34: Ep. 34 | The Art of Storytelling in Leadership with Matt Sinclair

Matthew Sinclair (00:00) Humans are storytelling apes, okay? And so in the absence of a good story, they’ll make up their own narrative, okay? And so what you have to do, and this is increasingly true, sadly, in a post-truth world which we seem to live in, there’s lots of narratives out there and people can pick, they’ve got a lot to choose from in terms of the stories. And like I said, if they don’t have a good one, they’ll make one up.

So the job of leadership, the first job of leadership is to explain why smart people that they’re working with should get out of bed in the morning and come and play their game, whatever their game may be, right? That’s the first job, tell a good story. And then the second, and if you can’t do that, the smart people have got plenty of options, they’re go and do something else. And so then the second thing is peers should collaborate to work out what to do, not how, but what. And then the trick of good management is to get out of the way.

I think and let smart people work out how to do the things that they’ve hired them to do.

Brittany Ellich (00:52) Welcome to the Overcommitted Podcast, your weekly dose of real engineering conversations. I’m your host this week, Brittany Ellich, and I am joined by…

Bethany (01:02) Hey, I’m Bethany.

Erika (01:04) And I’m Erica.

Brittany Ellich (01:05) The three of us met while working on a team at GitHub and quickly realized we were all obsessed with getting better at what we do. So we decided to start this podcast to share what we’re learning. We’ll be talking about everything from leveling up your technical skills to navigating your professional development, all with the goal of creating a community where engineers can learn and connect.

Today we’re sitting down with Matt Sinclair, an Aussie software engineering leader and former partner and VP of engineering at BCG Digital Ventures, who’s dedicated to building the digital future. Matthew’s journey started not with the Commodore 64 he craved, but with a Mercurial TI-99 that unexpectedly spurred his passion for programming, leading to a career that includes building everything from high-performance payment systems to shared mobility.

platforms. We’ll dive into what it takes to anticipate and build the future with BITS, the intersection of technology and business strategy and his passion for coaching the next generation of engineering leaders. Welcome, Matt.

Matthew Sinclair (02:02) Good day, thank you, that was a lovely intro. Thanks, Brittany.

Brittany Ellich (02:05) Absolutely. So we love to talk about tech that people are passionate about. And it sounds like you’re a big advocate for Elixir, calling it the most future-proof and AI-aligned language. And I would love to hear more about that.

Matthew Sinclair (02:19) You’re quoting my clickbaity headline from a post I did a few weeks back. Look, what can I tell you? I’ve been around for a long time and I get very frustrated with religious arguments about things. And one of the best things about Elixir is that it just answers a whole bunch of questions that you don’t have to re-litigate and debate with people out of the box. So I’m an old functional programmer from way, way back. did Lisp and Scheme at university.

40 years ago, maybe 35 years ago, whenever that was. When, crikey, when was 1990? Yeah, 35 years ago. So when I, and then I know I was a Java programmer like everyone and then a Ruby programmer and I kind of got a bit disenchanted with programming a little bit and the further up the leadership hierarchy you go, the less, the further away from software you get often. And,

Anyway, and so I came across Elixir about 2018, 2019, I think, and it was a revelation because I was just like, oh my gosh, this thing has everything in the one box. So you don’t have to make all these tech stacks decisions about Kubernetes or Redis or, know, a bunch of stuff just comes out of the box. And so I started tinkering with it and I wasn’t really writing anything serious because I was still at BCGDB. But I just kept plugging away at it and kind of…

after probably 10 years or so, or seven, eight years of not really writing code, I felt like I needed to get my gym fitness back from writing code after having written a lot of code in the past and then not writing much. And I just got into it and got more and more, fell more and more in love with it. has a, it’s kind of like this, as software engineers, you know that thing when you write something and it kind of feels right, you know, it’s got this sort of elegance to it and you know, it just kind of clicks. There’s a, I don’t know what that is, it’s very hard to put your finger on, right? Like when you write something and you crack it up,

that really feels great. And Elixir’s full of that kind of stuff. It’s just very joy inducing, which I know sounds crazy, right, for a software programming language, but it’s…

when you spend so much time on a daily basis writing something, if you have to battle with it the whole time, I think it becomes exhausting. And at least in the languages that I know, that’s the one that has the least, it’s the least exhausting that I’ve come across so far. I mean, that’s the conceptual, aesthetic version of it. But from a very practical point of view, it’s very fast, it’s very complete. The folks in the ecosystem are great.

