even if it’s not perfect, we just kind of have to do it. And if I struggle with that, I should realize that that’s an area to improve upon. But there are other contexts in which the right thing to do is actually to take a step back and be very thoughtful and to not rush into decisions.
And that’s so what I want to get to is not like let’s reject this strength or this weakness, but just know that that’s where we come from, that naturally, we might be wired in a particular way. Our growth often looks like getting better at doing the opposite, but not rejecting again the thing that we’re good at, but rather over time getting to this balance where we can read the context and the situation and know, “Should I lean a little more thoughtful or is this a time where I need to try and be a little more decisive and just share what’s on my mind right now?”
Lenny Rachitsky: I love this advice that things that we are incredible at and have a downside, and oftentimes the feedback we’re getting is something we’re not great at, there’s a good version of that that people appreciate. And I was going to ask you, and I think you answered most of this, but just when you got this feedback of, “Hey Julie, you’re not speaking enough in these meetings, you’re not contributing quickly enough,” it sounds like, so one option is just like, “Okay, cool, that’s me, that’s how I am, and I’m just going to solve the problem this other way and then just not going to change anything.” What I heard you say is, find opportunities where you want to actually change that behavior, even though it doesn’t come naturally in specific situations where things are moving fast. I guess just how far do you recommend people push themselves in things they’re not great at, versus leaning further into their strength, let’s say?
Julie Zhuo: Oo, I think that’s a really great question. So the way I think about it is it’s very dependent on what is your goal. So for example, let’s say that you are…
… on what is your goal? So for example, let’s say that you are… Let’s even take, for example, ICs versus managers. I think often about the pathway of an IC, an individual contributor, as wanting to deepen a craft. You love this thing and you just want to get better and better and better at this very specific skill or this craft, right? So think about in our dimension, infinite… It’s like you pick a couple dimensions, “I just want these to be… I want to be the top 0.01%,” and that’s kind of the pathway of extending it as an IC. Now, if that’s your high level goal and you’re like, “I want to be able…” Let’s say your high level goal is, “I want to be able to do this 10 hours a day because I love it and I want to be able to support myself doing it, meaning I get paid and I have a great job, and I want to have a bunch of impact in the world by doing this thing.”
So again, you still have goals. Then you have to see, okay, “Does my strategy of just deepening these things, is there a pathway to reach my goals according to that?”
And if there is, awesome. Then if someone’s like, “Hey, do you want to be a manager?”
You’re like, “Nope, don’t need to because these are my goals and this pathway actually allows me to do that.”
But if somehow you get to a point where the skill you really want to perfect is not something that may be commercially viable in the world, that’s going to somehow allow you to buy the big mansion that you want to buy to support your family, then I think you have to ask yourself, “Okay, so if I just do this, it’s not going to cut it. I might actually need to learn some of these other skills in order to be able to fulfill the job that is going to be valuable enough that people are going to pay me a bunch of money at this certain level so that I can afford my mansion.”
So I just think it has to go back to, what are your goals? And there are cases in which yes, it’ll support your goal to do this and to deepen your craft. And there are cases in which it won’t. And I think it’s important, it’s a very individual question for each person. But what I often think suffering is, is when these things are not aligned. So what you want is the giant mansion and all of that, but you’re like, “But I also just want to spend on my time perfecting my egg omelet.”
And then, you’re just in this tension place, and it’s very hard to feel satisfied and fulfilled because you’re a little bit like, “Oh, why doesn’t the world value my deep egg omelet skills?”
You can [inaudible 00:52:27] egg omelet, you should maybe not do this thing. Or if you want this thing, you may actually need to be better at just egg omelets. Perhaps you need to expand your repertoire of cuisines, and go and build a Michelin star restaurant or something.
Lenny Rachitsky: This is really good advice. It’s not just definitely always work on your weaknesses or don’t worry about them, it’s if you need to do this thing to achieve this goal that you have, make sure you understand what your goal is. And then is this thing a thing you need to work on? For example, [inaudible 00:52:52] become a VP, you probably need to be really good in big important meetings, and being on the spot, and just not waiting until everything’s over and then sharing an email of all your thoughts.
Julie Zhuo: That’s right.
Lenny Rachitsky: Yeah. For me, I actually went through a period where I was like, “I do not want to get promoted. I’m so happy in this very specific role, just leave me alone.” And that path is very different from the skills I need to build to be a manager. And then things changed and then, okay, now these are the things I need to work on.
Julie Zhuo: Yeah. I love that you knew that about yourself, because I think it’s so easy for a young person to go into their career and everyone is telling them, maybe their whole family has been telling them, “You need to level up, you need to get paid more. You need to get that manager title. You need to get a VP.”
And at a certain point, I think sometimes people opt into this without knowing what they’re actually signing up for. What are the trade-offs? And is that really what you want to do? Does that really align with your passions? And of course, sometimes we have to… Again, it’s a compromise for us, but we get to design. We get to design what are goals and what’s the right pathway. And I go back to, usually when people are unhappy, it’s because these things are a little bit out of sync. They want this big thing, but they’re not actually excited about what it takes to do that thing, and therefore it’s just going to be a mismatch.
Lenny Rachitsky: And along those lines it sounds like, oh, sure, I can design my life and design my role. But what I find is if you at least first of all know what you’d love and ideally do, and then at least mention that to your manager, it often is a lot more possible than you think.
Julie Zhuo: A hundred percent. I think it’s so important to be… We often also have this mental model like, “Oh, our managers are our judge, and they’re going to judge me on whether or not I did well, I should get a promotion, I should be fired.”
So there’s this sometimes fear that people have, but I think in the very best relationships, the manager is like a guide. It’s like, look, the manager has a job, and if you understand your manager’s job, which is how to get better outcomes from the team, and also you understand what exactly would your manager consider success for the team, it also makes it easier for you to then be like, “Oh, well if I do this project, then that clearly seems like it’s a very direct path to creating value for the team. And that also is a kind of project that suits my skills. It’s something I’m excited about.” You should suggest that to your manager.
But the other is true, right? So you would know that if you actually asked your manager, “What is your job and what do you consider success to be, and what is your greatest hopes and dreams?”
And then you might be able to help your own career and yourself because you would know that context. And conversely, if you say, “Hey, manager, these are my hopes and dreams. This is what I think I’m good at. I really want to get better at this skill. I really want to get that VP promotion, but I don’t know what it entails. Can you tell me, what does it take?”
That’s a really wonderful conversation as well because then you’ll get all of that context, and then you can actually decide whether you want to do it or not. And if you want to, then ask your manager for help, “Okay, if you see opportunities that are going to help me become a better presenter or increase my communication, please tell me.” Even better, “If you have feedback for me about communication, I want to hear it, because that’s what’s going to help me grow in this particular skill.”
And so, it becomes this collaborative relationship much more so than this almost adversarial, like I’m trying to get you to give me a promotion, and you’re trying to get me to work harder. That is not a very good vibe.
Lenny Rachitsky: It reminds me of a guest post by Ethan Evans that I’ll link to that has a really good framework for how to actually do exactly what you’re talking about called, The Magic Loop, where it’s kind of a framework for figuring out what to work on and how to help your manager see you’re capable of stuff and earn that trust.
Beyond analytics events, their data warehouse enables you to sync data from your Postgres database, Stripe, HubSpot, S3, and many more sources. Finally, their new AI product analyst, Max AI helps you get further faster. Get help building complex queries and setting up your account with an expert who’s always standing by. Sign up today for free at PostHog.com/Lenny and make sure to tell them Lenny sent you. That’s PostHog.com/Lenny.
So along the lines of timeless manager, especially new manager advice, you’ve shared a bunch. Is there anything else that you think is really important, really interesting, valuable?
Julie Zhuo: Feedback is one of the other topics that I am super, duper passionate about. And my general impression for both myself, everyone I’ve worked with, is that we don’t value feedback enough or we don’t think about it enough. Again, companies have these performance cycles, and so we’re all like, yes, every six months we’re going to go and do these reviews. That’s when I’ll get feedback. But feedback really, in my mind, ideally, should be a daily practice. Because the thing that matters for us in the long run as a team is how quickly are we getting better? So a team that just gets 1% better every week compared to a team that gets 1% better a month, even if they start off at a much lower baseline, is going to outperform in a very short amount of time the team that doesn’t get better.
And so, what is the best tool for us to get better? It is feedback. And what I think about in feedback, it’s very similar to what we said earlier about data metrics. It’s essentially trying to put your hypotheses and test them against reality. So as an example, maybe I have this perception right now that I am a positive and engaging speaker. So, I have this sense that I’m smiling and I’m very engaging, and I’m telling great stories, but is that really true? I don’t know. The reality is that I’m often biased, and we know these psychological effects where sometimes the Dunning-Kruger effect, people think they’re way more expert at something than they actually are. You ask people, “Hey, are you a better than average driver?”
And it’s like 70 or 80% of people, “Yes, I’m better than average.”
How could that possibly be? We have biases. And imposter syndrome is a bias on the other side, it’s like me feeling, “Oh, I suck. I don’t actually belong here.” Whereas, that also is a bias. It may not actually be true. In fact, I might very well be here and other people value my contribution.
So we are just wildly out of sync a lot of times in our perceptions of ourselves, our strengths, our weaknesses, what’s going on. And the way that we’re going to understand and truly get better is we need other people to reflect back what is actually their truth. And the way I think about it is like, I’m going to ask you for feedback after this podcast episode and you’re going to tell me something. And what you’re going to do is you’re going to give me a gift. Because it’ll be a gift of reflecting something back of what you see that I can’t see. Just like if I have a leaf in the back of my head, I can’t see that. And so if you’re telling me, “Hey, Julie, you have a leaf.”
“Oh, wow, thank you.” Okay, maybe I can get rid of the leaf or whatnot. But that is what feedback is. It is essentially reflection back. It helps us calibrate to reality, and it allows me to get this information about whether or not I’m moving in the direction of my goals.
Lenny Rachitsky: I love that. I completely agree. The challenge for most people, as you know, is giving feedback that people receive and don’t feel defensive about, and then receiving feedback and not being like, “Oh, no, they don’t know. They don’t know anything. How dare they say this about me?”
Could you give us maybe a tip or two for delivering feedback well and for receiving feedback well? And maybe even just seeking, how do you get more feedback? This all makes a lot of sense. Most of the time people don’t get any feedback.
Julie Zhuo: The best way… The first tip on getting feedback or delivering hard feedback is first go and actually establish that our relationship is one in which we value each other’s contribution, we want to help each other grow, and therefore we’re going to be the kind of people that want to give feedback to each other every week. So when you first start working with someone, don’t wait until something bad has happened [inaudible 01:01:42] given feedback, because that’s already a pressurized situation. Start by saying, “Hey, really excited to work with you. I feel like our best collaboration is I want you to help me get better. I think I’m good at this stuff. I’m not so good at this stuff. What about you? Okay, you think you’re good at this stuff? How about we just work together and we just help each other get better at these things? And the way we’re going to do that is, all feedback is open. I want you to tell me everything. Ideally, you’re going to then say, ‘Yeah, I want you to tell me everything.’” And we’ve already established that.
Lenny Rachitsky: And this is colleagues or manager or all colleagues?
Julie Zhuo: It’s like everyone. It’s like people you’re dating, it’s like your children. It can be with everyone, just establishing what kind of relationship do we want to have? I think most people want to opt into a relationship where you can be close, you can be tight with one another. You can say things to one another and not have to hide behind… I think most people will opt into it, and if you opt into it, everything gets easier down the road. So the first thing is get everyone to opt in that this is the kind of relationship that we want to have.
Lenny Rachitsky: One trickle throughout that I’ve heard that worked really well along these same lines is asking people, “Do you prefer feedback in the moment or do you prefer it kind of every month or every week or something like that?”
And everyone’s like, “No, no in the moment and just tell me as soon as something happens.”
And then that gives you that freedom to just, ” Okay, yeah, let me give you feedback here.”
Julie Zhuo: So if you get people to opt in, “Yes, I want us to have a great relationship. I want us to help each other get better. I want feedback.” That’s 60% of the hard part of delivering difficult feedback later on.
Then the second tactic I will say is that when you actually give the feedback, it helps a lot. First, you have to check, “Am I actually giving this feedback because it’s in the spirit of trying to help one another?” And if the answer is yes, then we’ve moved from 60% to 80%, it’s going to go well.
But what can often happen is something happens. You do something, it triggers me, because I don’t know, I had a bad experience about that type of thing before. And so, I’m kind of feeling mad and I want to be right. If my real rationale for why I want to give you feedback is I want to validate myself, I want to be right, I want to tell you you’re wrong, I want to punish you, it’s not going to go well. It’s just already there. There’s no way you can deliver it, unless you’re a tremendous actor. It’s just not going to go well. So you have to first check your intention.
But if you’ve done that, you’re like, “No, no, no. I thought about it. I’m calm now. I’m not seeing red. I really think that Lenny is just not aware that when he says this, it makes me and other people feel left out,” or whatever it is, right? Then I need to be able to give it to you.
And so usually then if you’re like, “Okay, now I might be nervous because I don’t want to offend you. I really value our relationship. How am I going to tell you. I don’t want you to get defensive?”
Then the third tactic is, just say that out loud. If I sit down with you and I say, “Lenny, I’m so nervous right now. I want to give you some feedback and I’m really worried that it’s going to impact our relationship, and I so value our relationship and I don’t want that to happen. But I also feel like it’s just going to help you to hear it if you can.”
That does so much of the work of… It’s humanizing. You’re going to realize that I’m going out on the limb, I’m being really vulnerable, and likely you’re going to hear that so much more than if I just find a way to drop it, just lobby it over because it’s so difficult. Just actually lean into the fact that it is difficult and expose that because that builds a lot of human connection.
Lenny Rachitsky: This is amazing advice. Very tactical. Okay, is there anything else? So we’ve talked about a bunch of timeless pieces of manager wisdom, things that people need to hear, especially as new managers. Is there anything else that you think is really important that you think people are just not fully grokking for being great managers?
Julie Zhuo: I think the idea of win-win, I think about that all the time in my mind. And I go back to it, because I think that often we have the story in our heads that sometimes things are adversarial. As a manager, I’m trying to get people to be more productive, so I’m trying to get them to do a thing that maybe they don’t want to do. I’m going to try and get them to work harder or I’m going to somehow put more pressure on them. If you start thinking like that, that’s not a win-win way to be thinking, right? That’s like you saying, “My getting better outcomes has to come at the expense of somebody else losing something.”
And I think if you start thinking like that, it’s very difficult to come up with a strategy or to truly be successful. But if you say, “Look, actually, my job is to figure out how to create win-wins.” So I actually don’t want somebody over the long run to feel like what I’ve done is just create a ton of pressure for them and now they’re super burnt out, real quick, because that’s not good for our team, that’s not good for me, that’s not good for our long-term relationship. How do we find the solution that can be a win-win? And I think if you think like that, a lot of things get easier. So for example, with new managers, I think this is true for me, too, the first time I had to tell someone that they shouldn’t be a part of this team was extremely fraught for me. And the main reason was because I’m putting myself in their shoes, and I’m imagining that this is truly horrible, and I’ve just done a huge disservice to this person, and that’s the most awful thing.
But there’s another way to look at it, which is, hey, if there’s persons on the team, they probably want to be successful. They want to do great work, they want to be valued, they want to grow their career. If this is not the place for them, because it doesn’t align with their true interests and the things that are going to help them be successful is just not the thing that they either want to do or can do at this point, it doesn’t do that person any good for me to somehow try to continue to make it. It’s actually going to be miserable. I’m going back to prolonging that misery state.
And so, sometimes a win-win thing is to just say, “Look, it’s not working, and I respect and value you so much that I know you want to do something that you can be proud of and you can grow in, and that’s going to be really valued. And right here, what we got, this isn’t it.”
That’s like a win-win way of looking at the situation, not a like, “Oh, my firing them is just definitely going to be a horrible…”
I’m not trying to say it’s not going to be hard, obviously it’s hard, but it’s in the mentality and the mental model I think makes all the difference. Because it’s going to be different in the way that I convey it to them. It’s going to be different and why this actually in the grander scheme of things may be great, and it’s going to reduce this adversarial feeling where they’re now going to see me as an enemy or somebody with all this power who’s making choices that impact them and they feel powerless. It has to be a collaboration. And I think if it’s not win-win… And I could be wrong. I would say I don’t think it’s right. The person could actually say, “No, you’re wrong.” And that would actually be great information, because then maybe we can go back and we can find a way to make it win-win.
