How to Manage Humans 🎽 — with Rands

I think a lot of, especially in tech, there's this sort of hero culture of like, oh, I saved the day. I'm like, good job saving the day. Why did the day need saving? Yes. We screwed up and you had to save the day. So good job cleaning up your screw up. Hey, Luca here.

Welcome to a new episode of the Refactoring Podcast, where every two weeks we interview a world-class tech leader and we discuss how to write good software and work well together. Today's guest is Michael Lopp, better known as Renz. who is senior director of engineering at Apple and writer of several popular books about engineering management, including Managing Humans, which is a personal favorite of mine.

With Renz we discussed his journey into writing and the impact it had on his personal and professional growth. We reflected on his leadership style and the complexities of modern engineering management. So let's dive right into it! Hi, Rans. Welcome. And thank you so much for being here with us today. Absolutely. Happy to be here. And thanks for asking whether to call me Ranslop or Michael, because nobody really knows what's going on.

Exactly. So I may still have some confusion and during this chat, I will switch to call you Michael, whatever. So I hope it's fine. So, um, Renz, you are a senior director of engineering at Apple. You've had an incredible career at many companies like Slack, volunteer, Netscape and more. Uh, but to me personally, you're first and foremost, an incredible writer.

your books on engineering management, like the art of leadership, managing humans, and of course your blog, Rans in Repose, which you've been writing for how long, like 20 years or something? Yeah, like two decades or something like that. It's crazy.

Yeah, it's crazy when you look back. The quality and the consistency as well. Long before it was cool, I should say. Yeah, I blogged actually before there was blogs. This is actually the second iteration. I did one a long time called the BitSifter Digest before things were blogged. like no one

blog software didn't exist. I was like publishing it myself. So, but it was the same format. It was like, you know, later today, latest stuff, you know, little paragraphs that I'm eating is what I'm doing, this sort of thing. So then I stopped and blogging got popular. I'm like, Oh, cool. This is kind of fun. And then. upriding rands for you know 20 years

Yeah. So we usually start these interviews by asking guests about how they got into tech right from the start. But since you've been writing for so long, I want to try a variation here. How did you start writing about work? think about thinking about writing building a routine yeah yeah so um uh writing i've done it since i was a kid i i've been journaling since i was like about 14 so um and so i've like

got piles of journals around and that's just i've had the habit for a long time that was just for me obviously i wasn't publishing that or not i think what one of the accelerants was just sort of tech arriving and sort of like the present omnipresence of computers and whatnot and

um typing just made it feel like i could do it on being more efficient so i was like writing essays and i wrote i wrote a book that you'll never read you know just on the computer because i i felt so empowering to be able to like as opposed to like

dear diary it was like a lot more so i got just a ton done that way and i think just got comfortable with it so it was always a thing that i i enjoyed doing and i i got compliments from people even like we had sent emails people like are you a writer i'm like uh write emails um so anyway the sort of the the change was the blog that we talked about ransom repose

And sort of getting in that habit of publishing stuff and publishing isn't as much, I mean, publishing is hard in terms of you're like publishing and sharing with the world, but the harder part is really the feedback piece. And that was the wonderful thing about blogs back. now but also back then was um you could see how it did by how many people showed up and they tell their friends and and i wrote a p i was starting to write regularly but like 12 people

And I wrote a piece about one-on-ones. And this was my first experience with viral. Because I don't know, 50 to 100 people showing up a day. And suddenly it was like 12,000. And it's just a flood.

They just all show up and they're putting comments and being happy with you and being sad with you and all that sort of thing. But more importantly, it taught me kind of what the things are that sort of resonated with folks. So I was writing about stupid stuff, but then I started writing about leadership.

And that was really resonating with folks, especially engineers who really didn't have a playbook. You know, there's a lot of MBA talk out there and books that sound good, but they're not really for engineers. So I thought I found that sort of... speaking, writing about leadership for engineers resonated. And it resonated not because I thought it was a good idea. It resonated because everyone showed up when I started writing about one-on-ones or public speaking or any number of topics there.

But I think the thing that keeps me going is that the feedback's great. Blogging's changed. Social media showed up. That's a whole other thing as well. um but for me it's a way of writing as a way of crystallizing my thoughts it kind of a thing i always say is like writing makes me smarter about the topic that i'm writing about because i'm anyone can just blah blah blah have an opinion but i take

I just heard this from a Johnny Ive video from last night. It takes opinions and kind of makes some ideas because they're two different things. right ideas are like concrete things opinions are like um i don't like red walls or whatever the heck it is right ideas are these concrete things and i think it takes some and it takes opinions that i have or feelings that i have

and turns them into ideas so that's that's why i keep on doing it because it feels like it makes me smarter or just more considered or or you know more ideas about the things i care about hey Before we continue I want to tell you that today's episode is brought to you by Swormia.

the engineering intelligence platform that serves some of the best software companies in the world from startups like lovable and superhuman to scale apps like mirror and webflow all the way to fortune 500s by combining data from the tools your product and engineering teams are already using, Swarmia gives everyone in your organization visibility into where engineering time is going, the progress of cross-team initiatives and the bottlenecks that are slowing teams down.