I don’t know what your preferred programming language is, but one of the things I learned over the years is there’s a very strong correlation between functional programming and high quality software engineers. I don’t know what that is. And I don’t know whether it’s just because functional programming’s got a bit of a steeper learning curve maybe to get into it, particularly if you’re coming from OO. I think if you come to it naturally first, it doesn’t have that at all. But if you’re coming, you have to sort of unlearn a bunch of stuff about OO when you go to a functional programming language. So, you know, it’s…

I don’t know, a bunch of stuff that just makes it fun to use. I don’t know. I get a bit evangelical about it, which I know is ironic considering I was suggesting that I don’t get religious about things. yeah, yeah. and there’s a really good, there’s really good.

use cases of Elixir being used at scale that people don’t really know about. So Discord and WhatsApp was actually on Erlang, but it’s the same underlying VM. And a few other ones, and there’s a few hedge funds that use it, I know here in London, and they keep very quiet about it for two reasons. One is because it’s so productive and they would prefer other hedge fund.

engineers to be using less productive languages for obvious reasons, but it has a kind of ecosystem lock-in thing from an employee point of view as well because it’s a niche programming language, there’s no doubt about it. But the folks who do use it love it and so you want to find other elixir shops so you can make use of it as well. it’s got a bit of a, maybe it’s a virtuous cycle I think from one direction to think about it.

I don’t know, other than that, I’ve spent the last nine months, what is it now, August, November, so 11 months writing software on a pretty full-on basis, whereas I had the previous 10 years, I wasn’t really writing software, and I think I’ve written more software in that time than probably in the previous 15 years of my life, just in terms of productive output. So yeah, I can’t speak highly enough of it. Not everyone’s using it, I think it’s in the kind of 5 % range,

it’s the most loved and Phoenix is the most loved. So when people who use it vote, say you use Python or Ruby or whatever and you say, how much do you like it? It comes in at whatever. The Phoenix, Elixir Phoenix.

folks using it seem to really love it. So, feels like it’s a good thing.

Brittany Ellich (06:42) That sounds…

Bethany (06:44) I used Elixir for a couple advent of codes and I had so much fun with it. It really is a fun programming language. But I was curious if you have any tips for folks who maybe never learned functional programming formally and how they can start learning or understanding that pattern.

Matthew Sinclair (06:50) There you go.

A few things clicked for me, like it was a long time since I’d used it from back at the university and as anyone who is a few years out of university, you sort of go in knowing nothing and sometimes come out of university knowing even less. So I wouldn’t say that I was an expert by any stretch of imagination. But after maybe say 20 odd years, 25 years or so of programming coming to it, a couple of things really just clicked.

for me, immutable data structures, I mean, I’m gonna get really nerdy, but immutable data structures and pure functions save you from about, I don’t know, 50%, like 50 % of the kind of bugs that you would get in normal code just aren’t possible with that. Now, it’s not that you don’t have any bugs, of course that’s not true, there’s plenty of bugs, but pure functions.

you can just test them in isolation. And with immutable data, you know that you’re not gonna get side effects. And so those two things alone, would sort of, I’d make the point that’s like, trust me on this, this is good. And once you get across the river and you realize, yeah, okay, I see, I see. So that’s the first thing. Those two things just improve productivity enormously. But the other thing is this chaining of functions, composable functions. once you get a…

you make a data structure and then you make these functions that do a little bit of mutation on the data structure and you chain them together. Once you get that kind of thing working, it’s super cool. And there’s a really non-obvious unintended consequence of all those three things is that it makes it very easy for a coding agent to reason about the software way, way easier than in mutable languages. So part of that productivity boost that I got.

This year has been because I’ve been using Claude code with with elixir and it’s a great combination a equals Claude’s probably the better I think the better programming Foundation model at least in my experience But it’s very good at reasoning about pure functional code because there’s no side effects So you can kind of you can basically I wrote a sub agent for Claude called elixir doctor and I got all my my coding standards and every now and again when I’ve got nothing to do I’ll go elixir doctor just just

do your worst on this module. And because it’s, you can change the internals of a function. As long you don’t change what comes in what goes out, it’ll just fix it. And so you can turn really, really terrible nested if-then conditionals, which is what Clued will do on its own if you just let it go. You can turn it into these really pure functional chained.