Lenny Rachitsky: Yeah, I was just going to say, they have to believe this. You can’t just make it sound like this, “Here’s the win you’re getting let go. It’s a huge win for you.” But in reality, the way you phrased it, it is actually almost always true, “This is just not a place that you’ll be happy and succeed at, and it’s better you go do something else.”
Julie Zhuo: Yeah.
Lenny Rachitsky: Okay. I’m going to keep fishing in this pool to see what else we got, but when we run out, let me know. Is there anything else that you think people should know, should hear, especially new managers that they’re still not fully getting?
Julie Zhuo: I think being aware of your own energy and conviction is really, really important. So a lot of these themes, as you can see, go back to, you have to first understand this about yourself and have the right mindset, and when you do, it becomes much easier to be able to be impactful with other people. So this is another one. I think it’s very difficult for managers to be able… We talked a lot about the three pillars of what are the major tools of a manager. The first is people. And so, we talked a lot about the importance of dimensionality and feedback and helping reflect and grow people.
I think the second one is around purpose. Purpose is like, “What are we here to do? What’s our North Star?”
And I think it’s very hard to actually convey that if you don’t have conviction yourself. And so watching your conviction is really important, particularly since a lot of people who are managers, you often start out not as the founder of the CEO of the company, but you might be a middle manager. So in some ways, you didn’t create the vision, but you are in some ways expected to execute it or take a piece of it and do it. And I find that sometimes what new managers don’t pay attention to enough is what is their true belief. They feel like they might have to be a soldier, so they just get orders and they have to execute it. But it really makes a difference if they themselves have gone through the work of thinking through, “Wait, why are we doing this? Do I believe this strategy? Does it make sense or not?”
And if it doesn’t make sense, to go and actually have the conversation with their manager or whoever else, just so they can get to alignment on, “I really believe in what I’m doing.”
Because if you don’t really believe in what you’re doing or you’re just parroting the thing that got passed through the organization, it’s very hard for you to then be able to help other people see what that magic is or to be actually really effective as a person who can hold that vision and that purpose. So I just think you have to really check in with yourself on like, “Wait, I know we’re told to do this and this is what we have to do, but how do we really feel about it?”
Because if you don’t feel good about it, then it’s not going to be very likely that the project’s going to succeed. I can tell you right now, every single manager I’ve ever managed where they’re like, “I don’t really think this is a good idea,” there’s no case where I can think of where the project somehow turned out to be wildly successful.
Lenny Rachitsky: This is such a classic challenge of managers, is getting things done that you don’t really agree with. And I can’t help but ask you for advice on someone that isn’t in that place of just, “Okay, we have this feature our CEO’s prioritizing. This is not a good idea, but I need to have a brave face and not make it sound like I’m just being told what to do and I’m just reporting orders. I don’t believe in this.” You don’t want to do that. You become a terrible unsuccessful manager and people lose trust in you. What’s your advice to folks that are in that place of just how to find that balance?
Julie Zhuo: So I think, first, if you feel that way, you got to actually find a way to get it out and engage in dialogue. So if you’re like, “My manager told me to do this, I think it’s a terrible idea,” you’ve got to talk to your manager about it or you’ve got to talk to the CEO or whoever and feel… Because once you engage in a dialogue, what will often happen is you’ll learn more, you’ll have new information, you’ll have new assumptions, and maybe you’ll have influenced a project in some manner. But often, the more you can learn about, “Okay, why did some other smart people feel like we should do this? And what parts of it do I believe and what parts am I more skeptical about?”
You can probably decompose it from a blanket it’s good or bad to like, “Okay, this is a hypothesis, this is a hypothesis, this is a hypothesis. I might kind of believe this one. The reason I don’t like the proposal, I don’t believe this particular hypothesis, but I believe these other ones.”
And so, when you can start to get one level deeper into breaking it down into a set of assumptions, that makes it much easier, because then likely find something that you do kind of resonate with. And you might be able to then steer things like, “Okay, if that hypothesis doesn’t… I believe in disagree and commit, but now we can be very specific. We can isolate the thing that…”
And what we can also often do is like, “Okay, the reason I didn’t like this proposal is because I believe that this assumption is wrong.”
I’m going to come up with a really stupid example. But your suggestion is, “I know we have a great idea. We’re going to go and put a lemonade stand on every block. And my core assumption is people do not like lemonade. That’s not the hot beverage right now. And so therefore, I think this is a stupid plan.”
But if I talk to you about it and you’re like, “No, no, this is the core assumption we disagree on.”
Likely what starts to unfold is like, “Well, can we get some data? Can we get some information? Is there a quicker way to validate whether people like lemonade? Perhaps we should just test it in one market before we go and open up the lemonade stands across the entire 50 states.”
And so what happens is we can likely get to the actual specific area and come up with something. And then, if I have to now share with my team, We’re going to try this hypothesis. I’m not sure how I feel about it, but I actually do think… I don’t know for sure and our CEO seems to think this is… But we’re just going to test it.”
For sure, and our CEO seems to think this is… But we’re just going to test it, and we’re going to do the test in a way where… That’s what we want to find out, is do 18 to 25-year-olds love lemonade if we put them on these neighborhood college campuses? It becomes very specific and everyone’s like, “Well, yeah, I don’t know for sure, but I’m happy to go in, and test that, and commit to it.”
Lenny Rachitsky: This is such a good advice, and there’s also, you could layer on, “Here’s the things I do agree with and believe. Here’s the ways that I see this as totally right. Here’s the piece that I’m not so sure about, but that’s why we’re going to run this test, and here’s why it’s the smallest version of this test and why it’s a great idea just to figure it out.” We’ll show them. You probably don’t want to say that. As you give this answer, it’s so interesting, I almost want to do a whole new episode with you later of just common conundrums managers have, challenges that every manager runs into that are really difficult to figure out on the spot. We could save that for the future. Okay, I’m going to take us to a couple of recurring themes on this podcast, occasional recurring that every episode corners that we take guests to.
The first is I want to take us to AI corner. What I like to do in AI corner is ask, what’s a way that you’ve figured out to use AI in your work or your life that’s just really interesting, really useful?
Julie Zhuo: Well, I already shared a lot about education and learning, but I’ll share maybe a more fun story. It’s my kid’s birthdays. One of them just passed. My middle son’s birthday is in two weeks and my daughter’s birthday is in a month.
Lenny Rachitsky: By the way, the birthday just passed. The kid didn’t pass.
Julie Zhuo: Okay. Yes, the birthday passed.
Lenny Rachitsky: [inaudible 01:16:35].
Julie Zhuo: That’s right, that’s right. The birthday passed, my kid’s birthday. One of my goals this year was to try and build them something, so give them a present that has me going back to being the IC and making something for them. AI makes this really fun, and so just from my youngest son who was six years old, this is an idea that I stole from Eric Antonow, if you know Eric. Have you had him on your podcast?
Lenny Rachitsky: I haven’t, I am trying to. He actually sent me the… What is it? The-
Julie Zhuo: Yes, yes.
Lenny Rachitsky: What is it called? The metha-
Julie Zhuo: Methaphone?
Lenny Rachitsky: Methaphone.
Julie Zhuo: Yes.
Lenny Rachitsky: Methaphone, check this out.
Julie Zhuo: Yeah, yes.
Lenny Rachitsky: It’s like instead of holding the phone in your pocket, you hold this thing, and then you walk around with it and everyone’s like, “What the hell is that?” Methaphone.
Julie Zhuo: Yeah. I, too, am the proud owner of a methaphone and the next version upgrades with the little stickers, but-
Lenny Rachitsky: No, I don’t have that one yet.
Julie Zhuo: … Eric is great. You should definitely have him on your… He’s such a creative character. One time, I saw him with a parrot on his shoulder, and I was like, “Why do you have a parrot on your shoulder?”
He’s like, “Well, you can talk to my parrot. It’s a talking parrot,” and then I spoke to the parrot and the parrot spoke back to me. What had happened is that he had hooked up a microphone, he surgically went into the parrot and added a microphone, a speaker, and connected it to voice mode on ChatGPT so that… It spoken I think like a pirate voice.
I was like, “This is the best idea.” My six-year-old son is really into raccoons. He has a huge amount of raccoon stuffies. I was like, “I want a raccoon that can talk to him,” so I made that using the Eric Antonow method, but it was great. It was a huge hit. Now, my middle son’s birthday’s coming up, and he is really into parody. He loves video games, so Minecraft, but what he often listens to on his Alexa are these parody songs. It’ll be like Justin Bieber’s hit or Gangnam Style, but they’ve changed the lyrics so it becomes a video game parody of some video game that he’s playing, and they’re horribly sung. They’re like off-tune, it’s just like some person who produced it. I was like, “Well, if he doesn’t seem to mind off-key singing, I’m going to create him an album of video game parody songs, and I’m going to create an…”
I created an app on Replit, and what it does is you just give it a song. This is Justin Bieber’s Baby and you link to a Spotify song, and I give him some context like, “Oh, Locke likes playing Kingdom Rush right now. We have an inside joke about the gargoyles being free money.” Whatever it is, I just give it a bunch of context. I’m like, “Write me a song that just personalizes it and it’s a parody of this particular video game.” It writes me the lyrics. It’s pretty good at doing this. It’s pretty high quality. Again, it does it according to the beats of the music, and then I just sing it and record it, and then I got myself a song, so I’m creating an album of this, which I’m going to give to him. He’s not going to hear this podcast, so no one spoil it to him. I think this is going to go publish after his birthday, but I’m very excited about this.
Lenny Rachitsky: Wait, so you’re going to be the one singing the song?
Julie Zhuo: Yes, yes.
Lenny Rachitsky: I thought you were going to use Suna or some AI thing to actually sing it.
Julie Zhuo: No, I think I’m going to sing it myself.
Lenny Rachitsky: Wow.
Julie Zhuo: All of this made it so easy. All I have to do is just record. Again, I’m not a very good singer, but it doesn’t turn him off to hear off-key singing.
Lenny Rachitsky: Yeah. Wow, that is so beautiful. This gave me so many ideas for gifts I can give to kids in my life, and I just love how AI is making it, I don’t know, easier to be a parent and, in some ways, more delightful. These are awesome examples. Okay, I’m going to take us to a different corner, contrarian corner. What’s something that you believe that most other people don’t, people would disagree with?
Julie Zhuo: I believe that there’s infinity in every direction. That makes me pretty contrarian on pretty much everything that anyone says. If someone says something like on Twitter, I sometimes play this game with myself, which is in what context would that actually not be true? I think the reality is that the world is so, or at least my reality and my understanding of the reality, is that the world is just infinitely complex. For example, if my kids say something like going outside is boring, or taking a walk is boring, or doing something is boring, my general response will be, “Well, it’s because you’re not seeing the infinity that’s in that direction.”
Even, for example, something really mundane like staring at a blank wall, I think that you can make that actually deeply, deeply interesting, because you can use that as an opportunity to go into your own mind and to figure out how you can make time pass, or you can meditate on the existence or meditate on your breath, or just be grateful for the purpose of being alive. Two people, one person you can say, “Sit in front of a wall for an hour,” and, like my kid, they will super complain and be like, “This is the worst thing ever,” but you can put somebody else like a monk and they’ll have a wonderful experience. It’s not really about the environment or the wall. It’s really about how we see it and whether we can find the thing that is deep, and rich, and infinite in that direction.
Lenny Rachitsky: Wow, these are some deep answers. This is very, I don’t know, Buddhist, very mindfulness-oriented. I did a retreat once and their advice was just anytime you’re bored, just notice all the things that are going on around you. What does your seat feel like right now? What does the air feel like? What are you hearing right now? It’s exactly what you’re saying, there’s infinite things to pay attention to and keep you interested. It’s hard.
Julie Zhuo: It’s hard.
Lenny Rachitsky: Hard to actually do that for a long time and practice. That’s why it’s a practice.
Julie Zhuo: That’s why it’s a practice. But I repeat that to myself, because oftentimes, if I have a bad experience feeling a certain way, it helps me to realize that it’s often probably in my head. It’s because I haven’t gained the skills to be able to see the richness and infinity in that… I can maybe work on that. That feels better than feeling like, “Oh, I’m a victim of my circumstances. This thing happened to me,” and that’s so awful but not powerless, I can’t do anything about it. That, to me, is a worse feeling than the alternative, which is I just don’t have the skill yet. I can recognize it for what it is. I don’t have the skill yet, but I can grow. I can maybe get better at it. There is a person out there who had the same situation as me and feels much more positively than I do, and don’t I want to be more like that person?
Lenny Rachitsky: It’s such a beautiful circle back to our very first episode, which a lot of it was on imposter syndrome and overcoming that and your story there, so I love that that’s maybe a way to close this conversation. But before we do that and before we get to our very exciting lightning ground, is there anything else that you wanted to mention, or share, or double down on that we’ve talked about?
Julie Zhuo: I just want to say thank you. Honestly, I’m so inspired by the work that you do. I know we’ve known each other for quite a while, and I just think from the very first idea that you had for this newsletter, for the podcast, has been incredible, and I think the world gets so much from it. I’m sure you hear that a lot, but I am very grateful.
Lenny Rachitsky: Well, I really appreciate that, and I say this every time we do a chat, is just this wouldn’t have been possible without you, Julie. I was inspired by your longtime newsletter, The Looking Glass. Essentially, my idea was what if I do this for product? I started on Medium just like you did, and then I moved to Substack, and then it’s like, “What if I charge for this?” That worked, and then I’m like, “What if I do a podcast?” and then that worked. But it all began with your concept, so thank you, Julie.
Julie Zhuo: Yeah. I think you do it with so much kindness and curiosity as you always have, so I love that.
Lenny Rachitsky: That’s just who I am. Well, with that, we have reached our very exciting lighting round. I’ve got five questions for you. Are you ready?
Julie Zhuo: I’m ready.
Lenny Rachitsky: What are two or three books that you find yourself recommending most to other people?
Julie Zhuo: The first is Zen and the Art of Motorcycle Maintenance. I absolutely love that book. It’s beautifully written. It’s so deep. My whole philosophy around quality is beautifully… A lot of it comes from that book, the idea and even all the stuff that we talked about change. What does it mean to be at that forefront of change and dynamic quality? I think he just talks about so beautifully and so masterfully in that book. Old classic, but I try to reread it every few years or so. Second is Conscious Business. It is my favorite management book. It’s a little bit of a sleeper head because I actually end up recommending this one far more than my own book.
Lenny Rachitsky: Oh, wow.
Julie Zhuo: I read this one after I wrote my book, and I always tell people that if I read it before, I’m not sure I would’ve written my book, because I would’ve been like, “Conscious Business is really the book that really, really so much resonates.” Many of the things I talked about, this idea of win-win, idea of being a player, not a victim, and how to think about work, not just it’s a job but how do you really think about aligning it with your own personal values and what you want to do in the world, I think that this book really speaks to that so beautifully. It is also very tactical. It’s got a lot of really wonderful examples. I will tell people, the cover isn’t very attractive, and I think that if you judge a book by its cover, this seems very corporate-y. The title also seems like, “What conscious business?” and the first chapter is a little bit more technical. But if you just get past it and get into chapter two and you start with examples of the soccer team, it’s just the best management book.
Lenny Rachitsky: That is good advice to get people over the hump when they look for it. They’re like, “Okay, okay, I’m going to stick with it.”
Julie Zhuo: Yes. Okay, third book. I love the book Good Inside by Dr. Becky. It’s a parenting book and it’s a very wildly popular parenting book, so I really recommend it to all parents, but I also think it’s just a wonderful book for thinking about relationships, because parenting is that. It’s like a very, very deep and intense relationship and interaction that you have with another human being, and there’s so many things that I read in parenting books, including Good Inside by Dr. Becky, that I think could just as well been a management or a team leadership book.
Lenny Rachitsky: I am thinking about trying to ask Dr. Becky to come on the podcast. I feel like there could be a lot of synergy exactly for that reason. She uses this term sturdy, which inspired maybe your bullet tree process.
Julie Zhuo: Oh, yeah, I probably got it… I think she talks a lot about sturdiness and that just incepted right in here.