So what makes Wormia special is that it combines proven software development metrics with feedback from developer experience surveys. It's a platform engineers actually want to use, not just another monitoring tool they have to tolerate. And instead of throwing you into a bottomless pit of data, Swarmia automatically identifies patterns and delivers actionable signals to you, complete with practical steps you can take to improve.

So get 20% off your first year by booking a demo at swormio.com slash refactoring. Yeah. I love this answer because, uh, when I speak with, uh, writers or people who write alongside the normal jobs, they all say the same thing that there is something that makes them smarter and then clarifies their thoughts.

And so I wanted to follow up by asking, so how has writing, your writing work has been good for your personal and professional growth? So not just making you smarter, but maybe has helped you on the job. Yeah.

I haven't had to write a resume in like a decade because people know who I am. So there's a certain amount of nerd fame that came with it. So that's both the topic and just luck. So I haven't really had to.

It's sort of a known quantity now, which that's a selfish sort of thing. I think it also... i have a lot of questions thoughts you know you can you i sort of repeated the prior answer but like i can say one-on-ones are good but i can tell you deeply yes and when excruciating detail why they are and by the way it wasn't just that one essay it was sort of deeply considering a thing for years and years

And even flip-flopping on it. Like I wrote, I don't know, 15 years ago, hey, managers, stop coding. This is super bad advice, like really bad advice. Because I think you keep on, you got to keep on, you got to learn about your craft all the time. And I discovered that video. discussion and thinking about and that sort of thing so like long consideration of ideas is a uh is a thing i get out of that and i also just think it's um i mean this in a positive way it's just sort of therapy to kind of

especially in this day and age where it's just sort of like you feel helpless about whatever the heck's going on in your state, country, city, whatever the hell. And it sort of gives you an ability to be empowered about. your thoughts and what you not necessarily action but at least consideration which could turn into action and consequence but to me it just gives me a grounding in terms of of what i care about and what i can sometimes do about a thing

Yeah, I've loved this answer. I mean, I write a newsletter full time, so I get a lot of questions by people about. advice about how to write more or whether writing is good or or useless for their career for themselves and so you have kind of answered this already but do you think i mean how much of this value that you mentioned

came from, let's say going the full distance as you have written books. I mean, you've done a lot of writing work versus, uh, you, you would have gotten a lot of value even by just writing casually, like, I don't know, one article per month or just as a hobby. That's a good question.

I like the line you said, finishing, like doing a book, right? I mean, you can just sit there and kind of write 100 words every day, which is a great habit, by the way. And if anyone asks me how to write, I say write 100 words a day. So many days you can do it. believe it or not you'll end up with a book um and everyone goes what i'm like it's hard work gotta do it a lot yeah so um and that's the piece i think about books and whatnot is my books are you've read them are sort of

mostly aggregations of things I've written online with some unique stuff for the book and maybe a different edit or different sort of layout or kind of order to the pieces. But it's really just the essays kind of combined together.

I haven't written a book book where it's like beginning, middle and end. And, but even that is, I've got two going on right now and I'm late on one. There's a. set of lessons to be learned by finishing a thing especially a hard thing and books are hard even with me sort of spreading it out over two or three years like right now stress is high because i'm

i'm i'm sitting here staring at like needing to write about 15 chapters in the next three months that's a crap ton of writing yeah um and but when i get done with it and i can there's something about holding it in your hands i know people like kindles and your ipads and those are great but i'm a book reader i like

Paper. I like holding it in my hands. Yes. Yeah. And to me, like they're on the walls here. I have the covers from the books here. So it's also sort of a sense of accomplishment and achievement, if you will. And that's, I think that's just basic human 101 stuff of like. feeling like you've done stuff so writing casually and great i do this all the time but there's also this sort of big moments in your life where you've accomplished something that i think are just

I think our brains are wired. My brain is wired to like doing that. And it is hard, like hard, hard, like miserable hard at the end. I don't want to, I get done writing a book. I don't want to write ever again because that last. two weeks, I'm just like churning out what I think is garbage. I've done some amazing things during that time and I've created some garbage too. But that stress, that deadline, whatever that is, I think is really important to kind of...

Building your confidence, building your skill set, and also just your kind of joy as a human. Yeah. So it's like when you run a marathon, at least that's what they tell me because I've never done. Uh, but when you, when you finish, you never wanted to do one ever again, but then like after a week you want to do another one. It's like run as high.

yeah i i get so this happens every book i get so tired of the rand's voice in my head because he's kind of a little snarky and he's a little pithy and he's a little and i just i i hate him i really do because he's just i kind of have to conjure this voice that i really like by the way but like you know in concentrated

form. It's just, okay. I don't ever want to think about this word, this guy again. So, um, but it's just part of the process. Is that voice in your mind, like a slightly different voice than The one that shows up at work? Yeah. I mean, I don't know. That's a great... Someone should think about that. That's not me. Because I do have a...