Chains across across the data structure anyway, and I don’t know the first time I did it I went I don’t really understand what I’m doing and then it kind of clicked and I went oh, Okay, now I see and I see everything now in these data structures and pipelines of functions So I don’t know whether that appeals to anyone, but it kind of it’s a it’s a very Fulfilling when you get one of these things written and it works pretty well and then the because of the the LLM interaction

I mean, I feel like I’ve got a team of five engineers working for me now. And I sit here and I write code and I’ll sit there and I’ll rub a duck of design with Claude there. And you end up spending way more time upfront, I think that I probably would have in the past in terms of design, but you’re not something out. And then all of a sudden it’s just writing this code. hang on, hang on. No, no, no, no, no, no. Don’t do that. Do this. And it makes some silly mistakes from time to time that you have to rectify, but at the end of it, you can kind of have at least two.

often I have two Claude sessions going, operating on these. I wrote this thing called Intent, which is a bunch of little helpers to help Claude, to allow me to help Claude help me. It’s just a free, what do call it, kind of open source thing I put on GitHub. And what it does, I have this concept called a steel thread. And so a steel thread is like a piece of functionality that I can describe.

What I the reasons called intent is because what I’m doing is I’m capturing the intent about a piece of software and so if you think about when you’re writing software you often don’t capture the intent or you or you you might capture it a little bit in comments, you know you don’t really do it, but doing it in a systematic way with a steel thread and then a bunch of work packages in that steel thread what it does is it stops context explosion so you can you can you get this little chunk of documentation and You can then come back to the problem when you reset your context when you come back to the problem and you just have to ingest

the steel thread and it gets back up to speed. And then what you end up doing over time is you sort of leave this trail of intent, which is super interesting. You have these steel threads and you complete them and so on. But at any point you can go, oh, I wanna pick up steel thread 17, go and read that and come back to me and tell me, right? And then you put it in ultra-think mode, you say ultra-think this steel thread and then come back and tell me what’s next.

And rather than have to relearn the entire code base every time you turn it on, which just doesn’t work. And I think when you hear people’s frustration when they’re talking about coding agents, a lot of it is around, the context we do is not big enough or I can’t, I spend all my time trying to keep it on the rails. What intent does is it just helps it help, sorry, it helps me help it keep it on the rails so it can be much, much more effective. And so these combination of things, like it’s kind of happy accident. I didn’t set out for that to be true, but Elixir was fun.

I started using Claude and it was a disaster. the first, the early part of it, I did this, I was writing this thing and I reckon I wrote 200,000 lines of code of which 150,000 were complete rubbish. And something occurred to me actually a little bit later on, I’ll go a bit about it, but what occurred to me a bit later on was I was like seduced by this machine that was emitting all this code, just really, really seduced by it. And I felt like I was on.

on superpowers and then I realised that most of it was complete crap. what happened, it took me a while, it took me a while to realise it, two or three rewrites of this thing before I realised and I came up with this analogy to try and describe it to someone that, I’m not trying to suggest that I’m Michelangelo so please if you want to send a letter, send a letter to Brittany, don’t send it to me, I’m not suggesting I’m Michelangelo, but he said that in that David, that David was inside the block of marble.

and he just had to find it. And so I feel the same way about coding agents in that the difference between a muppet with no experience and someone with 30 years experience is they’ll both generate 200,000 lines of code. And the difference between the two is that the person who knows what they’re doing knows which 150,000 to throw away. And that’s the thing. so I get, I don’t know if we want to have a conversation about AI and coding agents and so on, but there’s a real polarized

Real polarized points of view about coding agents, like the classic Hacker News really, I mean the whole world’s polarized at the moment, but Hacker News is an example of this as well. You’ve got one side of the argument is sort of, I’ll never have to write code again, we don’t need software engineers, Claude, the GPT is just gonna chat or software into existence. Obvious nonsense. But the other extreme is also quite stupid, that these things don’t work, they don’t help you, they hallucinate, that and that. That’s nonsense as well, and if you know what you’re doing, you can be crazily productive.

with these things. Now, even after 12 months of doing this, I still find myself like today, like yelling at it, anthropomorphizing this thing.

Yelling at it because it done something stupid and I have to get up and walk away from a desk and go I’m not actually talking to a human. This is a machine You know escape escape control C controls. They get out start again, right reset But the truth of this is I think you know in the notes that you sent me this this mech suit idea Which is another metaphor that I came up with it’s a coding agents like a mech suit. It’s like Ripley an alien You you you if you know what you’re doing like if you don’t know what you’re doing you’re gonna hurt yourself and everyone around you

Right? They’re literally lethal. But if you do know what you’re doing, then you can kill the alien and live happily ever after. And that’s how I see it. It’s like a mech suit that you need to learn how to use. And once you pick up, and there are definite skills that come from this. Being able to, it’s not so much prompt engineering, but it’s sort of like controlling the agent. It’s like a parenting exercise. It’s parenting instead of programming. Teaching the agent to work in your best interest.