Lenny Rachitsky: Yes. Yeah. Her whole thing is being a sturdy parent. Strong but flexible, I imagine. Yeah. I love her and I love her stuff. I watch all her videos on TikTok and Emily Oster. Okay, next question. Is there a movie or TV show you recently enjoyed?
Julie Zhuo: I have not watched anything. I have no good answer for you. I think the only thing I watched this year was a rewatch of La La Land, which I do truly love.
Lenny Rachitsky: So delightful. Okay. Is there a product you recently discovered that you really love?
Julie Zhuo: I don’t think there’s anything too new. I love Granola, I love Replit. I’ve used all of the different coding lamps. Cursor is big on me for now. I just got a Matic Robot. I think that’s been really delightful so far, at least the setup. I haven’t used it long, long term, but it’s the setup, the way that it worked. The fact that it had little stickers and you could make it into a dog or a cat was a wonderful experience.
Lenny Rachitsky: The Matic Robot, willing to it, I am also a huge fan. I’m not an investor that’s… Essentially, Waymo meets Roomba. For folks that don’t anything about it, it’s like a very sophisticated robot vacuum built by AI vision people.
Julie Zhuo: Oh, I just thought of one more as well, the Limitless Pendant. Disclaimer, I am a small investor in Limitless, but what I love about it is that… Okay. It’s a pendant, you wear it, and it just records everything that’s going on, and later it summarizes things and it gives you feedback. I don’t usually wear it out because I find that maybe other people feel awkward that I’m recording everything, I usually try and get people’s permission, but I do wear it at home when I’m with my kids, and one of the best things that the pendant does is it gives me feedback on parenting.
Lenny Rachitsky: What? Automatically or run into ChatGPT?
Julie Zhuo: No, automatically. There’s an app and it will sometimes notify me, or if I check it, it’ll… Or I can also engage with Ask It, but what it does is essentially… It’s like Granola, but for your life in terms of capturing everything, summarizing it, and then giving you tips and feedback. It’s said things like, “Hey, there was that time you were talking about the game and you cut your kid off a lot. Maybe next time, think about letting them speak fully and listening better.”
Lenny Rachitsky: The app itself natively does that?
Julie Zhuo: Yeah.
Lenny Rachitsky: I did not know that, because I have one. I haven’t used it much recently. That is incredible. I wonder if it gives you relationship advice too if you’re talking to your partner. I wonder how it even knows.
Julie Zhuo: Yeah. It did a pretty good job of inferring. I think I said person two, but it was kind of eye-opening for me.
Lenny Rachitsky: Incredible. There’s a recent episode of our How I AI podcast, our sister podcast, where somebody wears that in their meetings with their CEO and automatically turns what they’re asking for into a prototype from the meeting notes, and then sales teams can start showing it to people to see if they’re interested. How about that?
Julie Zhuo: That’s awesome. That is super cool.
Lenny Rachitsky: Holy moly. [inaudible 01:30:22], what is even happening? Okay, I’ll keep going. Do you have a favorite life motto that you find yourself repeating to yourself, sharing with others?
Julie Zhuo: I like make it happen. Just a reminder that, at the end of the day, we could have a lot of motion. Maybe this is another one that I really like. I think about this poster. It used to be a poster at Facebook that says “don’t mistake Motion for progress”. There’s this idea of be the change We want to be in the world, I guess is other ways of saying the same thing, which is I can do things. We can all do things. We have better and better tools to go out there and make things happen. Make it happen.
Lenny Rachitsky: The common meme on Twitter, you can just do things.
Julie Zhuo: Yes.
Lenny Rachitsky: Final question. I like to ask this question to folks that are really deep in AI, and been working with AI, and getting a sense of where things are going. Is there something that you teach your kids or teaching your kids, think about encouraging them to learn, knowing that AI is going to be a big part of their life?
Julie Zhuo: Emotional regulation is still really, really, really important. That’s probably the thing that I think about the most in terms of what I want my kids to learn. I want my kids to be able to introspect, to have a better understanding of where their state of mind is, because we’re still human. We still have the same hardware that humans have had for thousands of years, and that’s not changing even as the tools and the environment around us change, and so I feel that you have to really understand yourself and you have to understand what’s going on for you and where you are biased and where you’re not, because AI can make it… This is my great fear, is that it makes things so much more comfortable. I have this great fear that this has been the trajectory that we’ve been on with technology. This is, again, going back to every strength is a weakness.
Technology makes things a lot easier. That’s why we invent, that’s why we create. Human race has always been about trying to better our circumstances and, in some ways, control our destiny, control our future. But at the same time, all of that control gets to a point where we have so many shortcuts in our lives and you can shortcut a lot of things. You can shortcut relationships, you can shortcut hard feelings, because now you can just watch TikTok instead of actually dealing with a very difficult emotion or tension that you had with a colleague, or with your partner, or with your children. AI makes it even, I think, more attractive, because now there’s a person or there’s a thing that can be very, very personalized, and if you’re like, “Oh, I want a distraction, I want to do something,” you got that.
But how do we actually still learn to sit with what is our true biology that’s not changing, and how do we continue to be the kind of people that want to take on the freedom of doing challenging things? Because I find that if we don’t do challenging things, we suffer. We suffer in a different way, and so, to me, true freedom is you can pick the things that are hard and you can feel pride in becoming the thing that you want to be. It’s not forced upon you. It’s not for survival’s sake anymore, but you still have to pick. I want to figure out for my children the fact that it is really important to still find the challenge. Yes, you can use AI to do that, but really, don’t think about it as a shortcut tool, because if that’s the case, I don’t actually think that they’re going to be able to become the kind of people they want to be in the world.
Lenny Rachitsky: What a beautiful way to end this conversation. Julie, it feels like this is just some kind of huge milestone of this podcast. Just like having you back three years later, it’s like, I don’t know, a chapter in the journey. I appreciate you coming back. I appreciate you sharing all this wisdom with us. Two final questions. Where can folks find you online if they want to reach out and maybe chat about maybe Sundial, maybe whatever else you’re up to, and then how can listeners be useful to you?
Julie Zhuo: Well, I would love to work with people who are at companies building really cool things and want better answers to how we build better, and so if you think your company would be interested in working with us at Sundial and figuring out how do we make every single decision maker into their own expert analyst, please reach out. That’s one area, sundial@sundial.so. I am on X, so I’ve been tweeting a lot more, sharing thoughts. Going back to that skill of practicing, just share what’s on your mind.
But for the long form stuff, I have my blog, The Looking Glass. It’s on Substack. I share articles and thoughts about AI, product building, leadership periodically, and then, of course, I have my book, the revised edition with two additional chapters. One is around managing remotely and the other one is around managing in a downturn or managing in difficult change scenarios. That will be coming out in two weeks’ time. The new content will be in the paperback. That’s important. I’ll send you a version of this when I get a copy myself, Lenny-
Lenny Rachitsky: Sweet.
Julie Zhuo: … but the paperback has a gradient type of cover. The hardback will eventually get the new content, but it just takes a while to phase out from all of the different retailers, so if you buy one, I cannot guarantee that it’s going to have the new content. But certainly, the Kindle and the paperback will have all of the new content.
Lenny Rachitsky: Just for the publish day, because this might come out later, what’s the date that’s coming up just for folks?
Julie Zhuo: September 9th.
Lenny Rachitsky: Okay, amazing. I think it’ll be out by the time this is out, so go buy it. I imagine available on Amazon, all your local retailers.
Julie Zhuo: Yes, yes.
Lenny Rachitsky: Amazing. Julie, thank you so much for being here.
Julie Zhuo: Thank you so much, Lenny. This was so fun. I hope to be back in another three years or whatever the next chapter is.
Lenny Rachitsky: Hopefully sooner. Bye, everyone.
Julie Zhuo: Bye.
Lenny Rachitsky: Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.
Glossary
| English | 中文 |
|---|---|
| ”don’t mistake motion for progress" | "don’t mistake motion for progress”(Facebook 经典海报标语,意为”不要把忙碌误认为进步”) |
| agent | agent(智能代理,此处保留原文) |
| AGI (Artificial General Intelligence) | AGI(通用人工智能) |
| alignment | alignment(在此处管理语境中指目标一致/对齐) |
| ARR (Annual Recurring Revenue) | ARR(年度经常性收入) |
| Ask It | Ask It(Limitless app 内的 AI 问答功能) |
| be the change you want to be in the world | be the change you want to be in the world(源自甘地名言,意为”成为你想在世界上看到的改变”) |
| builder | 建造者 |
| bullet tree | bullet tree(Julie 提到的流程/方法,保留原文) |
| Conscious Business | 《Conscious Business》(书名,保留原文) |
| Cursor | Cursor(AI 代码编辑器,保留原文) |
| dimensionality | 维度(指多维度评估个人能力的框架概念) |
| disagree and commit | disagree and commit(一种决策文化,即表达异议后仍全力执行) |
| Dr. Becky | Dr. Becky(育儿专家,保留原文) |
| Dunning-Kruger effect | Dunning-Kruger 效应(一种认知偏差) |
| Emily Oster | Emily Oster(人名,保留原文) |
| Eric Antonow | Eric Antonow(人名,保留原文) |
| Ethan Evans | Ethan Evans(人名,保留原文) |
| Good Inside | 《Good Inside》(Dr. Becky 的育儿书,保留原文) |
| Granola | Granola(AI 会议笔记工具,保留原文) |
| High Output Management | High Output Management(Andy Grove 的经典管理著作,此处保留书名原文) |
| How I AI | How I AI(Lenny 的姊妹播客名称) |
| IC (Individual Contributor) | IC(个人贡献者) |
| imposter syndrome | 冒名顶替综合征 |
| instrument/instrumentation | 埋点 |
| Kevin Wheal | Kevin Wheal(人名,保留原文) |
| Kingdom Rush | Kingdom Rush(一款塔防电子游戏) |
| La La Land | 《La La Land》(爱乐之城) |
| Limitless Pendant | Limitless Pendant(AI 可穿戴设备,保留原文) |
| make it happen | make it happen(Julie 的人生座右铭,意为”去把它做成”) |
| Marc Benioff | Marc Benioff(Salesforce CEO,此处保留原文) |
| Matic Robot | Matic Robot(机器人吸尘器产品,保留原文) |
| Methaphone | Methaphone(Eric Antonow 设计的一款产品名称) |
| Mike Krieger | Mike Krieger(Instagram 联合创始人,保留原文) |
| Nick Turley | Nick Turley(ChatGPT 团队成员) |
| north star | 北极星(指代核心愿景) |
| observability | 可观测性 |
| Replit | Replit(在线编程平台) |
| retention | 留存 |
| Sundial | Sundial(Julie Zhuo 创办的公司) |
| Suno | Suno(AI 音乐生成工具) |
| The Looking Glass | The Looking Glass(Julie Zhuo 的 newsletter 名称) |
| The Magic Loop | The Magic Loop(Ethan Evans 提出的职业发展框架名称) |
| The Making of a Manager | The Making of a Manager(Julie Zhuo 的著作) |
| VP (Vice President) | VP(副总裁) |
| Zen and the Art of Motorcycle Maintenance | 《Zen and the Art of Motorcycle Maintenance》(禅与摩托车维修艺术) |
Reformatted by reformat_english.py
从管理人员到管理 AI:当下人人都需要的领导力技能 | Julie Zhuo
文字稿
Lenny Rachitsky: 我们正在看到组织结构的扁平化趋势,每个人都重新变回了 IC(Individual Contributor,个人贡献者)。
Julie Zhuo: 过去是这样的——我没有能力同时做十种不同的工作,但现在 AI 让我自己就能胜任其中很多角色。我们需要打破传统角色的边界,称自己为”建造者”(builder)。我很希望我们能走向那样的世界,让”建造者”成为所有人的头衔。
Lenny Rachitsky: 我还刚看到一条数据,Google 裁掉了大量的中层管理者。
Julie Zhuo: 管理仍然至关重要。你需要有一个北极星、一个愿景,然后想方设法用手头的资源去实现它。过去资源是人,但现在基本上是各种模型,而不同的模型各有强项。你得像组建复仇者联盟一样,把合适的工具用在合适的地方。
Lenny Rachitsky: 你觉得如今管理者的角色和日常工作中最大的变化是什么?
Julie Zhuo: 管理者的职责一直就是应对变化,我只是觉得变化的节奏在加速。今天的管理,核心理念就是”坚定而灵活”。我常常想到柳树的比喻——它能挺过很多风暴和灾难,但同时又非常柔韧。
Lenny Rachitsky: 你从设计负责人到如今痴迷于数据和分析,这个轨迹非常有趣。
Julie Zhuo: 你要用数据来诊断,用设计来治疗。数据不会告诉你应该构建什么。我其实认为,目前很多高速增长的公司并没有真正用好数据。过去事物增长的速度远没有这么快,这些公司完全靠好的直觉和好的氛围就能撑住,但最终总会发生的是——增长总会停下来。
嘉宾介绍
Lenny Rachitsky: 今天的嘉宾是 Julie Zhuo。Julie 是我播客的第一位嘉宾,那期节目录制于三年多前,所以这是一次非常特别的对话。正如我之前多次在其他场合提到的,Julie 的 newsletter The Looking Glass 是我 newsletter 的灵感来源,基本上引领了我现在所做的一切。如果你不太了解 Julie,她曾是 Facebook 应用的长期设计负责人,该应用被超过三十亿人使用。她也是畅销且非常重要的著作 The Making of a Manager 的作者。最近她创办了自己的公司 Sundial,这是一个 AI 家长分析师,被 OpenAI、Gamma 和 Character.AI 等公司使用。Julie 是我所见过的最有思想、最有洞察力的产品领导者之一,她也拥有关于产品构建最有趣的视角之一。
曾在 Meta 这样的超大型企业担任设计负责人,如今又作为创始人在一家专注于利用数据辅助决策的小型创业公司工作,能拥有如此跨度经历的人非常罕见。在我们的对话中,我们聊到了为什么学习成为优秀管理者与学习如何出色地使用 AI 工具直接相关,未来几年哪些具体技能会变得更有价值,她给新管理者最有价值且历久弥新的建议,为什么她在自己的创业公司不招产品经理,以及她判断何时用数据、何时凭直觉做决策的简单经验法则。这期节目对每个人都会有收获。
(广告段落已跳过)
重逢闲聊
Lenny Rachitsky: Julie,非常感谢你能来,欢迎回到播客。
Julie Zhuo: 谢谢你,Lenny。我非常激动能来这里,期待了整整一周。我很喜欢你的播客,也很喜欢你从我们第一次对话至今把它发展成的样子,非常期待今天有一次有趣而深入的交流。
Lenny Rachitsky: 你能相信那第一期节目——这个播客的第一期——到现在已经三年多了吗?我的天。
Julie Zhuo: 我不太确定你那时候背后就有那团火。
Lenny Rachitsky: 有趣的是,我不知道有多少人注意到我保留的这个小彩蛋——在第一个录影棚里,我回看那期节目时,发现有一个滑稽的小镜子。我不确定第一期里是否有壁炉出现在镜子里,因为镜子里反射出了一些蠢东西。所以后来每换一个录影棚、每搬一次家,我都把这个壁炉保留了下来。
Julie Zhuo: 我甚至记得我们当时聊过,视频还是比较新的东西。你说”我们会录视频,但核心还是音频。“而现在我们已经进入了视频时代。
Lenny Rachitsky: 就在你刚才说那些的时候,我发现我的壁炉画面坏了,所以刚把它打开。中间剪掉了一小段。那个壁炉是我给自己留的一个小乐趣,我觉得从来没人发现过。
Julie Zhuo: 很温馨,我很喜欢。
Lenny Rachitsky: 就是这个效果。我刚刚还看了下数据,从第一期节目到现在,这个播客的下载量已经超过两千万次了,正接近三千万次。
《The Making of a Manager》再版
Julie Zhuo: 真的很了不起。我觉得这充分说明了你的好奇心,以及你对打造优秀产品这门手艺的投入和分享精神。我知道我自己和我的团队都会听你的播客、读你的 newsletter,我们经常分享你邀请的那些嘉宾带来的精彩内容,所以谢谢你做这些。
Lenny Rachitsky: 这是我的荣幸,非常感谢。三年后我们再次对话,原因是你重新发行了那本出色的书《The Making of a Manager》。我手边就有一本。这本书卖出了天文数字的销量,各种榜单上都能看到。这次你推出了平装版,还增加了一些章节。首先想问问,回顾这本书取得的成功,你有什么感受?