This is another thing I think about doing a lot of writing is you kind of find a way that you like to write. Like I was just writing this piece I just published recently about AI. And I was getting to this point where I was just like.

angry about my topic and I started like three words period three words period three words period just to be like hey you know i'm making this point and like by the way you pass that through grammarly and grammarly is like trying to fix it i'm like don't touch my shit like like and the grammarly does great things i'm a horrible speller i like help me with you know ands and semicolons and whatever

me too but there's other parts which are very much me and very much totally not what the robots are expecting which is the point um is having that sort of voice but that's in my head i can kind that's that's the voices

this guy who's me obviously but like he's he's kind of got a different way of speaking yeah and i read what i write and then i see how i speak like i do public speaking they feel like different people to me but i mean it's all the same person i just have a personality disorder or something Yeah, I love this. It's complicated, but I can talk to the rest of it.

I want to talk about some of this writing work, especially managing humans was particularly special to me. I mean, I've been writing this newsletter for about four years. And I remember when I started thinking about writing like a regular newsletter, it was like... 2020, uh, completely, I was thinking not only about the topics, but also the style, the tone of voice and many things, you know, that writers think, I don't know. Uh, and luckily, I mean, completely by chance, uh, in that time.

I also read Manage Humans and it blew my mind because it was not the... the first book on engineering management and these topics that I had read. I had read many that I loved, like the manager's path and others, but it was different in a way that, I don't know, it didn't shower me with frameworks, but it was. more about giving me a taste about what real engineering management looks like and and it kind of said hey this thing it's a mess but it's okay we're we're in this together so yeah

give me this kind of vibe. So did I get this right? That is, I mean, for us engineers who thrive with systems and controlling things, is this management thing instead about...

ambiguity and chaos and this kind of stuff yeah it's i mean there's this is a you just asked a really good question and there's it's a complicated answer number one the book is a reflection of me and i'm an engineer but as you can read i'm also i have the empathy thing i have the the people skills thing i had a you've read this but like i had a manager at netscape and and he was like got a read on me in like three months he's like hey lop michael whatever your name is um

just people thing your ability to read the room to understand how people work and all this sort of thing you think this is um you think this is boring because no one else knows how to do it because no sorry you think it's boring because you know how to do it and it's just obvious to you yeah and he's like you're a good engineer you're not the best engineer i've ever met but this people thing is you're world class at

And that's the thing you need to lean into because we lack that a lot in high tech is that sort of that people side of the equation. So that's why I lean into less process stuff and frameworks and whatnot and more just sort of. a way of describing things hopefully artfully colorfully and um in a way that um that is appealing to folks as opposed to showering you with frameworks and this sort of thing but i don't

Those frameworks and that process piece are really actually super important. I'm just not that good at it. I mean, I can do it and I can tell you what to do and blah, blah, blah. But more of what I'm likely going to do if we're working together, I'm going to say, hey, can you go talk to Julia? Because her process skills, her execution skills are 10 times better than mine. And I delegate and I give that to her.

it's it's i i'm glad you like the book but the whole being a good leader is this whole set of skills some of which you're good at and some of which you're bad at and um you have to um i'm glad it appealed to you but it's not the only way you need others you need other people to actually be a successful leader is what i'm saying yeah of course of course but so so so you said

you had the skills, but maybe you were not even aware of that, that you were not aware that you scored high on these skills. And did you know what you were getting into when you started like... managing people or running a team at netscape no it's just it was It's horrific. I always ask this when I do a talk. I say, how many of you, usually leadership talks, right? I ask the room, how many of you actually had training to be a leader?

I'd say maybe 30, 40% raise their hand consistently. So no training, like 60%. The majority of folks have not had any training to be a leader. Seems like an important role, right? So maybe that's gotten better. But it's just, it's such a, and it's weird because it's such an important thing that why are we not really doing a ton of work to build these folks? It really seems that it's sort of like throw people in the deep end and like pray.

that you're like, is he or she good at this sort of thing? And that first couple of years, three or three years, dude, like this is before I was early writing, just horrible mistakes, horrific mistakes, like career defining mistakes.

eroding trust just just bad right so but i'm learning you know and then it's actually one of the reasons i started writing i'm like hey i think i think i figured this out and i wanted to tell somebody um so that's where it started is that piece but this it's such an industry is such a we don't do the training bit and maybe that's okay maybe part of this i really and this is maybe one of the reasons you like the book

is like doing it is actually the best way to learn. And the, my book is like, here is sort of a journal of a lot of, a lot of that sort of stuff. So you can see that glimpse of it and be like, oh gosh, because your work situation is different than mine, which is different than hers. So, but. getting that data and sort of, you know, cross pollinating, I think is important to kind of that early learning. Yeah. I think like many people have this notion that these skills are kind of.