Julie Zhuo: 说实话,它的表现远远超出了我的预期,所以我非常开心。我写这本书的主要动力,很大程度上是因为我觉得如果必须把这件事写出来,我自己很可能会因此成为更好的管理者。这其实是很大一部分原因——我写博客已经很久了,而我知道自己的一个学习方式是,当我真正坐下来,试图把自己所有的感受写下来,给自己写封信,这对我的帮助非常大。所以这确实是一个巨大的动力。我当然希望它能出版并卖出一些,我当时想的是,也许对于在像我这样的公司——比如 Facebook 这种高规模的硅谷公司——成长起来的人来说,它可能会有共鸣。但我完全没有预料到它的影响范围会远远超出这个群体,这真的很棒。
有很多人告诉我,“我以为只有我一个人有这种感受,但这本书让我意识到,原来这些感受非常正常。“我自己当然也是这种感觉——跌跌撞撞地走过那么多年,一直觉得自己像个冒牌货。所以能从读者那里听到这样的反馈,真的非常令人欣慰。
管理技能与 AI 的交汇
Lenny Rachitsky: 我觉得它就是现代版的 High Output Management。那本书是这个播客中被提及次数最多的书,而我觉得你这本就是它的现代版本。老实说,那本书在许多方面已经过时了,所以我能理解为什么人们会如此被你这本书吸引。这也正好引出我想花时间聊的第一个话题——你作为管理者学到的大量技能,似乎同样适用于擅长使用 AI、用好 AI 工具。我想聊几个与这个大主题相关的趋势,想听听你的看法。
第一个趋势是,感觉在不久的将来,所有人都会变成管理者,因为 agent 已经深度融入了我们的工作流程。我们正在进入一个 agentic 社会,而那些让你成为优秀管理者的技能,似乎同样能让你擅长与 agent 协作。你对这个趋势怎么看?你觉得未来会怎样发展?
Julie Zhuo: 我百分之百相信这一点,也完全同意。在我心中,管理本质上是关于达成一个结果。你想做成一件事,就这么简单。你有一个北极星,有一个愿景,你要做的就是想办法利用手头的资源把那件事做成。通常当我们在传统语境下谈论管理时,我们说的资源是人——招募合适的人才,确保你组建了一支复仇者联盟一样的团队,拥有你需要的各种技能组合。第二个杠杆是关于目标——每个人都清楚自己应该用才干做什么吗?我们有共同的目标吗?我们有明确的使命吗?第三个要素是流程,也就是所有这些不同的人和工具应该怎样协作配合。
这些仍然是与 agentic 系统协作的基本功。你仍然需要目标,需要非常清楚地定义结果是什么,而且你必须了解各种模型——以前是了解人的强项,现在基本上是了解模型的强项。不同的模型有不同的擅长领域,就像它们有不同的性格一样。你得去熟悉它们,培养直觉,这样才能针对不同目的选用合适的工具。我们说 agent,但也要考虑 agent 能访问哪些工具——你仍然需要在这方面做出决策。当然还有流程,即具体怎么做。随着模型越来越好,agent 变得更聪明,它们能处理越来越高层次的任务自主规划,但我仍然认为,我们能够提供正确的上下文、给出正确的高层指令,对于获得我们想要的结果来说至关重要。
所以本质上,原则是一样的。我完全同意你的看法——我们中越来越多的人需要深耕这些技能,才能高效地使用这些工具。
Lenny Rachitsky: 顺着这条线说,我手边就有你的书。你列出了管理者的职责:建立一个协作良好的团队,支持成员实现他们的职业目标,以及创建流程让工作顺畅高效地完成——这基本就是刚才说的那些。有趣的是,中间那条是你在面对 agent 时不需要再操心的。你不需要关心 agent 的职业发展和成长。
Julie Zhuo: 确实如此。不过确实有人开玩笑说,如果我们不好好对待 agent,等到 AGI 来了会怎样?也许善待它们对我们仍然有好处。
Lenny Rachitsky: 我就是那种人——离开 Waymo 的时候会说谢谢,用 ChatGPT 语音模式的时候也会说”谢谢你,真的很有帮助”。沿着这个方向,我知道可以从很多角度展开,但就管理者所需技能而言,你觉得在与 agent 和 AI 系统协作时,哪些技能最值得培养?我会想到比如清晰度、沟通能力之类的。当你想到”作为管理者要重点发展的能力,这些能力同时也能让你擅长使用 AI 工具和与 agent 协作”时,你脑海中浮现的是什么?
定义目标:最核心的技能
Julie Zhuo: 首先是定义目标和结果,对”成功长什么样”做到极其、极其清晰。如果你让一家公司来做这件事,我们会发现这对人类来说本身就很困难。很多时候当我们讨论为什么大公司的 alignment 这么难,归根结底就是这个原因——不同的人对成功的画面可能有完全不同的理解。哪怕我用人类的语言描述——Lenny,我想打造这个产品,它会很棒;或者这期播客,你邀请了我,希望很多人听到并有所收获——这还是非常笼统。我们怎么才能更具体,具体到我们能毫无争议地判断是否达成了目标?这其实是一个非常、非常困难的问题。
这对我们来说之所以困难,是因为我们倾向于用非常高层的方式思考。所以要把它层层拆解,让 agent 真正理解成功和失败分别长什么样——这才是关键所在。我觉得这也和很多相关实践联系在一起,比如这就是为什么我们要写 eval,也是为什么 eval 如此重要,因为它们帮我们界定什么是客观标准。现在我从事数据领域的工作,我的公司专注于尝试自动化数据分析。而数据的意义、指标的意义、KPI 的意义,归根结底都是我们试图建立更客观的衡量标准,尽可能清晰地定义”成功长什么样”。我觉得这与其说是一门科学,不如说是一门艺术,但这是第一件事。如果你对成功的定义不够清晰,那你给出的 prompt 大概率也不会产出最出色的工作成果。我认为这对管理团队如此,对管理 AI 更是如此。
组织扁平化趋势
Lenny Rachitsky: 好,让我换一个角度,聊聊我们看到的另一个趋势——组织扁平化,管理者被裁掉,所有人又变回了 IC(个人贡献者)。我之前刚请了 Airtable 的 CEO 来做播客,他的核心观点就是 CEO 必须重新变成 IC。他现在写的代码比以往任何时候都多,他的理念是你必须亲自扎到细节里去,了解什么是可能的,才能弄清楚产品应该做成什么样。我也刚看到一个数据,Google 裁掉了大量小型团队的中层管理者。就是这样一股扁平化的浪潮。那么问题来了——未来我们还需要管理者吗?以及你觉得这一切会怎么演变?
Julie Zhuo: 我认为 AI 在工作场所带来的真正承诺和魔力在于,它让每个个体都获得了远超以往的能力赋能。过去的情况是,我没有能力做十种不同的工作,所以我需要通过招人来补齐。我需要一个擅长设计的人,一个擅长写代码的人,一个擅长数据分析的人,然后我组建起这个团队。但现在有了 AI 作为我的伙伴,情况变成了——等等,AI 让我自己就能完成其中很多工作。当然,我做不到博士级别或者最高的 1%、10% 的水平,但如果我原本处于第零或第十百分位,AI 今天就能非常迅速地把我提升到第六十、第七十百分位,达到当前的前沿水平。
我觉得这打开了很多扇门。所以我最感到兴奋的一点,也是我一直在团队里反复强调的——我们需要打破传统角色之间的边界。过去我们有一个传统团队结构:工程师、产品经理、设计师、研究员、数据科学家。而现在的团队可以更像是——就两个人。他们可以来自这些传统职能中的任何一个,但关键在于他们现在可以用 AI 来帮自己完成过去其他角色才能做的事情。所以在某种意义上,我们可以放下所有这些不同的角色标签,称自己为建造者。我觉得这是思考我们每个人能成为什么的最通用方式。我们都可以是建造者。我很期待我们能进入一个那样的世界——建造者就是你的头衔。
Lenny Rachitsky: 有意思。这恰恰是我最近越来越多使用的词。我最初把这个播客和 newsletter 定位为面向产品经理的,后来我开始用”产品”这个更宽泛的词,再到现在我确实在用”建造者”这个词,我很喜欢这个词,原因正是你说的。而且这在最近的对话中越来越成为一个高频主题——边界正在模糊。我很好奇在你的公司里,这具体是什么样的?你们在做什么不同的事情?你在一线看到了什么和几年前不一样的东西?
Julie Zhuo: 我们裁减了一些角色。比如,我们原以为会需要一批产品经理。结果发现,如果没有产品经理——我知道这可能有点背离 Lenny 起家的初衷——但我发现有时候当你有设计师或产品经理在团队里,而我是工程师,那当我遇到问题、需要弄清楚产品定义时,我的第一反应就是——我有这些人,这大概就是他们的职责描述,所以我直接交给他们就好了。而我觉得这样做——虽然我们是出于礼貌,出于尊重每个人的职责边界——其实错失了一个机会,就是作为工程师的我本应该说:“等等,我也应该在上面花很多心思。“我需要理解并形成自己的判断——该做什么产品,用户体验应该是怎样的。
所以我们发现,如果把团队做得更小,甚至在 AI 出现之前就这样做,减少这些专职角色,反而让每个人都意识到——“我们团队没有产品经理,所以沟通是我自己的事。弄清楚如何为用户创造最大价值,现在也正式成了我的职责范围。“这就是为什么我非常推崇我们可以让团队更小,可以消除这些界限。当然,我不是说每个人都要做所有事情。我们仍然可以承认你在某项技能上可能比我强得多,但重点不在于角色头衔,而在于我们所处的具体情境。
我发现每当你在团队中赋能他们,让他们能够根据自身所处的具体情境采取更多行动,而不是依赖那些高层的规则、政策或”事情应该这样做”的教条,你就会得到更好的工作成果。更快的工作成果,以及更快乐的员工——因为人们觉得自己真正拥有权力去创造想要的东西。
招聘中的能力考量
Lenny Rachitsky: 这很有意思,就是”没有 PM”这个约束条件,反而迫使工程师意识到不能等别人来做,必须自己想办法解决。但显而易见的挑战是——他们得擅长这件事。从工程到真正擅长表达”这是我们要解决的问题、为什么这个问题重要、我们如何排优先级、我们如何达成 alignment”——这是一份非常不同的工作。在招聘这些工程师的时候,你知道可能不会招 PM,你会做什么不同的事情吗?因为要招到一个在所有这些方面都很强的人,感觉真的很难。
Julie Zhuo: 确实如此,我并不是说每个人都得擅长所有事情。我觉得这不现实。比如,如果我们组建一个团队,有几个工程师,但他们都不太擅长思考产品需求或用户视角,那我们可能确实需要给团队补充一个有这方面能力的人。这个人可能是设计师,也可能是另一个擅长这些的工程师,也可能是传统的产品经理,有时候甚至是一个在这方面很强的数据分析师。所以这个技能依然重要,团队仍然需要具备这个能力,否则可能产不出最好的结果。但我更倾向于这样思考:这个团队需要哪些技能?我们能不能现在找到几个人来覆盖它们?
打破职能边界
但这并不意味着我们就自动套用那个模板——需要一个 PM、一个设计师、三个工程师,需要这些。对我们来说另一个例子是,甚至在思考前端和后端工程的时候也是如此。以前的情况是有些人就是前端工程师,有些人就是后端工程师。所以如果你有一个项目,既有前端部分又有后端部分,捷径就是:我需要一个这个、一个那个,就这样搞定。但如果你说,你是一个工程师,你是一个建造者,这个东西有一点前端,但你知道吗?你大概率能自己搞明白。用 AI 帮你搞明白。当然找一个专家来 review 代码,或者给你一些高层面的指导,但你就去做吧。自从我们也开始推行这种方式之后,我们发现同样需要在一开始做一些投入,因为人们一开始没那么适应。他们需要学习,所以初期事情会花更久一些,需要多花一点时间,相比之下如果你确实安排了一个前端专家来做这个前端项目,可能时间会短一些。但长期来看,这个投入真的会得到回报,因为现在你有更多工程师变得更加全面,能够独自承担更多的工作。当然在特定场景下,比如这个项目前端比重极大,那当然还是让更专业的人来做。
AI 工具加速最多的职能
Lenny Rachitsky: 我很喜欢你拥有这样的经历——在 Meta 这样超大型公司工作过,现在又创办了自己的小型创业公司,而且正好处于这个”保持小规模、保持精益、每个人都做更多事情”的趋势当中。你能同时体验这两者,真的太棒了。关于这个我有几个问题:你观察到哪些职能在这些 AI 工具的加速下效果最明显?是工程吗?还是别的什么?另外,有没有哪些工具对你们帮助最大?就是那些让人听了会说”我应该去看看”的 AI 工具。我猜是 Cursor,但好奇还有没有别的。
Julie Zhuo: 对,工程肯定是其中之一,毕竟我们公司大部分都是工程师,所以这是我们投入较多关注的领域。我确实也看到更多人在做原型。我们有两个设计师,但我们也看到工程师也在做原型。我们有一个团队叫 product science,是一个很有趣的混合体——你可以把它想象成一个有很强数据分析背景的前线部署人员,既在扮演客户成功的角色,同时也在扮演产品的角色。你看到他们开始承担更多构建原型的任务,或者参与到一些工程工作中。所以看到每个人都能做一点其他领域的事情,而且大家都在互相鼓励,这真的很棒。另外我们最近也一直在努力推动的一件事就是,我们会跟工程师说:“嘿,你现在可以做分析了。”
他们的第一反应是:“哦,我不太懂分析。“这就是 ChatGPT 派上用场的地方。以前我们会说:“我必须跟一个真人学,我得问这个人,然后占用他大量时间让他给我解释所有东西。“而事实上,我认为现在 ChatGPT 或其他 AI 工具是更好的老师。我觉得我们可能还没有充分把它们用于加速我们的学习,甚至没有系统地利用它们来过一遍知识。有时候我会找一个在线课程,如果你去上一个课,它可能是 12 周的课程,我就把它喂给 ChatGPT,然后说:“帮我根据我喜欢的学习方式定制一个学习计划。“
个性化学习加速
我是一个非常需要例子的人。我需要很多”给我当五岁小孩解释”的风格,给我类比。我知道我团队里有些人会说:“这些例子根本说不通。“我们是不同类型的学习者,所以一个能为我们每个人个性化学习的工具,确实帮助我们——我觉得——加速了学习进程,比以前快得多地掌握这些技能。所以是的,这些工具很棒,我们可以用 Cursor,它帮助我们自动补全,帮我们写很多东西,但学习的加速,我觉得是我们所有人武器库中另一个可能被低估的工具。因为我知道每次我跟人聊起来,我们都会忘记——我们不会想到,等等,对,我们其实可以这样做,坐下来,可能 30 分钟或一个小时就能比以前快得多地学到东西。
Lenny Rachitsky: 这个观点太有意思了。有些工具是即时帮助你提速的,但有时候你也需要学习如何做某件事,需要一些基础性的课程。你团队在哪些方面做过这样的事?他们学习了什么?
Julie Zhuo: 我给你举个例子。今天早上我刚跟一个工程师聊过,他写了不少这类算法。我们公司做的事情之一是尝试自动化数据分析,所以我们显然需要理解最佳实践。如果有一个类型的问题……
……我们需要做的事情显然是理解最佳实践。如果有一类问题,比如”嘿,哪些功能才是用户真正愿意付费的?“我们需要搞清楚该做什么样的分析。这个工程师跟我说:“Julie,我觉得我真的理解了怎么做。我知道算法,我知道怎么做根因分析,知道怎么操作。但我不太理解的是为什么,或者什么时候这个最有用。在公司的什么场景下会用到这个?“因为他是工程师,他没有做过 PM 或高管那种会问这类问题的工作。而这恰恰是最完美的场景——对,传统上你可能去问某个人,但这个问题其实更具通用性,互联网上有大量相关资源。这恰恰是那种如果你直接跟 ChatGPT 聊,它很可能给你一个更好的答案,还能让你更深入地探讨的问题。
用 AI 检验学习成果
另外我们也在学习的第二个做法是,几乎像是把 ChatGPT 当作……用它来检验你的学习成果。它解释了一大堆东西,然后我经常喜欢做的是:“好,我读了这个,所以这个……”我尝试用自己的话把听到的复述回去。“所以这是不是意味着……这样想对不对,这个东西有点像这个类比?“然后 ChatGPT 会点评我:“对,这是正确的。“或者”不,你理解得不太对。事实上……”而且它总是试图委婉地说。这其实挺有意思的,它会先说”这差不多对了”,然后最终变成”你完全错了”。就是这种风格。但这真的帮助很大,因为它是交互式的,所以我们可以通过尝试用自己的方式复述回去,来真正检验我们是否理解了这个概念。
AI 时代的数据分析挑战
Lenny Rachitsky: AI 的这些突破正在以如此多的方式帮助我们进步、做更多事、学到更多、变得更好,这真是令人难以置信。我知道有一些负面效应,但这确实太了不起了。让我们变得更好、更快的途径如此之多。我想再多聊聊数据分析这个话题。你从在大公司工作到创办自己的小公司,从设计负责人到现在对数据和分析如此痴迷,这个轨迹真的很有趣。让我花点时间聊聊这个。那些已经摸索出如何用 AI 做数据分析和数据工作的公司,到底在做哪些不一样的事情?在提升数据工作能力方面,人们忽略了什么、错过了什么?我还想补充一点——感觉我们几乎在逐一扫除团队前进道路上的各种障碍:比如等待 PM 写 PRD,比如等待数据科学家给出分析结果。而这又是一个让每个团队成员都能突破瓶颈的利器。
Julie Zhuo: 你第一个问题是,一批 AI 公司是怎么使用数据的?有趣的是,我的回答可能有点出人意料——我其实并不认为目前很多高速增长的公司在很好地使用数据。主要原因在于,传统上事物的增长并没有那么快。如果你达到了一亿用户,你的公司大概已经存在了一段时间,而既然公司存在了一段时间,你就有时间去搭建诸如日志系统之类的东西,招了增长团队和数据团队,他们做了大量的工作来做埋点和数据转换,讨论我们业务的可观测性是什么样的。通常你有很多年的时间来逐步构建和完善这些,因为增长速度就是那样的。
而今天我们看到一些公司增长极其疯狂,团队可能只有十来个人、两三个人,但已经有了数亿的 ARR 和数亿用户。但你知道吗?他们其实并没有所有那些基础设施、那些日志系统来真正做数据分析。所以我会说这些公司完全是在靠好的直觉和好的感觉在支撑,我们也确实看到了这一点。有时候你并不真的需要数据分析才能做出管用的产品。但我认为数据帮助我们在做的事情,在我看来就是帮助我们反映真实的情况。所以当然,如果 ARR 在增长,太好了,继续做你正在做的事。但总会发生的是,增长最终会停下来。增长不可能永远持续。通常当增长停滞时,每个人都会问:“怎么回事?为什么会这样?”
这时候你就能看到,如果你之前做好了全面的埋点,对业务有非常好的可观测性模型,就更容易进行根因分析,甚至更容易预测增长是否会在某个时点放缓,更容易及早捕捉这些趋势。如果你对业务的运转方式和公司的关键杠杆没有良好的可观测性,那你就会手忙脚乱,而通常到了那个时候,人们才会开始在数据上大量投入。所以我不认为很多这些热门公司已经到了那一步。但我同时觉得有一个趋势是,每次出现新的技术变革,我们实际上都需要改变我们思考的方式——分析必须回答我们所面临的问题,而如果技术变了或环境变了,我们就需要新的分析方法论。
对话式分析的新范式
比如,当移动端成为主流时,会话数、每日会话数、移动端使用时长、会话长度这些指标就变得很重要,帮助我们理解用户是否在这个新媒介中获得了价值。我觉得今天的情况也是一样的。对话式分析完全不同。过去,比如说在 Google 的世界里,我知道你对购物感兴趣因为你点了购物标签,我知道你对地图感兴趣因为你点了地图标签,我们可以衡量点击。今天一切都是对话,所以我们实际上更难分辨用户的意图是什么。如果我在任何一个 LLM 产品上工作,我想最大的问题之一就是:哪些用例在增长,哪些用例在萎缩?这在今天要判断起来困难得多,因为不再是标签页或页面上的点击那么简单了。我们可能需要用 LLM 或机器学习模型来对用户意图进行分类。我们可能需要问这样的问题:对话的流程顺畅吗?比如,如果我只问了一个问题,没有来回多轮对话,用户是否获得了价值?归根结底我们一直在试图搞清楚这是否是一次好的体验,但现在我们需要真正发明新的方法论来帮助我们分析这些。
Lenny Rachitsky: 对,我觉得问题始终是,对于对话来说,你希望它是长对话还是短对话?什么是正确答案,什么更好?
Julie Zhuo: 是的。
Lenny Rachitsky: 我在播客上请过 ChatGPT 的 Nick Turley,结果发现他们早期发现最常见用例的方式之一,就是看 TikTok 上的评论和在 TikTok 上走红的内容。怎么样?
Julie Zhuo: 对,对。
设计师与数据的关系
Lenny Rachitsky: 好,我想回到你这条非常有趣、不同寻常的路径——从 Facebook 的设计负责人,到成为众多设计师的榜样,到现在你把时间花在一个数据创业公司上,痴迷于数据。通常来说,设计师并不是实验和数据的最大拥趸,也不太喜欢基于数据做决策。当你看到设计师们反驳说”不,我们不想被数据驱动,我们比数据更懂……我们有对美和优秀的直觉判断”之类的话,你觉得设计师们在说这些话、害怕写实验和用数据、想要把这些推开的时候,他们错过了什么?
Julie Zhuo: 有一句话是我和联合创始人很早期就一直私下讨论的,后来也分享给了很多我们合作的公司,那就是:你真正想要的是——用数据诊断,用设计治疗。数据不会告诉你应该构建什么、解决方案是什么、或者怎么解决留存不好的问题。它就是做不到。但它可以告诉你是否存在问题,以及那个问题或机会可能在哪里。而你仍然需要回到一个非常有创造力的过程中去,找出解决问题的最佳方式。所以我想说的第一点就是这个框架:数据帮助你搞清楚实际发生了什么,人们喜欢什么,他们在使用什么,等等。它只是给你一个更好地反映现实的故事。因为说到底,我们都有自己的故事。我们会说”哦,我的公司很棒,人们都喜欢我们”之类的。那是我想相信的故事,但现实可能是另一番景象。所以数据试图做的就是捕捉现实。顺便说一句,我不认为数据只是 AB 测试和我们可以衡量的量化指标。对我来说,数据也包括人们在 TikTok 上发了什么、哪些内容走红了,以及他们在 Twitter 上——或者说 X 上,我想现在应该这么叫了——说了什么。
如果你做了一次用户访谈,那依然是数据,只是那种数据更难提炼和量化。不过现在有了 AI,我们有了更好的综合分析工具。所以在我心目中这些都是数据,它们都在试图帮助我们理解到底发生了什么,现实中正在发生什么现象,以及我们如何理解它。你仍然需要去发明、创造和想象,没有公式,也没有哪门科学能准确告诉你怎么做出一个爆款。你可以做实验,这可能让你尝试更多东西,更严谨地理解短期效果。但这一切都非常依赖具体情境。AB 测试不会告诉你长期会发生什么,而且说到底,这一切仍然是数据,你仍然需要综合分析并决定怎么做。
所以这就是我想说的——用数据诊断,用设计治疗。我想跟设计师们说的第二点是,有时候我觉得——也许是那种数字带来的虚假精确感,我们很容易陷入其中,对吧?因为就好像,好,我们拿到了这些数字,数字涨了。但不是这样的——你仍然需要选择看哪些指标,这本身就是一门艺术,不是科学。你对”数字涨了 5%“这件事的解读——这是好还是不好——同样是一种解读,是艺术,不是科学。只是有时候我觉得我们会给自己制造一种感觉——我理解这种心理,有时候人有一种想要控制一切的本能,我们希望一切都有条不紊,我们希望知道只要做了 ABC,一切就会很好,我们的职业生涯会很棒,我们的产品会一飞冲天。
而我觉得设计师们对此的反驳是合理的,他们会说:“不,现实是这些东西本身就是模糊的,存在不确定性,我们永远无法确切知道。“我认为这些说法都很有道理。所以我还想说的一点是,我真的认同你不可能靠 AB 测试试出一个真正优秀的产品。我从根本上相信这一点,但我觉得我们不应该因噎废食。我认为其实……你明白吗?这不是非此即彼,不是数据或设计二选一。
这些都是供我们使用的工具,而且我想说的是,我遇到过的每一位杰出的设计师,都极度痴迷于更好地理解现实。他们想知道用户真正在想什么,他们真正在做什么。如果能读懂每一个用户的心思——这是作为设计师我们所有人都梦寐以求的事情——如果我能知道每个人在使用时的想法和感受,我的生活会轻松得多,因为那样我就能构建出越来越好的东西。所以数据就是试图帮助我们做到这一点。它并不完美,没有任何一个指标能像我们希望的那样给我们真正的确定性和精确度,但这并不意味着我们不能利用它来改善我们的产品开发。
Lenny Rachitsky: 我正想说你说到的这一点,就是我合作过的每一位优秀设计师都极度痴迷于数据,主动拥抱数据。而那些只是说”不,我觉得我挺好的,我知道什么是对的,为什么要让数据来告诉我们怎么做”的设计师则不然。而且就像你说的,数据不会告诉你该怎么做,它会告诉你机会在哪里。让我把话题拉回到管理这个话题上,我也许先问一个比较大的问题——随着 AI 的兴起,你觉得管理者在角色、日常工作以及生活方面最大的变化是什么?
Julie Zhuo: 我认为管理变革这件事——管理变革一直都是管理者的工作,而且总是伴随着各种混乱。我只是觉得变革的速度正在加快,过去几十年我们一直在看到这个趋势。所以我觉得人们对很多事情有了更多的不确定感,比如两年后 AI 会发展到什么程度?我不知道。谁真的知道呢?我们五年内会不会实现 AGI?那将很大程度上改变整个格局。更不用说,我认为很多组织都感受到了相当大的恐惧。比如我一直在设计领域发展职业生涯,现在这些工具在我做的事情上越来越强,那我该怎么办?我的职业生涯和未来会怎样?我是不是需要转型?我是不是需要学习不同的东西?
所以就是这种变革,这种不确定感。而且我觉得很多时候管理者需要应对这些,除此之外还有你之前提到的,他们还需要学习新技能,也就是管理 AI,管理他们工作武器库中这些更强大的工具。我认为这与十年、二十年、三十年前相比是非常不同的。所以我觉得变得更加重要的技能,显然包括沟通、反馈、同理心,以及与人类协作的能力,让他们理解——是的,我们正处于变革之中。我认为现在每位领导者都需要做这件事,我认识的每一位创业者、每一位 CEO 都在思考:如何传达”事情正在变化,我们需要对变革非常开放”这个信息?
如果我们固守旧的方式,我们很可能会被抛在后面,我们的产品会被抛在后面,甚至我们的做事方式也会被抛在后面。所以我们需要改变。我们需要改变我们的产品,也需要改变我们工作的方式,就像我们之前讨论的那样——更小的团队、更敏捷等等。但同时问题是,我们如何做到这一点,而不让所有人都吓坏了?不会变成”啊,一片混乱,什么都在变”。
所以我经常想到一个隐喻——柳树。柳树是一棵非常坚固的树,它能挺过很多风暴和灾害,但同时它又非常柔韧,枝条非常非常柔韧,而在某种程度上正是这种柔韧让它能够如此坚固。所以我认为今天的管理,核心就是”在柔韧中保持坚固”这个理念。这是一个很难做到的事情,但我觉得至少当我面对……我会对自己说:“像柳树一样,Julie。想象那棵柳树,试着去感受那种我们要一起做到的事情。”
Lenny Rachitsky: 这让我想起之前几位嘉宾说过的话。我邀请过 Marc Benioff 上播客,我问他:“你就是怎么应对所有这些变化的?现在有 agent,有——我不知道——正如你说的 AGI 即将到来,你怎么在这样的环境中生存下来?“他的建议就是,他说:“我总是说,‘好,太好了。这就是我们想要的。这很令人兴奋。我们有这么多机会,一点也不无聊。我们可以不断重塑自己。‘“他总是以一种”这是好事”的态度去拥抱变化。他回应那个问题的方式我永远忘不了。
管理者首先需要认识自己
Julie Zhuo: 我觉得如果你不认为这是好事,那生活会相当痛苦。接下来的日子会非常非常难熬。所以我确实认为,在一切条件相当的情况下,去拥抱它吧。如果你每天醒来都能把它看作机遇和令人兴奋的事,而不是恐惧——当然,它们就像硬币的两面,总是同时存在——但如果我们能更多地倾向于思考”它可能变成什么样”,同时承认另一面的存在,承认恐惧和不安依然在那里。而且我觉得,如果管理者试图假装那一面不存在,假装一切安好、没人心烦等等,那也是缺失了什么。你需要能够坦然地说:是的,这很难。改变就是难的。我们大概率会不安。我们会有一些混乱。这些都会发生,但我们会一起度过,因为我们会保持柔韧,我们会把目光放在更大的图景上,看到那些令人兴奋的可能性。
Lenny Rachitsky: 你刚才说的让我想起另一句话,我忘了具体是谁说的了,也许是 Kevin Wheal,也许是 Mike Krieger。他们说,现在这个时候将是未来最”正常”的时刻了。也就是说,以后只会越来越离奇。我觉得给人一种这样的感觉——好吧,享受现在的正常吧,因为以后只会更离奇——至少能让人们对未来可能的方向有一个真实的预期。
Julie Zhuo: 是的,是的。
Lenny Rachitsky: 真是个值得活着的时代。
Julie Zhuo: 确实是。
超越 AI 之外,管理中最历久弥新的课题
Lenny Rachitsky: 好,让我们把视角再拉远一些,聊聊……我想问你,抛开 AI 不谈,管理在很多方面其实没有变。它仍然是很多相同的工作——管理人,帮助他们成功,产出优秀的成果。有哪些你觉得管理者、尤其是新管理者仍然没有完全理解、需要多听听的、最经典、最重要的经验教训?你脑海中首先浮现的是什么?然后我们看看这个话题会通向哪里。
Julie Zhuo: 我首先想到的,是管理自己、认识自己的重要性。这是我书的第五章,叫”管理自己”。事实上,写这本书的时候,我本想把它放在第一章,但我的出版商说:“也许你应该先讲一些更实操的内容……”人们不一定认为管理他人或管理团队要从自身开始,但我真的从根本上相信这一点。因为我觉得我们每个人,就像所有人一样,有擅长的事情,也有不擅长的事情。而我非常相信,每一个优势本身就是一种劣势,每一个劣势本身也是一种优势。
不存在什么方法能让你的每个维度都达到100%。事实上,我认为最有趣的概念或框架之一——对我自己来说,甚至这也算是一种数据框架概念——就是”维度”(dimensionality)这个概念。维度的意思是,你是一个人,但我们可以在无限多个维度上来看你。比如说,Lenny 扔飞斧的水平怎么样?这就是一个维度。
Lenny Rachitsky: 还不错。
Julie Zhuo: Lenny 做播客主持人的水平怎么样?非常棒。
Lenny Rachitsky: 还行吧。好吧,谢谢。
Julie Zhuo: Lenny 在 AI 领域做一个从零到一的项目水平怎么样?对吧?你可以把这些想象成无限多个维度。而现实是,我们每个人的画像都是独一无二的,就像指纹一样。对你来说,这些是你真正擅长的领域,远超他人,排在前1%。然后有些领域你在前10%,有些领域你属于平均水平,还有一些维度你低于平均水平。我们每个人都是如此。我喜欢这个框架的原因是,如果你接受了这个模型,你会意识到,这些维度中没有哪一个就完全代表了你。所以我可以对你说:“Lenny,你的飞斧技术真的需要提高。“理想情况下你不会觉得”Julie 在说我是一个糟糕的人,我的身份认同受到了威胁”,对吧?因为它只是你的一个维度而已。但有时候会发生的情况是,我们会非常执着于某些维度,因为我们开始认为那就是我们是谁。我觉得管理者会这样,他们团队中的个体显然也会这样。一旦发生这种情况,就很难进行更客观的对话了——好吧,你可以在哪些方面做得更好?哪些方面可能会变差?所以我讲这些,是因为我觉得这个框架——至少对我来说,以及我交流过的许多人——帮助他们认识到,别人可以给你反馈,或者你在某些维度上可能不那么出色,你有进步空间,但这并不代表你就是谁,因为你是一个拥有无限多个维度的整体,它们中没有一个能代表你作为一个个体的真正价值。
我深信我们每个人都是美好而有价值的,当然我们拥有各种技能,我们也希望提升这些技能,但这些技能的强弱并不决定我们是否值得被肯定。所以我觉得,如果你能把这一点真正内化,那么你作为管理者就能更客观地审视自己,你会意识到有些领域你真的很强,有些领域你有偏见,而且它们往往是同一回事。我举个例子。人们经常告诉我,我过去在绩效评估中会从管理者那里收到这样的反馈:“嘿,Julie,你很深思熟虑。当你思考一件事的时候,你有自己思考的方式,你显然做了深入的思考,有这些框架和所有的东西。这是很好的优点。“然后另一方面,我会收到这样的反馈:“嗯,Julie,你在动态讨论中不太发言。你比较安静,临场反应不够快。“你会发现,这两件事其实是一体两面的。因为我不那样做,我不是随口就说,这才让我很多时候能够做到非常深思熟虑——或者至少,好吧,在我年轻的时候很明显——那个特定的弱点同时也映射着一个特定的优点,那就是我是那种不会总是立刻做出判断的人。我必须真正地去思考、去消化,有时候需要时间才能理清自己的感受,然后才能把它分享出来、呈现给世界。
所以,了解自己这一点是极其有帮助的。当然,这并不意味着我永远无法在这方面进步。但我经常思考的是,所谓的掌控,就是意识到这两方面我们都可以提升,而我们需要做的就是在具体情境中判断怎样做才最合适。比如我收到这个反馈后,我就想:“好,我需要改进的一点是,想办法在面对面交流时更开放一些,说得更清楚一些,也许可以说类似’我还不完全确定我的感受,但我目前的想法是……’这样的话。“这些都是明确的策略,能让我成为更有效的团队成员,在我和团队一起做事的情境中做得更好。
Julie Zhuo: 所以我一直在努力培养这些技能,但更上层的元能力是能够退后一步说:好,在某些情境下,快速行动、果断决策、先做了再说真的很重要。即使不够完美,我们也得先做起来。如果我在这一点上感到吃力,我应该意识到这是我需要改进的地方。但在其他情境下,正确的做法恰恰是退后一步,深思熟虑,不急于做决定。
所以我想要达到的,并不是要否定这个优势或那个弱点,而是认清我们的出发点——我们天然可能就是以某种方式运作的。我们的成长往往表现为:在不擅长的那一面变得更好,但不是再次否定我们擅长的事情,而是随着时间推移达到一种平衡——能够读懂情境,判断”此刻我是不是该再多想一步,还是说这正是我需要果断一些、直接说出我当下想法的时候?”