like magical that you can only learn them on the job. Right. So, but you think instead there is a gap in the sense that we should be more intentional and provide training to fresh managers. I mean, there is more that we can do about this. I think I don't. Yes, kind of. I think it's more about having a mentor because I really do believe that, you know, the battle tested way of going through it is a good way to do it.

and i don't think sitting and going to an offside or sitting in a conference room is like gonna or reading maybe reading a book um depends on how you learn but i really i think the one that really resonates with me is i'm a as a new manager who's that person that's not your manager, who's sort of your mentor type, who you can...

slack or message or have coffee with or whatever and ask him or her like hey what's going on here because i that's that's kind of similar to the book i think in terms of a little more real time and a little more personalized to you how you kind of kind of

course correct as you're going through these really bumpy couple of years. But it's still, I mean, make it sound like mentors are like your first five years. Your entire life, these are people, they're valuable. But I think that it's really critical early on is to have. that buddy, someone that you can talk to because you just, you're going to screw up constantly. Yeah. Yeah. No, I totally agree. So one thing I wanted to ask is that we often say that.

I mean, frameworks are important, but there is no like a conclusive unique framework about how to lead teams and humans. There are many styles and different types of leaders and some might be more. calm, empathetic, care about relationships while others, I don't know, they break down walls, but they're also abrasive, controversial, Elon style, you might say. So do you have a style yourself? How would you define it?

I mean, you can already infer it a lot from talking to me and also the books and stuff. I'm sort of Mr. Listen. I'm Mr. Kind of... understand the the situation that i walked into and i'm less sort of come in there bowl in the china shop hey everybody this is what we're going to go do and

By the way, that's delightful for me, but I've gotten feedback, especially as a senior leader, as a VP at other places where they're like, you don't talk a lot. And I'm like, that's because you're talking all the time. which is a snarky passive aggressive response. It's really hard for me to actually, the most recent book that I'm, new book that I'm writing right now is sort of around sort of being a lot more proactive, directive, more, you know, like stand up and, you know.

valley the troops type because it's not my natural state i'm a fine leader but mine is sort of like listening to okay i got four people here what's going on here okay who needs what here what do i need out of this situation how am i going to gently shape this and push it in a direction i think a lot of time when people are telling me hey you're kind of quiet it's because they're pretty loud and they're also looking for like big huge sort of like

PR events about things my teams are doing. I don't care about the PR. I don't want the recognition. I want us to be successful and I want us to get things done. I think a lot of, especially in tech, there's this sort of hero culture of like, oh, I saved the day. I'm like, good job saving the day. Why did the day need saving? Yes. Like we screwed up and you had to save the day. So good job cleaning up your screw up.

yeah it's you know it's so like this is a thing a line i say a lot is like nobody claps when nothing happens right and a lot of my style is sort of like you probably hopefully never going to notice it because people are just getting their work done yeah and we're not we're not flashy and we're not generally you know driving off the cliff so um it's a really easy way it's a kind of a clever way to do so i'm doing a lot you don't notice

But I think that's the case, is that a lot of the work I do is in terms of leading, guiding, delegating. whatever is pretty light touch, even though I have clear things I want to do and I know what a healthy team looks like. And if I need to step up and save the day on the blah, blah, blah, I have many playbooks to get that done.

But that means I screwed up earlier and I prefer not to screw up earlier. I love this definition. I love this answer because in my view, like good engineering is very often like boring. engineering boring stuff right when things are exciting chances are we are kind of on the wrong side of things i mean you've got to you want to have some excitement yes you want to figure out how to push the team and to do new things

100% and and when you do those things the probability of it being harder to do than something which is you know obvious is much higher so but even even with the innovation even with the pushing of folks people good leaders who've done it before know like okay cool this is not standard operating product that we're building. This is going to be a stretch. What does that mean? Means that any estimate that you give me, I'm going to double in my head, maybe triple.

And you're going to go, no, no, no, I can do it. I'm like, yeah, you're not counting the other 17 people you're working with and the interrelated cost of doing that sort of thing, which isn't your job. That's my job. So I'm going to buffer a lot more for something which is a lot more complex so that we have more time to deal with the things we can't predict right now.

And so we don't have those explosions and sometimes the explosions still happen and that's fine. But I know how to deal with those too. But they are to me always bug reports, always things that I'm like, okay, cool. Why did I not?