Lenny Rachitsky: 我很喜欢这个建议——我们极其擅长的事情也有其代价,而我们收到的反馈中那些不太擅长的方面,往往也有一个别人欣赏的正面版本。我本来想问你一个问题,我觉得你已经回答了大部分,但还是想确认一下:当你收到”Julie,你在会议上说得不够多,贡献不够及时”这样的反馈时,一种选择是说”好,那就是我,我就是这样的,我用别的方式解决问题就好了,不做任何改变。“但我听到你说的是,要找到那些你确实想改变那种行为的时机,即便那不是你的天性,尤其是在事情推进很快的具体场景下。我想问的是,你建议人们在多少程度上推动自己去改善不擅长的方面,而不是进一步发挥自己的优势?
Julie Zhuo: 哦,我觉得这是一个非常好的问题。我思考这个问题的方法是,这很大程度上取决于你的目标是什么。比如,假设你……我们拿 IC(个人贡献者)和管理者来举例。我经常思考 IC 的路径,一个 IC 想要不断精进一门手艺。你热爱这件事,你想在这项非常具体的技能或手艺上变得越来越好。所以在我们的维度模型中,就像你挑了几个维度,“我就想把这几个做到前 0.01%“,这基本上就是 IC 的深耕路径。如果你的高层次目标是这样——比如你的目标是”我希望能每天花十个小时做这件事,因为我热爱它,我希望靠它养活自己,也就是说我能拿到报酬,有一份好工作,同时通过做这件事在世界上产生影响力”——那你仍然是有目标的。然后你就需要审视:“仅仅持续深耕这几项技能,这个策略能否帮我达成目标?“如果能,太好了。那如果有人问”你想不想做管理者?“你可以说”不,不需要,因为这是我的目标,而我目前的路径已经能让我达成。”
但如果你走到某个节点,发现你最想精通的技能在商业世界里并不那么有价值,不足以让你买下你想要的大房子来支撑你的家庭,那你就得问自己:“好,如果我只做这件事,那是不够的。我可能实际上还需要学习一些其他技能,才能胜任一份更有价值的工作,让人们愿意在那个层级付给我足够的钱,让我买得起那栋大房子。“所以我认为一切都要回到你的目标是什么。有些情况下,深耕你的手艺确实能支撑你的目标;有些情况下则不然。我认为这一个非常因人而异的问题,每个人都需要自己去思考。
但我经常认为,所谓痛苦,就是当这些东西不对齐的时候。你想要那栋大房子和所有一切,但你又说”可我只想把所有时间花在精进我的煎蛋卷技术上”。于是你就处于一种拉扯状态,很难感到满足和充实,因为你会觉得”为什么这个世界不重视我的煎蛋卷技术?“你要么放弃那栋大房子——也许你不该追求那个;要么你想要那个,那你可能就不能只做煎蛋卷了。也许你需要拓展你的烹饪技能,去打造一家米其林星级餐厅之类的。
Lenny Rachitsky: 这个建议真的很好。不是简单的”一定要弥补你的短板”或”别管它们”,而是——如果你需要做这件事来达成你的某个目标,那你得先搞清楚你的目标是什么,然后判断这件事是不是你确实需要努力的。比如,如果你想成为 VP,你大概率需要在大型重要会议中表现出色,能够临场应对,而不是等一切都结束了再发一封邮件分享你的所有想法。
Julie Zhuo: 没错。
Lenny Rachitsky: 对我来说,我确实经历过一个阶段,当时我想”我不想晋升,我待在这个非常具体的角色里就很开心,别管我。“那条路径所需要的技能,和成为管理者需要培养的技能完全不同。后来情况发生了变化,好,那这些就是我现在需要努力的方向。
主动设计自己的职业路径
Julie Zhuo: 是的。我很欣赏你对自己的那份了解,因为我觉得一个年轻人进入职场时太容易被周围的声音裹挟——也许全家人一直在跟你说”你需要升级,你需要赚更多钱,你需要拿到那个经理头衔,你需要当上 VP。“到了某个阶段,有些人会在不了解自己到底在签什么的情况下就选择了这条路。取舍是什么?那真的是你想要的吗?那真的符合你的热情所在吗?当然,有时候我们不得不做出妥协,但我们可以去设计——我们可以设计自己的目标是什么,什么是最适合自己的路径。我总是回过头来说,当人们不开心的时候,通常就是因为这些东西之间存在错位。他们想要那个宏大的目标,但对达成那个目标需要付出的东西并不真正兴奋,于是就会出现不匹配。
Lenny Rachitsky: 顺着这个思路,听起来似乎是——当然,我可以设计我的生活和我的角色。但我的发现是,如果你至少首先清楚自己真正热爱和想做的事情是什么,然后至少把这件事告诉你的管理者,它往往比你想象的要可行得多。
Julie Zhuo: 百分之百同意。我觉得非常重要的一点是要……我们还经常有这样一个心理模型:“管理者是我的评判者,他们要评判我做得好不好,我该不该升职,我该不该被解雇。”
Julie Zhuo: 所以人们有时候会有这种恐惧,但我觉得在最理想的上下级关系中,管理者更像是一个向导。你要知道,管理者有自己的职责,如果你理解管理者的工作——也就是如何让团队产出更好的结果——同时你也清楚你的管理者认为团队怎样才算成功,那你就会更容易判断:“哦,如果我做这个项目,那显然是直接为团队创造价值的路径,而且这个项目也适合我的技能,是我感兴趣的事情。“你应该主动把这个建议提给你的管理者。
反过来也是一样的。如果你真的去问你的管理者:“你的工作是什么?你认为怎样才算成功?你最大的希望和梦想是什么?“了解了这些背景信息,你就能更好地推动自己的职业发展。反过来,如果你说:“嘿,管理者,这些是我的希望和梦想,我觉得自己擅长这些,我真的很想在这个技能上有所提升,我特别想拿到那个 VP 的晋升,但我不知道它具体需要什么。你能告诉我,需要达到什么条件吗?“这也是一段非常棒的对话,因为你会获得所有这些背景信息,然后你才能真正决定自己是否想做这件事。如果你想的话,就请你的管理者帮忙:“好的,如果你看到能帮助我提升演讲能力或加强沟通能力的机会,请告诉我。“更进一步,“如果你对我的沟通有任何反馈,我想听,因为那才是帮助我在这个特定技能上成长的东西。”
所以,这应该变成一种协作关系,而不是那种近乎对抗性的——“我试图让你给我升职,你试图让我更卖力地工作”。这种氛围可不太好。
Lenny Rachitsky: 这让我想起 Ethan Evans 的一篇客座文章,我会附上链接,里面有一个非常好的框架,恰好就是讲你说的这件事的,叫做 The Magic Loop。它基本上是一个帮助你弄清楚该做什么、如何让管理者看到你的能力并赢得信任的框架。
反馈的力量
Lenny Rachitsky: 顺着这个经典管理者建议的话题,尤其是针对新管理者的建议,你已经分享了很多。还有没有其他你觉得特别重要、特别有意思或有价值的内容?
Julie Zhuo: 反馈是我另一个超级、超级热衷的话题。我的总体感觉是——无论是我自己,还是我共事过的每一个人——我们对反馈的重视程度都不够,或者说我们对反馈的思考不够深入。同样,公司都有绩效考核周期,所以我们都会觉得,“好,每六个月我们去做一次评审,那时候我会得到反馈。“但在我看来,反馈理想情况下应该是一种日常实践。因为从长远来看,对一个团队真正重要的是进步的速度有多快。一个每周进步 1% 的团队,相比于一个每月才进步 1% 的团队,即使前者的起点低得多,也会在很短的时间内超越那个不怎么进步的团队。
那么,让我们进步的最好工具是什么?就是反馈。我对反馈的看法,跟我们之前聊数据指标时说的非常类似——本质上就是把你对自己的假设拿去跟现实做检验。举个例子,也许我现在觉得自己是一个积极而有感染力的演讲者。我感觉自己在微笑,很有感染力,讲的故事也很精彩——但这真的是事实吗?我不知道。现实是我常常带有偏见,我们知道一些心理学效应,比如 Dunning-Kruger 效应,人们会认为自己对某件事的精通程度远远超过实际情况。你去问别人:“嘿,你的驾驶技术是否高于平均水平?“大概有百分之七八十的人会说:“是的,我高于平均水平。“这怎么可能呢?我们都有偏见。而冒名顶替综合征则是另一种方向上的偏见,就像我觉得”唉,我很差劲,我其实不配在这里”。这也是一种偏见,它未必是真的。事实上,我可能完全有资格在这里,其他人也很认可我的贡献。
所以,我们对自己的认知——我们的优势、弱点、当下正在发生的一切——很多时候跟真实情况存在巨大的偏差。而我们要真正理解并取得进步的方式,就是需要别人把他们看到的真实情况反馈给我们。我的想法是,比如这期播客结束后我会找你要反馈,然后你会告诉我一些东西。你给我的其实是一份礼物——一份把我看不到的东西反射回来的礼物。就像如果我后脑勺上有一片树叶,我自己是看不到的。所以如果你告诉我:“嘿,Julie,你头上有片树叶。""哦,哇,谢谢。“好吧,也许我就可以把树叶拿掉。反馈就是这样,它本质上是一种镜像反射,帮助我们校准到现实,让我获得关于自己是否正朝着目标方向前进的信息。
Lenny Rachitsky: 我很喜欢这个说法,完全同意。但正如你所知,大多数人面临的挑战是:如何给出反馈让接收者不会产生防御心理,以及如何接收反馈而不是立刻觉得”哦不,他们不懂,他们什么都不懂,他们怎么敢这样说我?“你能不能给我们一两个关于如何好好给出反馈、以及如何好好接收反馈的建议?甚至包括如何主动寻求反馈——如何获得更多反馈?这些道理都说得通,但大多数时候人们根本得不到什么反馈。
如何建立反馈关系
Julie Zhuo: 最好的方式……关于获取反馈或给出困难反馈的第一个建议是,先去真正建立起这样一种关系——我们珍视彼此的贡献,我们希望帮助彼此成长,因此我们会成为那种每周互相给反馈的人。所以当你刚开始和某人合作时,不要等到糟糕的事情发生了才去给反馈,因为那时已经是一个高压情境了。一开始就可以说:“嘿,很期待和你合作。我觉得我们最好的合作方式是我希望你帮我变得更好。我在这些方面比较擅长,在那些方面不太擅长。你呢?好,你觉得自己擅长这些?那我们就这样一起合作,互相帮助在这些方面提升?我们的做法是——所有反馈都开放。我希望你把一切都告诉我。理想情况下,你也会说:‘好啊,我也希望你把一切都告诉我。‘“这样我们就在一开始就确立了这种默契。
Lenny Rachitsky: 这适用于同事、manager,还是所有同事?
Julie Zhuo: 所有人都适用。就像你在约会的人,就像你的孩子。可以和所有人建立这样的关系,就是确定我们想要什么样的关系?我觉得大多数人都会选择一种可以亲近、彼此紧密的关系。可以彼此坦诚,不用藏在……我觉得大多数人会选择这种方式,而一旦你选择了,之后一切都会变得更容易。所以第一件事就是让每个人都”选择加入”——我们想要的是这种关系。
Lenny Rachitsky: 我听到过一个在这方面非常有效的小技巧,就是问对方:“你希望反馈是即时给的,还是每月或每周定期给?“大家的反应通常是:“不不,即时给,一有事就告诉我。“然后你就获得了那种自由——“好,那我就直接给你反馈了。”
Julie Zhuo: 所以如果你让人们选择加入——“是的,我希望我们有一段好的关系。我希望我们互相帮助变得更好。我希望得到反馈。“这就解决了给出困难反馈时百分之六十的难题。
然后第二个技巧是,当你真正给反馈的时候,有很大帮助。首先你要检查:“我给这个反馈,出发点是不是真的为了互相帮助?“如果答案是肯定的,那我们就从百分之六十推进到百分之八十了,这件事会很顺利。
但经常发生的情况是——某件事发生了,你做了某件事,触发了我的情绪,因为可能我之前在类似的事情上有过不好的经历。所以我感到愤怒,我想要证明自己是对的。如果我给你反馈的真正动机是为了验证自己、为了证明我是对的、为了告诉你你是错的、为了惩罚你——那这件事不会顺利的。这种意图本身就在那里。除非你是个极其出色的演员,否则不可能顺利。所以你首先要检查自己的意图。
但如果你做了这个检查,你告诉自己:“不不不,我仔细想过了。我现在冷静了。我不再怒火中烧。我真的认为 Lenny 只是没有意识到,当他说那句话的时候,会让我和其他人感到被排斥在外。“或者诸如此类的,对吧?那我就需要能够把这个告诉你。
给出困难反馈的技巧
所以通常接下来,如果你会想:“好吧,现在我开始紧张了,因为我不想冒犯你。我非常珍视我们的关系。我该怎么告诉你呢?我不想让你产生防御心理?“那第三个技巧就是——把这些话说出来。如果我坐下来跟你说:“Lenny,我现在特别紧张。我想给你一些反馈,但我真的很担心这会影响我们的关系。我太珍视我们的关系了,我不想让那种情况发生。但我也觉得,如果你能听到这些,对你会有帮助。“这能起到极大的作用——它是人性化的。你会意识到我在冒着风险,我在展现脆弱,而你很可能比我巧妙地把话抛过去更容易听进去。不要试图回避它的困难,而是真正去承认这件事很困难,把这一点暴露出来,因为这能建立很多人与人之间的连接。
Lenny Rachitsky: 这个建议太棒了。非常实用。好的,还有别的吗?我们讨论了很多经得起时间检验的管理智慧,尤其是新手 manager 需要听到的东西。你觉得还有什么真的很重要、但人们还没有真正领悟的关于成为优秀 manager 的事情?
管理中的双赢思维
Julie Zhuo: 我觉得”双赢”这个概念,我脑海里一直在想这个。我经常回到这个概念上,因为我觉得我们脑子里常常有一个故事,认为事情有时候是对立的。作为 manager,我在努力让人们更高效,所以我在试图让他们做他们可能不想做的事情。我要想办法让他们更努力,或者我要用某种方式给他们施加更多压力。如果你开始这样想,那这就不是双赢的思维方式了,对吧?那等于你在说:“我得到更好的结果,必须以别人失去什么为代价。”
我觉得如果你开始这样想,很难找到好的策略,也很难真正取得成功。但如果你说:“看,实际上,我的工作是想办法创造双赢。“所以其实我不想让一个人在长期来看觉得我做的只是给他制造了巨大的压力,然后他很快就筋疲力尽了——因为这对我们团队不好,对我不利,对我们长期的关系也不好。我们怎么能找到双赢的解决方案?我觉得如果你这样想,很多事情会变得更容易。
比如,对于新手 manager 来说,我觉得我也是如此——我第一次不得不告诉某人他不应该再留在这个团队时,对我来说极其煎熬。主要原因是我把自己放在他们的位置上,想象这对他们来说真是太糟糕了,我对这个人造成了巨大的伤害,这是最糟糕的事情。
但还有另一种看待方式,那就是——嘿,如果这个人在这支团队里,他大概也想要成功。他想做出出色的成果,想被认可,想让职业生涯有所成长。如果这里不适合他,因为这里与他真正的兴趣不吻合,能帮助他成功的那些东西不是他想要做的、或者现阶段能做的,那么我勉强让事情继续下去对他没有任何好处。实际上只会很痛苦。我又回到了那种延长痛苦状态的做法。
所以有时候,双赢的做法就是直接说:“看,这样行不通,而我是如此尊重和重视你,我知道你想做让你感到骄傲的、能让你成长的事情,那才是能被真正认可的。而在这里,我们目前的情况,不是那个。“这就是用双赢的方式看待这个局面,而不是觉得”哦,我让他们离开就一定是一件可怕的事情……”
我不是想说这不会很难——显然这很难——但心态和心理模型上的差异,我认为会带来天壤之别。因为我向他们传达的方式会不同。为什么从更大的格局来看这可能是一件好事,角度也会不同。这会减少那种对抗性的感觉——他们不会把我看作敌人,或者一个握有所有权力、做出影响他们的决定、而他们无能为力的人。这必须是一种协作。而我认为如果不是双赢的话……当然我也可能判断错误。我愿意接受他们说,“不,你错了。“这其实是很好的信息,因为也许我们可以回头找到一种方式让它变成双赢。
Lenny Rachitsky: 对,我正想说,他们必须真心相信这一点。你不能只是嘴上说说——“你被解雇了,这是你的大赢。“但实际上,按你刚才的表述,这几乎总是事实:“这个地方就是不适合你在这里快乐地工作和取得成功,你去做别的事情会更好。”
Julie Zhuo: 是的。
还有什么值得分享的
Lenny Rachitsky: 好。我继续在这个话题池里捞一捞看看还有什么,如果聊完了你告诉我。还有什么是你觉得大家应该知道、应该听到的,尤其是新 manager 们还没有完全领悟的?
Julie Zhuo: 我觉得觉察自己的能量和信念真的、真的非常重要。你可以看到,我们聊到的很多主题都回到同一个原点——你必须先了解自己,拥有正确的心态,然后你才能更轻松地在与他人合作中产生影响力。这又是一个类似的例子。
我觉得 manager 很难做到……我们之前讲了很多 manager 的三大支柱。第一个是人(people)。所以我们谈了很多关于维度的重要性、反馈、帮助人反思和成长。
第二个支柱是目的(purpose)。目的是指”我们在这里做什么?我们的北极星是什么?”
如果你自己没有信念,其实很难把它传达出去。所以审视自己的信念非常重要,尤其因为很多 manager 并不是公司的创始人或 CEO,你可能是一个中层 manager。所以在某种意义上,你没有创造愿景,但你被期望去执行它,或者承担其中一部分来实现它。我发现新 manager 有时候不够注意的一个问题是——自己真正的信念是什么。他们觉得自己可能需要做一个士兵,所以只是接收命令然后执行。但如果他们自己真的花功夫思考过”等等,我们为什么要做这件事?我相信这个策略吗?它合理还是不合理?“这会产生很大的不同。
如果不合理,就要去跟自己的 manager 或其他相关人员真正展开对话,这样你才能达到 alignment——“我真的相信我在做的事情。”
因为如果你自己都不真正相信你在做的事情,或者只是在鹦鹉学舌般地复述组织传下来的话,你很难让别人看到其中的魔力,也很难真正有效地作为一个持有愿景和目的的人发挥作用。所以我觉得你必须真正地跟自己核对——“等等,我知道我们被告知要做这个、要做那个,但我到底怎么想的?”
因为如果你对此感觉不好,那这个项目成功的可能性就不大。我可以直接告诉你——我管理过的每一个 manager,只要他们说”我真的不觉得这是个好主意”,我想不出任何一个案例,那个项目最后大获成功了。
如何执行你不认同的事情
Lenny Rachitsky: 这是 manager 们非常经典的一个挑战——去做你自己并不真正认同的事情。我忍不住想问你给出建议:当一个人并不是那种”好吧,CEO 优先排了这个功能,这不是个好主意,但我需要装出一副信心满满的样子,不能让人觉得我只是在传话、只是在汇报命令、我自己根本不信这个”——你不想那样做,你会变成一个糟糕的、不成功的 manager,人们也会对你失去信任。对于那些身处这种境地、想要找到平衡的人,你有什么建议?
Julie Zhuo: 我觉得,首先,如果你有这种感觉,你必须找到出口,展开对话。如果你的想法是”我的 manager 让我做这件事,我觉得糟透了”,你必须跟你的 manager 谈,或者跟 CEO 谈,或者跟相关负责人谈——因为一旦你展开对话,通常会发生的情形是:你会学到更多,获得新的信息,形成新的假设,也许你还会在某种程度上影响了项目的走向。
通常来说,你越能深入了解”好吧,为什么其他一些聪明人觉得我们应该做这件事?其中哪些部分我认同,哪些部分我更怀疑?“——你大概就能把它从一个笼统的”好或不好”拆解成:“好,这是一个假设,这是一个假设,这又是一个假设。这个我可能有点相信。我不喜欢这个提案的原因是我不相信这个特定的假设,但另外几个我信。”
所以,当你能深入一层,把它拆解成一组假设时,事情就变得容易多了,因为这样你很可能找到你确实有共鸣的部分。你可能就能够引导事情的方向——“好,如果那个假设不成立……我相信’不同意但执行’(disagree and commit),但现在我们可以非常具体。我们可以把那个东西单独拎出来……”
我们往往还能做的是——“好,我不喜欢这个提案的原因是因为我相信这个假设是错的。”
我举一个非常蠢的例子。你的提议是:“我有一个好主意,我们要在每个街区都摆一个柠檬水摊。而我的核心假设是——人们不喜欢柠檬水。现在这不是热门饮品。所以我认为这是一个愚蠢的计划。”
但如果我跟你谈了,你说,“不不不,这是我们意见不一致的核心假设。“接下来很可能展开的对话就是——“好吧,我们能拿到一些数据吗?能获取一些信息吗?有没有更快的方式验证人们是否喜欢柠檬水?也许我们应该先在一个市场测试,然后再去全美五十个州到处开柠檬水摊。”
于是我们就很可能定位到具体的分歧点,并想出一些办法。然后,如果我现在要跟团队分享——“我们要测试这个假设。我不太确定自己怎么看,但我确实觉得……我不完全确定,我们的 CEO 似乎认为这是……但我们就是去测试,而且我们要用这样的方式测试——“这正是我们要搞清楚的问题:如果把柠檬水摊放在这些大学校园附近,18 到 25 岁的年轻人会喜欢柠檬水吗?这样就变得非常具体,每个人都说——“嗯,对,我也不确定,但我很愿意去试试,测试一下,然后全力以赴。“
AI 角落
Lenny Rachitsky: 这条建议太好了,而且还可以再加一层:“以下是我认同并相信的部分。以下是我认为完全正确的地方。这个部分我不太确定,但正因如此我们才要跑这个测试,而且这是最小化的测试版本,所以去验证一下是个好主意。“我们走着瞧。你大概不想这么说。听你回答的时候,我觉得特别有意思,几乎想以后专门跟你录一期新的节目,就聊管理者常见的两难处境,每个管理者都会遇到但很难当场想清楚的那些挑战。这个可以留到以后再说。好,我接下来要进入这档播客的几个固定环节,每个节目都会带嘉宾去的几个角落。
第一个环节,我想带大家进入 AI 角落。在 AI 角落里,我喜欢问的是:你在工作或生活中找到了什么用 AI 的方式,真的特别有趣、特别实用?
Julie Zhuo: 嗯,我已经分享了很多关于教育和学习方面的内容,那我就讲一个更有趣的故事吧。我家孩子的生日,其中一个刚过不久。我二儿子的生日在两周后,我女儿的生日在一个月后。
Lenny Rachitsky: 顺便说一下,是生日刚过去了,不是孩子过去了。
Julie Zhuo: 对对,是生日过去了。
Lenny Rachitsky: 没错没错。
Julie Zhuo: 对,生日过去了,我孩子的生日。我今年的目标之一是亲手给他们做点什么,送他们一个让我回到 IC(个人贡献者)身份、自己动手做东西的礼物。AI 让这件事变得特别有趣。比如我小儿子,他六岁,这个主意是我从 Eric Antonow 那儿偷来的,如果你认识 Eric 的话。你请他上过你的播客吗?
Lenny Rachitsky: 还没有,我正在努力。他其实给我发了那个……叫什么来着?那个——
Julie Zhuo: 对对。
Lenny Rachitsky: 叫什么?Metha——
Julie Zhuo: Methaphone?
Lenny Rachitsky: Methaphone。
Julie Zhuo: 对。
Lenny Rachitsky: Methaphone,你看看这个。
Julie Zhuo: 对。
Lenny Rachitsky: 就是说,你不再把手机揣在口袋里,而是拿着这个东西,然后你拿着它到处走,所有人都像,“这什么鬼东西?“Methaphone。
Julie Zhuo: 对,我也是 Methaphone 的骄傲拥有者,而且下一版升级还带小贴纸,不过——
Lenny Rachitsky: 不,我还没拿到那个版本。
Julie Zhuo: Eric 特别棒。你一定要请他来你的节目……他是个特别有创意的人。有一次我看到他肩膀上站着一只鹦鹉,我就问他,“你肩膀上为什么有只鹦鹉?”
他说,“你可以跟我的鹦鹉说话,这是一只会说话的鹦鹉。“然后我就跟鹦鹉说话,鹦鹉也回了我话。原来发生的事情是这样的:他接了一个麦克风,他”外科手术式”地拆开鹦鹉,塞进去一个麦克风和一个扬声器,然后连到了 ChatGPT 的语音模式,所以……它说的是海盗口音。
我当时就想,“这简直是最好的主意。“我六岁的小儿子特别喜欢浣熊。他有超级多浣熊毛绒玩具。我就想,“我要做一只会跟他说话的浣熊。“于是我用了 Eric Antonow 的方法做了一只,效果很棒,反响非常大。现在,我二儿子的生日快到了,他特别喜欢恶搞。他很喜欢电子游戏,比如 Minecraft,但他在 Alexa 上经常听的是那些恶搞歌曲。就是 Justin Bieber 的热门歌曲或者《江南 Style》,但他们改了歌词,变成他玩的某个电子游戏的恶搞版,而且唱得非常难听。完全走调,就是某个随便做的人弄出来的。我就想,“好吧,既然他似乎不介意走调的演唱,那我就给他做一张电子游戏恶搞歌曲专辑,我要做一个……”
我在 Replit 上做了一个应用,它的用法是你给它一首歌。比如 Justin Bieber 的《Baby》,你给它一个 Spotify 的链接,然后我给它一些背景信息,比如”Locke 最近喜欢玩 Kingdom Rush。我们之间有个内部笑话,说石像鬼就是白捡的钱。“之类的,我就给它一堆上下文。然后我说,“给我写一首歌,把它个性化,做成这个特定电子游戏的恶搞版。“它就给我写出歌词来,做得相当不错,质量相当高。而且它会按照音乐的节拍来写,然后我只要唱出来录下来,一首歌就有了。我正在做一整张这样的专辑,准备送给他。他不会听到这期播客的,所以别有人给他剧透。我想这期节目会在他生日之后发布,但我对这件事非常兴奋。
Lenny Rachitsky: 等等,所以是你自己来唱这些歌?
Julie Zhuo: 对,对。
Lenny Rachitsky: 我还以为你会用 Suno 或者什么 AI 工具来唱。
Julie Zhuo: 不,我想我会自己唱。
Lenny Rachitsky: 哇。
Julie Zhuo: 所有这些都让事情变得太简单了。我只需要录就行了。再说了,我也不是什么好歌手,但他听走调的歌并不在意。
Lenny Rachitsky: 嗯。哇,这太美了。这给了我好多灵感,让我可以送给我生活中的孩子们。我就是喜欢 AI 怎么让做父母变得——我不知道怎么说——更轻松了,而且在某种程度上也更有乐趣了。这些都是超棒的例子。好,我要带大家进入另一个角落,反共识角落。你有什么看法是大多数人不认同的,别人会反对的?
反共识角落
Julie Zhuo: 我相信每个方向上都存在无限。这让我在几乎任何人说的任何事情上都显得很反主流。如果有人在 Twitter 上说了什么,我有时候会跟自己玩一个游戏——在什么情境下这句话实际上是不成立的?我认为现实是,世界是如此的——或者至少我的现实、我对现实的理解是——世界就是无限复杂的。比如,如果我的孩子说出门很无聊,或者散步很无聊,或者做什么事很无聊,我通常的回应会是,“那是因为你没有看到那个方向上的无限性。”
甚至,比如说盯着一面白墙这种非常平淡的事情,我觉得你也可以把它变得非常非常有趣,因为你可以把这个当作一个机会,走进你自己的内心世界,想出打发时间的方法,或者冥想存在本身,或者冥想你的呼吸,或者只是感恩活着这件事。同样是两个人,你对一个人说,“在墙面前坐一个小时。“像我孩子那样的,他们会超级抱怨,说”这是世界上最糟糕的事情。“但你换一个人,比如一个僧侣,他会有非常美好的体验。所以这其实不在于环境或者那面墙。真正重要的是我们如何看待它,以及我们能否找到那个方向上深邃的、丰富的、无限的东西。
Lenny Rachitsky: 哇,这些回答真的很深。这非常——怎么说呢,很有佛教的味道,非常正念导向。我曾经参加过一次禅修营,他们给的建议就是,任何时候你感到无聊,就去注意你周围正在发生的一切。你此刻坐着的感受是什么?空气的感觉是什么?你现在听到了什么?这跟你说的完全一样,有无限多的事情可以去关注,让你保持兴趣。不过这很难做到。
Julie Zhuo: 确实很难。
Lenny Rachitsky: 真要长时间坚持去做、去实践,是很难的。所以才叫”修行”。
Julie Zhuo: 所以才叫”修行”。但我会反复对自己说这句话,因为很多时候,如果我有某种不好的感受,意识到这往往可能只是我自己心理上的问题,会对我有帮助——是因为我还没有培养出那种能力,去看清其中丰富和无限的东西……我可以去练习这个。这比另一种感觉要好得多——“哦,我是环境的受害者,这件事发生在我身上”,感觉很糟糕却又无能为力,什么都做不了。对我来说,那是一种比另一种更糟的感受——另一种是,我只是还没有这项技能而已。我可以认清它本来的样子。我还没有这项技能,但我可以成长。我也许可以变得更好。世上有一个人和我处境相同,却比我积极得多,而我难道不想变得更像那个人吗?
Lenny Rachitsky: 这真的是一个非常美的呼应,回到了我们第一期节目的很多内容——那期很大一部分讲的是冒名顶替综合征以及如何克服它,还有你当时的故事。所以我很喜欢用这个来作为这段对话的收尾。但在那之前,在我们进入非常精彩的快问快答环节之前,你还有什么想提的、想分享的,或者想在之前聊过的基础上再强调的吗?