I've done this a lot. Why did I not predict or account for whatever the explosion is? Yeah, no, I agree. And when you think about your style and you were doing things, how much has this evolved over time or is this mostly driven by your personality your your your character yeah yeah yeah it's a good one i mean i think i'm i think i'm probably majority

This is just who I am. And it's sort of a comfortable way for me to have my leadership style. Let's call that 60% of it. There's another 40%. And this is the hard part, which is 100% the opposite of what I. how I feel. I'll give you a really good example. Like feedback, constructive feedback for every human, I think is hard, right? When you're scurrying up and I need to tell you that you're scurrying up.

i feel bad you know you're screwing up when i start talking about it you know what i'm about to say and then you feel bad And so what usually happens, what mostly happens is people avoid actually doing it early until it's actually a four alarm fire. And then it's really bad. So back to, you know, no one claps when nothing happens. I have learned that to tell you that it's a screw up way before it's like even a bad one. Like I say, Hey, did you, you, you.

said this thing in this meeting i think this means you mean this is that what you mean yeah okay great i've seen this before and let's not do this because of this and this and you're like i was kind of just yolo you're like a little put back because i'm sort of nipping something in the bud that is really not an issue yet

It would have been an issue in three months because I could see what your intent was or your strategy or your desires. And I'm giving you that feedback early. And you're a little pissed at me because I'm like slapping your hand. I'm not slapping your hand, but I'm like giving you this feedback. way before it's an issue. So you don't think it's a bad thing. You go, okay, great. No problem. Whatever, dude. And then you change your thing.

nothing happens that's hard because that's kind of like i know that i'm kind of being a little more stricter or a little earlier about giving the feedback and that's yeah there's other things like that which are just learned behaviors that if i was just kind of coasting and not working i'd be like oh whatever you'll that'll blow up in your face later not my problem um so i have things habits skills playbooks that i

that are countered to kind of my natural state, if you will. Yeah. I love that you mentioned feedback because this is a very frequent comment and thing that people do not get right. I feel like it's early in their careers in management. Do you think that... I mean, when it comes to people who get into management, there are some like usual suspects of people that get wrong more often than others, or it depends on your...

uh on their characteristics i mean even this feedback thing it was because of how you are as a person or is this like a common thing that most people get wrong it's really i mean everyone's sort of different but like

The greatest hits of sort of new managers are feedback. Another one that is a classic is delegation. um engineers are particularly bad at this because we really like building stuff with our hands so that idea of giving giving up their legos and giving them someone else on the team is hard hard hard um it's more of a manager of managers things where you really have to fully delegate that piece. So that delegation thing is sort of, is one of my

It's like when someone becomes a manager, I'm like, are you ready to feel like you're not doing anything anymore? You're doing stuff. But it's going to take a reframing of what you consider to be work.

um to actually do that as opposed to like closing bugs or writing code or doing whatever that is so there's that delegation piece which is someone have to get over and that's just that's like like if i had to pick one thing that's it The feedback piece is definitely a one, but it's a larger sort of...

bucket of like performance management which isn't just like when it's bad it's also like cool i need to get you to the next level so here are the three things you need to go do and you understand that and you say no i don't understand i mean like great let's have a conversation about that so Performance management, both the positive and the negative versions of that is another one that is there. Really, really tricky. There's a couple. There's a bunch like that, though.

Yeah, love this. Feedback, delegation and performance, man, definitely good estates, I think. At least when, if I think of my own experience, I agree pretty much. And so I asked you how you personally changed and how you learned things throughout your career, but how do you think instead engineering management as a, as a whole has changed in, in the last, I don't know, 20 years. I mean, if you, if you think, for example, at the things you have.

have written even back in the past, uh, that you can, are there things that you go back and look at them and you think this is just not true anymore? Well, I mean, I said they got one by now, which is sort of, you know.

engineering leadership not coding or not kind of being a very deeply aware of the technology kind of flip-flopped on that i think that's incredibly important and that's I think that's changed over the years. I don't know, like it's for every kind of Ranzian sort of team that is small and tight and not a huge amount of direct reports.

the org is pretty flat you've got googles of the world that are ginormous teams that are many layers or this sort of thing and you can't argue they're not doing i mean i don't know how they're doing but like you know there's lots of different styles that work so um

I would never do that. I like small teams with empowered engineers generally building consumer software, so things I can understand and not necessarily enterprise software. But it's just, like you said earlier, there's just so many ways to kind of... get it done the directive jerk types are all over man and you know and i'm not a fan but you can't say there's not a lot of these folks that are actually what for whatever definition of success you want are really successful

And maybe that probably means there's a me hiding in the, a me type person hiding in the hallway who doesn't, who's kind of like smoothing it over and making that jerk more palatable. i am a fan you can go look at my career i am a fan of sort of strong product ceos and and you know whether that's slack or pinterest or my former current employer um i mean apple before and currently um i like that i think it cuts through a lot of the bs when you have a really talented leader out there kind of

um kind of cutting through and saying this is what we're doing i love i love that you know um and sometimes it takes that's a different personality than sometimes i am and i those folks are essential there so it takes it's really hard to how's it changed to be blunt i um it's moving a lot faster um that's what's definitely going on which is making it a lot harder i think for teams to kind of

Learn from there. I was listening to this thing last night. It's going so fast right now that we're not really able to. adapt and kind of learn what the new playbook is as quickly and that's yeah that's a challenge that's not a that's not a function of how long you've been in industry that's just like how quickly things are changing here and it's it's it's fascinating but it's also just like