Julie Zhuo: 我只想说谢谢。说真的,你做的事情给了我很大的启发。我们认识也挺久了,我从你最初产生做这个 newsletter 和播客的想法开始,就觉得非常了不起,我觉得世界从中获益良多。我猜你应该经常听到这样的话,但我真的非常感激。
Lenny Rachitsky: 非常感谢你这么说。而且每次我们聊天我都会讲——这一切没有你是不可能实现的,Julie。我当初正是受到你长期经营的 newsletter The Looking Glass 的启发。基本上我的想法就是:如果我针对产品领域做类似的事情会怎样?我和你一样从 Medium 开始写,后来转到了 Substack,然后想”如果我对此收费呢?“——这行得通。然后我又想”如果我做一个播客呢?“——那也行得通。但这一切最初都源自你的理念,所以谢谢你,Julie。
Julie Zhuo: 是的。我觉得你做这些的时候一直带着那么多的善意和好奇心,一如你一贯的样子,我很喜欢这一点。
Lenny Rachitsky: 这就是我本来的样子嘛。好了,说到这里,我们终于到了非常精彩的快问快答环节。我准备了五个问题。准备好了吗?
Julie Zhuo: 准备好了。
推荐书籍
Lenny Rachitsky: 有两三本你发现自己最常向别人推荐的书是什么?
Julie Zhuo: 第一本是《Zen and the Art of Motorcycle Maintenance》(禅与摩托车维修艺术)。我非常喜欢那本书。文笔优美,思想深邃。我关于”品质”的整套哲学,很大程度上源自那本书——包括我们今天聊到的关于变化的一切,什么是站在变化前沿、什么是动态品质,我觉得他在书里讲得既优美又精妙。这是一本老经典了,但我试着每隔几年就重读一次。
第二本是《Conscious Business》。这是我最喜欢的管理书。这本书有点被低估了,因为我实际上推荐这本书的频率远远超过推荐我自己的书。
Lenny Rachitsky: 哇哦。
Julie Zhuo: 这本书是在我写完自己的书之后才读到的。我一直告诉别人,如果我先读到它,我不确定自己还会去写那本书,因为我会觉得”《Conscious Business》真的是那本让我深深共鸣的书”。我谈到的很多东西——双赢的理念,做一个参与者而非受害者的理念,以及如何看待工作——不只是把它当作一份差事,而是真正思考如何让它与你的个人价值观、与你想在这个世界上做的事情对齐——我觉得这本书把这些讲得非常美。它同时也非常实操,有很多非常精彩的案例。不过我要提醒大家,它的封面不太吸引人,如果你以貌取书的话,会觉得它非常”企业风”。书名也让人觉得,“什么’自觉商业’?“而且第一章也比较偏技术性。但只要熬过去,进入第二章,从那个足球队的案例开始读,那就是最好的管理书。
Lenny Rachitsky: 这个建议很好,可以帮助人们跨过那个门槛。他们会说,“好吧好吧,我会坚持读下去的。”
Julie Zhuo: 对。好,第三本书。我很喜欢 Dr. Becky 的《Good Inside》。这是一本育儿书,也是一本非常火的育儿书,所以我向所有父母强烈推荐。但我也认为它是一本关于思考人际关系的非常好的书,因为育儿本质上就是这样——它是你与另一个人之间非常非常深层、非常强烈的互动关系。我在育儿书里读到的很多东西,包括《Good Inside》,我觉得完全可以当作管理或团队领导力的书来读。
Lenny Rachitsky: 我正想着试试邀请 Dr. Becky 来上播客。我觉得可能会有很多协同效应,正是因为这个原因。她用了一个词叫”sturdy”(稳固),可能启发了你提到的 bullet tree 流程。
Julie Zhuo: 哦,对,我可能就是从那里学来的……她讲了很多关于”稳固”的东西,然后就这样植入了我脑子里。
Lenny Rachitsky: 是的。她的核心理念就是做一个”稳固的”父母。坚强但有弹性,我猜。是的,我很喜欢她,喜欢她的内容。我在 TikTok 上看了她所有的视频,还有 Emily Oster 的。好,下一个问题。
最近喜欢的影视
Lenny Rachitsky: 最近有没有一部让你很享受的电影或电视剧?
Julie Zhuo: 我什么都没看。我没有好的答案给你。我觉得今年我唯一看的就是重看了《La La Land》(爱乐之城),我确实非常喜欢。
Lenny Rachitsky: 真的很棒。好的。
最近喜欢的产品
Lenny Rachitsky: 最近有没有发现一个你非常喜欢的产品?
Julie Zhuo: 我觉得没有什么特别新的。我喜欢 Granola,喜欢 Replit。各种编码工具我都有用过。Cursor 目前对我很重要。我刚入手了一台 Matic Robot。目前为止我觉得它非常令人愉快,至少设置过程是的。我还没有长期使用过,但它的设置方式、运作方式,还有它附带小贴纸、可以把它装扮成狗或猫的样子——整个体验非常棒。
Lenny Rachitsky: Matic Robot——我也非常愿意推荐,我也是超级粉丝。我不是投资人……基本上就是 Waymo 遇上 Roomba。不了解的人可以理解为,它是一个由做 AI 视觉的人打造的非常先进的机器人吸尘器。
Julie Zhuo: 哦,我还想到一个——Limitless Pendant。声明一下,我是 Limitless 的小投资人。但我喜欢它的一点是……好,它是一个吊坠,你戴着它,它就会记录周围发生的一切,之后它会做总结并给你反馈。我一般不戴出门,因为我觉得别人可能会对我在记录一切感到不舒服,我通常会先征得别人的同意。但在家里和孩子们在一起时我会戴,而这个吊坠最棒的一点是——它会在育儿方面给我反馈。
Lenny Rachitsky: 什么?自动的还是丢到 ChatGPT 里跑的?
AI 育儿助手
Julie Zhuo: 自动生成的。它有一个 app,有时会给我发通知,或者我主动查看时也会看到……我也可以用它的 Ask It 功能。但本质上它做的事情就是……它就像 Granola,只不过是针对你生活的——捕捉一切,做总结,然后给你建议和反馈。它说过这样的话:“嘿,那次你在聊游戏的时候,你经常打断你的孩子。也许下次可以考虑让他们把话说完,更好地倾听。”
Lenny Rachitsky: app 本身就能做到这些?
Julie Zhuo: 对。
Lenny Rachitsky: 我都不知道,因为我也有一个,但最近没怎么用。这太厉害了。我好奇如果是在跟伴侣说话,它会不会也给你恋爱关系方面的建议。我甚至好奇它是怎么判断的。
Julie Zhuo: 会的。它的推断做得挺不错的。我想我当时说的是”第二个人”,但对我来说确实有点醍醐灌顶的感觉。
Lenny Rachitsky: 太厉害了。我们 How I AI 播客——我们的姊妹播客——最近有一期节目,有人戴着那个吊坠去跟 CEO 开会,然后把 CEO 的需求自动从会议记录中转化成一个原型,接着销售团队就可以直接拿去给客户看,测试他们的兴趣。厉害吧?
Julie Zhuo: 太棒了。超级酷。
人生座右铭
Lenny Rachitsky: 天哪。[听不清] 这都是什么时代了。好,我继续。你有没有最喜欢的人生座右铭,会经常对自己重复、或者分享给别人的?
Julie Zhuo: 我喜欢 make it happen。就是提醒自己,到头来,我们可以很忙碌,但也许这是另一个我很喜欢的说法——我想起一张海报。Facebook 以前有一张海报,上面写着”don’t mistake motion for progress”。还有一个说法是 be the change you want to be in the world,大概也是同一个意思,就是——我可以做事。我们都可以做事。我们拥有越来越好的工具去把事情做成。Make it happen。
Lenny Rachitsky: Twitter 上常见的那个 meme——you can just do things。
Julie Zhuo: 对。
AI 时代的孩子教育
Lenny Rachitsky: 最后一个问题。我喜欢问那些在 AI 领域很深入、一直在与 AI 共事、对行业走向有感觉的人——在知道 AI 将会成为孩子们生活的重要组成部分的前提下,你在教孩子们什么,或者在鼓励他们学什么?
Julie Zhuo: 情绪调节仍然非常、非常、非常重要。这大概是我在思考希望孩子们学到什么时想得最多的东西。我希望我的孩子们能够内省,更好地理解自己当下的心理状态,因为我们仍然是人类。我们仍然拥有和几千年来人类一样的”硬件”,这一点不会随着我们周围的工具和环境的变化而改变。因此我觉得你必须真正了解自己,了解自己当下的状态,了解自己在哪些地方有偏见、哪些地方没有,因为 AI 会让一切变得……这是我最大的担忧——它会让一切变得太舒适了。我一直有一种很大的担忧,觉得这其实就是技术发展的轨迹。说回那句话——每一个优势也是弱点。
技术让很多事情变得更简单。这就是我们发明创造的原因,这就是为什么我们一直在创新。人类一直以来都在试图改善自身处境,在某种意义上掌控自己的命运、掌控未来。但同时,所有这些掌控到了一个地步——我们生活中有了太多捷径,你可以走捷径绕过很多东西。你可以在人际关系上走捷径,你可以在面对困难情绪时走捷径,因为现在你完全可以刷 TikTok,而不是去真正处理一个很困难的情绪,或与同事、伴侣、孩子之间的紧张关系。我觉得 AI 让这一切变得更有吸引力了,因为现在有一个”人”或一个”东西”可以非常非常个性化地满足你——如果你说”哦,我想要一个消遣,我想做点什么”,你马上就能得到。
但我们究竟如何学会与自己那不会改变的真正生物本性共处?我们如何继续做那种愿意拥抱自由、去挑战难事的人?因为我发现,如果我们不去挑战难事,我们会痛苦。我们以一种不同的方式痛苦。所以对我来说,真正的自由是——你可以选择那些困难的事情,你可以在成为自己想成为的人的过程中感到骄傲。它不是被强加给你的。它不再是为了生存,但你仍然需要做出选择。我希望让我的孩子们明白,去寻找挑战仍然是非常重要的。当然,你可以用 AI 来做到这一点,但真的,不要把它当作一个捷径工具来看待,因为如果是那样的话,我不认为他们能成为自己在世界上想成为的那种人。
尾声
Lenny Rachitsky: 用这么美好的方式结束这次对话。Julie,感觉这期节目是这档播客的一个重要里程碑。三年后再次邀请你回来,就像是旅程中的一个新篇章。感谢你再次做客。感谢你与我们分享这些智慧。最后两个问题。大家如果想联系你,想聊聊 Sundial 或者你正在做的其他事情,可以在哪里找到你?然后,听众可以怎样帮到你?
Julie Zhuo: 我很希望能和在很酷的公司做产品的人合作,想要更好地回答”我们如何做出更好的产品”这个问题。如果你觉得你们公司有兴趣跟我们在 Sundial 合作,探索如何让每一个决策者都成为自己的专家分析师,请联系我们。这是一个方向,邮箱是 sundial@sundial.so。我也在 X 上,最近推文发得多了一些,分享自己的想法。回到练习那个技能——就是把脑子里想的东西分享出来。
至于长内容,我有我的博客 The Looking Glass,在 Substack 上。我会定期分享关于 AI、产品打造、领导力的文章和思考。然后当然,我的书——修订版增加了两章新内容。一章是关于远程管理的,另一章是关于在下行周期中管理,或者说在困难的变革场景中管理。这本书两周后就会上市。新内容会出在平装版里。这点很重要。等我拿到样书,我会寄一本给你,Lenny——
Lenny Rachitsky: 太好了。
Julie Zhuo: ——平装版的封面是渐变色的。精装版最终也会有新内容,但从各个零售渠道铺货需要一段时间,所以如果你买一本,我不能保证它一定包含新内容。但 Kindle 版和平装版肯定会有全部新内容。
Lenny Rachitsky: 确认一下出版日期,因为这期节目可能会在之后播出,具体是哪天?
Julie Zhuo: 9 月 9 号。
Lenny Rachitsky: 好的,太棒了。我想这期节目播出的时候书应该已经上市了,大家去购买吧。应该可以在亚马逊和各大零售商处买到。
Julie Zhuo: 对对。
Lenny Rachitsky: 太棒了。Julie,非常感谢你来。
Julie Zhuo: 非常感谢你,Lenny。这次太开心了。希望三年后还能再来,或者下一个篇章是什么都好。
Lenny Rachitsky: 希望不用等那么久。大家再见。
Julie Zhuo: 再见。
Lenny Rachitsky: 非常感谢你的收听。如果你觉得这期节目有价值,可以在 Apple Podcasts、Spotify 或你最喜欢的播客应用上订阅本节目。同时,也请考虑给我们评分或留下评论,这真的能帮助更多听众发现这个播客。你可以在 lennyspodcast.com 找到往期所有节目或了解更多关于本节目的信息。下期见。
术语表
| 原文 | 中文 |
|---|---|
| ”don’t mistake motion for progress" | "don’t mistake motion for progress”(Facebook 经典海报标语,意为”不要把忙碌误认为进步”) |
| agent | agent(智能代理,此处保留原文) |
| AGI (Artificial General Intelligence) | AGI(通用人工智能) |
| alignment | alignment(在此处管理语境中指目标一致/对齐) |
| ARR (Annual Recurring Revenue) | ARR(年度经常性收入) |
| Ask It | Ask It(Limitless app 内的 AI 问答功能) |
| be the change you want to be in the world | be the change you want to be in the world(源自甘地名言,意为”成为你想在世界上看到的改变”) |
| builder | 建造者 |
| bullet tree | bullet tree(Julie 提到的流程/方法,保留原文) |
| Conscious Business | 《Conscious Business》(书名,保留原文) |
| Cursor | Cursor(AI 代码编辑器,保留原文) |
| dimensionality | 维度(指多维度评估个人能力的框架概念) |
| disagree and commit | disagree and commit(一种决策文化,即表达异议后仍全力执行) |
| Dr. Becky | Dr. Becky(育儿专家,保留原文) |
| Dunning-Kruger effect | Dunning-Kruger 效应(一种认知偏差) |
| Emily Oster | Emily Oster(人名,保留原文) |
| Eric Antonow | Eric Antonow(人名,保留原文) |
| Ethan Evans | Ethan Evans(人名,保留原文) |
| Good Inside | 《Good Inside》(Dr. Becky 的育儿书,保留原文) |
| Granola | Granola(AI 会议笔记工具,保留原文) |
| High Output Management | High Output Management(Andy Grove 的经典管理著作,此处保留书名原文) |
| How I AI | How I AI(Lenny 的姊妹播客名称) |
| IC (Individual Contributor) | IC(个人贡献者) |
| imposter syndrome | 冒名顶替综合征 |
| instrument/instrumentation | 埋点 |
| Kevin Wheal | Kevin Wheal(人名,保留原文) |
| Kingdom Rush | Kingdom Rush(一款塔防电子游戏) |
| La La Land | 《La La Land》(爱乐之城) |
| Limitless Pendant | Limitless Pendant(AI 可穿戴设备,保留原文) |
| make it happen | make it happen(Julie 的人生座右铭,意为”去把它做成”) |
| Marc Benioff | Marc Benioff(Salesforce CEO,此处保留原文) |
| Matic Robot | Matic Robot(机器人吸尘器产品,保留原文) |
| Methaphone | Methaphone(Eric Antonow 设计的一款产品名称) |
| Mike Krieger | Mike Krieger(Instagram 联合创始人,保留原文) |
| Nick Turley | Nick Turley(ChatGPT 团队成员) |
| north star | 北极星(指代核心愿景) |
| observability | 可观测性 |
| Replit | Replit(在线编程平台) |
| retention | 留存 |
| Sundial | Sundial(Julie Zhuo 创办的公司) |
| Suno | Suno(AI 音乐生成工具) |
| The Looking Glass | The Looking Glass(Julie Zhuo 的 newsletter 名称) |
| The Magic Loop | The Magic Loop(Ethan Evans 提出的职业发展框架名称) |
| The Making of a Manager | The Making of a Manager(Julie Zhuo 的著作) |
| VP (Vice President) | VP(副总裁) |
| Zen and the Art of Motorcycle Maintenance | 《Zen and the Art of Motorcycle Maintenance》(禅与摩托车维修艺术) |
此文档由 AI 分片翻译(translate_long_document)