Dang. Yeah. It feels like, I mean, technology and the things that you can do are changing way faster than companies who are, I mean. like small societies and people systems can adapt. to this so yeah i think there would be a latency like a lag between the how you can operate at your best i don't know and how your company is functioning

which has still not adjusted. I don't know. Yeah, I agree. I agree. And you mentioned, for example, uh, a lot of, um, layering, big tech, Google, do you think we are we are getting into a future where we are doing without a lot of this layering, getting leaner. There's a lot of talk, of course, with AI that you need less middle managers, less layers, less overhead. Uh, yes, I think

change is coming. That's correct. But I think the type of change that I like, the type of teams I like and the changes coming are very aligned. I truly believe that a manager or leader is capable of managing about seven plus or minus three folks a year. quoting myself here. So like more than that, and you do not have time for humans. Now, if you don't think that humans are, you know, important to grow and lead and manage, then you're going to not listen to me. You should hang up right now.

um so that that is a thing is that i i am a function of my success is a function of the team success and i grow by growing my team So, and I can only do that with seven plus or minus three and that apply the plus or minus is for a brand new manager versus race season manager. But like if a team, one of my teams has 11 folks, I am in my math head going, one person is not getting.

uh the leadership attention that they need one if it's 13 i'm like rattle fire if it's 15 i'm like we have a huge disaster here um that contradicts other leadership organizational styles so But the other part of that is sort of the layering from the CEO and the individual who is working. And I try to keep that in the places I've worked, obviously not Apple.

because it's a big place. But everyone else, try to keep three layers from Bob, who's a brand new engineer, and Julia, who's the CEO. Three layers. So that's Julia. VP director manager, three layers. And that doesn't, that sounds like a lot for a startup, but go look at.

Chase, go look at it. It's like just layers and layers and vastly trillion dollar businesses there. But anywhere that I'm working, I'm trying to keep it light just because I want that communication from up and down is a lot easier to get done. And it's less error prone. So that's just how I build teams. And it's worked out well. Yeah. One of the things I love about this answer is that basically if I get this right, you said it's not like AI is changing.

uh, what the ideal organization or the ideal team looks like in terms of shape and size. It's more like, uh, it is making what was already possibly the best type of team, the even bester. yep yeah yeah i i mean my version i agree with that but like i've been doing a ton of work with ai i've been using all of them and just kind of getting a sense of sort of the art of the possible um the

From a zero to 60, from a prototyping and idea thing, I've never seen better in terms of just like, I just kind of want to riff on this thing. I don't have as much data about how it's going to do well about an existing ginormous code base. that thing but as a as a prototyping tool but more importantly as a research tool in terms of asking questions saying why is this i'm like i've forgotten how this shell skip works just explain this to me again it knows it always knows

And I think that's an accelerant. I think that's an accelerant. And I think that the engineers and the leaders who embrace that, well. This is going to come back and haunt me in like 10 years. We'll save this clip. No, it's okay. I really do believe this because I've just been seeing it for 30 years. When the Apple II came out... There was a whole backlash of people, probably accountants or other folks that were like writing, enjoyed writing things down on paper, who were super, the Apple II.

No modem, like nothing, no internet, nothing, just a device to a computing device of the simplest form, not the simplest. It's very complex. So. that's the same thing that's happening right now and i think there's a group that's going to be like i'm cranky about it and there's gonna be a group who's gonna see it

as an accelerator. And I see it as an accelerator right now. I've done more coding in the last month than I have in the last 10 years. Like we're talking thousands and thousands of lines. Important to note, I didn't write any of it. like um but i did work and now i have this tool that's doing this i have this tool that's doing that and it's teaching me things like i was this is this is dumb but i was doing some facial care stuff and i forgot what i bought so i took a picture of the little

face cream thing i said what is this and it's like this is uh the stern thing and would you like me to tell you what order to do these in i'm like yes i would like you to do that and now everything's all arranged in the medicine cabinet and like like no ads Talk like a human. I asked a human question, got a human answer, and I got huge value out of it. I challenge you to do the same thing inside of a Google or whatever your search engine is. That's...

Forget about the acceleration, just the knowledge assistant, that ability to help you answer questions and wonder about things. That's good for us. Usually when I say this sort of impassioned, foamy at the mouth thing that I'm doing with you right now, someone goes, well, how do you know it's not lying to you? I'm like, how do I know that anybody's not lying to me? I need to have that skill.

24 seven because people lie to me all the time and i need to know like oh you're trustworthy you're not and if you are i don't trust you please source tell me why you believe that thing or what is your source on that thing it's exactly the same thing it's just a robot that you're probably scared about which is why you're ripping on it and by the way

It makes mistakes. Welcome to the planet Earth. We all do. Like it's the most human thing that it does. That's a good line. That's good. I think because I think that many people like intuitively compare. the like the qualities of the these ai systems to other computing systems that people always had and yeah one quality of software is that it generally it's even like it's either like 100 correct

right, or 100% wrong. While this is not a good metaphor for AI, which is probably a good analogy, it's more like, as you said, a person sometimes is right, sometimes is wrong, makes mistakes, but it's hard. maybe to see things that way i don't know well and by the way it's there's this really important skill when with both humans as leaders and with now with robots is

You tell me something and you say, hey, I believe this thing. And I, uh-huh. Tell me why you believe that. Uh-huh. This, this, and this. I'm like, uh-huh. That's interesting. Did you think about this? You're like, no, I didn't. That updates my model. I mean, sorry, that updates my priors on that. I'm like, great.

We just learned something together. I was sharing my experience. You were testing a thesis with me. We debated it a little bit and you went back. I've had many interactions with AI over the last three, four months where I'm like, cool, generate this good for me. Okay. Huh. why did you do this and it's like oh yeah you're right i i thought you wanted this to be spaced and not this sort of thing and i'm like yeah right like okay cool thanks and then it fixes it

It generated bugs. It didn't do what I expected. And we had a conversation and it didn't learn because it's not, these are not learning models yet. But it learned in the context of that one conversation for me. There's a skill there about me going, hmm, is this bullshit? Is this real or not? And like looking at it and verifying it and testing it and going to say, oh yeah, you're right. Or no, this is wrong.

It's a normal thing. And this folks who are worried about their jobs being replaced, again, this will blow up in my face in like 10 years. Like, I think they want the quick hack. They just want to say, cool, I don't want to fix that bug anymore. I just want the robot to do it. I'm like, yeah, you're in trouble. you're in trouble because yes it can be an accelerant yes it can help you automate or make that more efficient you still have a responsibility to think

to ask great questions and to not trust the answer without verifying it. The human condition is... And I hope that that's actually a good thing in a world where a lot of folks are just consuming what they're reading on the internet and going like, oh, this is the truth. Like, yeah, it's really not. Yeah, not only I agree, but I think that, I mean, many of the skills that... are required of you to to to make the most out of this uh out of ai are actually similar to how you

you make the most out of working with people because it's about giving great context, great feedback, you know, being thorough about what you need and how you want things to be done. I mean, if you're good at that, chances are you can be a good work with human. as well no i think it's really you make a really um you make a really really good point which is um some of my biggest mistakes with ai is when my question wasn't good

or my prompt wasn't good. It did exactly what I told it to. It did exactly what I told it to. And you know what I did? I was too vague. Like it was, there's a specificity thing that is really interesting to watch it do. And the more that you get, there was a prompt, it was a... I think I was using Claude, and it was like, uh, write the prompt like you're talking to one of your...

a junior engineer and be super specific or something like that and that's right that's because it's kind of a baby right and yes it does exactly what it tells you and i mean it like makes leaps and it does intuition and whatever the right words are to describe it but the more you're precise

the more you're clear with your robots or your team, the happier everyone's going to be. Yeah. Yeah. No, I totally agree. Also because I think, I mean, one of the differences when you work with humans is that we take for granted how much uh, implicit context we have already shared with another person about our workplace, our team, our, the goals of things. And, uh, AI knows nothing about this. So it's not like you can condense what you need in like one sense.

sentence as you would with one person that you'd be working with for like five years. Right. So that's the gap. I think in many cases, not the pure intelligence or role power of the model. And there's another thing on the human side of that is people, this is just being a human, they don't like to feel like they don't know. Like, I feel good, right? You know?

And they often don't ask. So I, you know, I can get foamy at the mouth and talk about something. They're like, well, he's super fired about this topic that I barely understand what he's saying. Humans sometimes when they trust you will raise their hand and be like.

you mean this i'm like yes that's what i mean they're like okay that's not what i heard i'm like okay cool thank you for asking it's something i'd like to see out of um out of the bots the robots a little bit more it's a little bit more like did you mean to do this like these are these agent

tools right now or doing more of it in terms of saying like, hey, I'm about to remove this file. Are you sure about this? I'm like, yes, that's exactly what I wanted you to do. So that back and forth is for both the humans and the robots, I think is...

is a way for us to learn about each other. Yeah. Yeah, I agree. So since we are launching ourselves into predictions that are going to be embarrassing in a few years or months from now. So we have talked about things that are changing, but what are the things instead? uh that on the midterm you you think are not changing are staying the same about our jobs about team engineering teams that's a good question it's even harder to answer

What is not changing? Yes. I do not, I do not, I'm sure that we're going to, there's gonna be some shift in terms of. how many humans are necessary to get the job done in a world where you have these really interesting tools to kind of help you get started or to help you kind of go faster. But I just don't, I think there's the last 10% that of, of a very hard project or a design or an application that involves such deep knowledge and intuition. It's like writing.

I can write a thousand words and they're pretty good. What makes it great is the editing and the last 10% of the work with that work to kind of refine it to read it. to let it sit there a week and go back and read it again. And like, I don't know. I feel like I haven't done this math. Like it's maybe twice as much work as writing the piece.

And it's not like writing it out. It's like thinking about it and getting on your bike and doing that sort of thing to take it from average to great. That last 10% of... Not the work, but of the time, if you will. It's a huge amount of time. But that's the difference. And I don't think, I don't see a lot of art and flair in AI. Someone just literally mailed me yesterday.

It was the, Hey, you want to do a podcast with us? And I'm like, this is a robot. I am. I'm 100% sure. Well, I mean, they were, it was, I'm like, wow, this is pretty personal. Oh, wow. This is just like. poorly reading my wikipedia page and then they followed up today this morning and it was like hey i saw the aaron mailed you and i'm like this is 1000 a robot and because it just it was clearly it was tinny it was it was dumb um

And so, again, I think there's this last 10% of flair, hard work, blood, sweat, and tears that is not changing. And I'm... I dare to say this, I'm excited about that. I don't think we're actually seeing a lot of the things that these robots allow us to do when you combine it with that last 10% of a human making. phenomenal. That's the piece I don't think is changing right now because it's mimicking stuff. It's showing you patterns that already exist.

And you know what that looks like. What your brain really loves, what we really love is truly innovating and building that new thing, whatever that is. And I think that is going to be... uh the speed by which we do it maybe is faster but i don't think it makes the hard work go away i gotta litigate it we gotta

get in a room and talk about what makes it great and argue about that piece. And the robots are not doing that yet. So, and I also, so anyway, that's one thing. There's probably others. And when you think about this 10% that is left. What does it look like? Do you think it's more like a specific part?

For example, if you think, if we think of the development process, it's more like system design, you know, there are many predictions that say, I don't know, engineers are going to work mostly on system design, mostly on architecture and less on coding. Or do you think this is like a horizontal? thing where we are going to do 10% of everything that is just like Q&A and injecting our own creativity on all the things. I think I understand your question. I think the, the work, which is sort of

easy to describe and structured the robots are going to be great at that and that that whether that's testing or like here i dude here's something this is a quick story from last weekend i was writing this i was writing this script i was writing this i was describing how i wanted the script to be written to look at a watch site and to kind of look at it every hour and kind of scrape it and pull it down and do something there and it was screwing up

It kind of was working and it was screwing up. And I was getting increasingly frustrated with it. I'm like, no, don't do this. And it's like, okay, great. And then the second time I'm like, yeah, this is still not doing this thing yet. Here's the thing, man, that kind of blew my mind.

It, this is Claude, by the way, it said, okay, yeah, this is still wrong. It says, okay, what I want to do is I want to actually go and run the script now against the site, see what the output is, and then report back to you. which is testing by the way and I hadn't done this yet and it's very easy for us to do but it's like I didn't say go run the script and tell me what's happening it chose to do that right chose is a strong word but it was like it knew that it was

it knew that it was screwing up and it said, okay, I need a different path here. What's a good path? How about testing? I was like, whoa. I would have suggested that if I even know it was a thing it could do because I was in learning mode. That's there. It did it.

it ran it and it said okay cool lops expecting these four parameters i'm only getting these three what's wrong and it's okay let me try this let me try that let me try this it did all that and so i i think there's again i think for very structured things uh it can be very um I think that work is going to fall into the robots. And so if that's your world right now, get used to the robots. Jump in and start using them. But then there's sort of the art of it.

I have this script that's running. This is insane what I'm about to describe. And what it does, it checks like every 15 minutes and it's sitting in a terminal window. And every 15 minutes it goes tick or talk.

randomly and I say randomly say tick or talk like a clock right and I'm like okay great and then i'm sitting there going like you know it would be funny is every time that it does five in a row i'm going to call it disco and it's going to keep track of that so it goes tick tock tick tock tick tock and then i'm looking down and every time it does five in a row the sixth one it goes disco and then it saves that

now you're smiling because you know this is goofy and and just insane but you're smiling and because i'm doing a very human thing which is sort of this irrational kind of weirdness and like yes never in a million years well again famous last words like that's a very human thing yeah to be like putting a little bit of joy or humor or something into something that unexpectedly right

And I think those are the things, whether it's writing or coding or whatever you're building with these things, that's the human aspect of you kind of pouring a little bit of your sort of. uh error prone insane self into these things and when you look at great products i think you see the care but you also see a little you see a little bit of humans and they're going like

oh, this is kind of funky. Like, you know, it doesn't feel like a template. It doesn't feel like boilerplate. It doesn't feel like enterprise software. It feels like something that is human. And that is that last 10% we were talking about. Yeah, I love this. I agree 100%. I think also this is a perfect closure and parting thought. So thank you so much for this chat. I had a lot of fun, took a lot of notes. Thank you again for coming on the show. Absolutely. Great.

You did great prep. That's unusual in podcasts. So good job. Thank you so much. It means really a lot. Thank you. Thank you. Thank you so much for listening. If you found this chat valuable, you can subscribe to the show on YouTube, 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 show. You can find all past episodes and learn more about the show at refactoring.fm.

See you in the next episode.