Anyway, like I said, happy accident between Elixir and Chord. They work very, very well together. It’s been very effective for me.

Erika (15:04) If for no other reason than like the pure fun of imagining myself in a mech suit, I’m like going to steal that imagery. Yeah, well, and thanks for sharing a little bit about your workflow too. Yeah, it’s, I mean, it seems very situational and kind of like you said, like,

Matthew Sinclair (15:14) Yeah, yeah.

Erika (15:29) setting up the situation, providing all the tooling, providing the feedback for a successful AI assisted development situation. That in itself is a skill. Yeah, and like you said, the view of the person in command.

like it’s still so valuable of like knowing when to go in one direction or another, when to shut it down. Yeah, yeah. How do you like in that flow that you’re talking about, like what are the intervention points? So you’re talking about like you sort of like build, it almost sounds like I’m understanding like mini context windows based on like iterations and then.

And then, how do you test along the workflow that you’re moving in the right direction?

Matthew Sinclair (16:25) I’ve just dropped in the chat there, the GitHub link, so you can have a at intent and maybe put in the show notes if anyone’s interested in it. And full disclosure, no warranties, it’s very much my workflow and I can’t guarantee that it’ll work for anyone else, but it works for me. But to your point, Erica, so…

I find it very useful to generate tests as I go, rather than sort of in advance. I get it to write something and then I get it to test it. And then what I’ve learned is it’s actually really, really bad at writing tests cold. It’ll sort of write a lot of tests, but they’re not very good. And so I constantly have to, when something doesn’t work, I’ve got to constantly go back to it and say, you’re testing implementation, not behavior. I want you to test the behavior. And so I end up, it’ll generate.

a thousand lines of tests and I’ll throw away the same ratio, right, two, three quarters of it and get it, when a bug happens, I’ll get it to write a test that tests for the bug, which is sort of what a human would do anyway, right, but it’s just way quicker at it than me writing to And what I find as well is, this is a whole other story here, but traditionally, if you sort of, if you have to stack the cost of writing software in a stack, right,

you’ve got, okay, I need to think about the problem, I need to sort of write down what I think the answer might be, an iterator, talk to someone, rub a duck, you know, whatever, and then I have to write the code. And so depending on how good you are as an engineer, you might allocate different amounts of time to those buckets, okay? And I would argue that the vast majority of engineers, unfortunately, spend most of their time writing code and not anywhere near enough time.

in the first phases because we’ve all been sort of told that you just need to write code and iterate it and so on, right? What Clawd does or coding agents do is they effectively reduce the time cost of writing the code to zero or almost zero, okay? And so what that does is it gives you a bunch of extra time that you otherwise, you didn’t have before to spend.

more in the thinking about what you’re gonna do. And this rubber ducking thing, like Claude’s really good at it. say, wanna see, you know the rubber duck metaphor, you sit there and explain a problem to a rubber duck, yeah, okay. So I even tell it that, right, I wanna rubber duck with you, this thing, and it gets quite amusing. It’s trying to be funny, not always funny. But we’ll sit there and go backwards and forwards with the design. And then often it comes out, the designer comes out, right, so a pretty good designer, go, bang, go and do it. And then, but then you’ve got, this is where the pruning starts, right, because it’ll mix.

a bunch of code and then you have to print it back. But anyway, that’s a rather circuitous way to say that the time cost of writing code, like your fingers on keys time, is sort of like that bit of the software engineering, which I know is not the total time, it’s the hard parts about software engineering, the human parts, not the software parts. But that bit’s now zero, okay, or not zero, but it might as well be zero. So you get all this other time to spend thinking about it upfront and you’ve got this sparring partner.

this intellectual sparring partner which you can think about it with. I don’t know if that answered your question, but that’s the, if your question was, how’s my workflow changed or what’s my workflow, how to keep it on track, I spend way more time now thinking and rubber ducking design than I ever did before. I used to be the design with code school, like I’d.

because you never really know what problem you’re solving, so you sort of experiment and the code’s the canvas, and so you’re sort of iterating on the code. I don’t do that anywhere near as much as I used to, if at all, to be honest, whereas I used to do that probably all the time.

Erika (19:51) For sure, yeah. And I also have found that the idea of designing something in multiple ways to come up with the best version out of multiple ideas, that cost is so much lower with AI assisted development. Because you can of spike out like.

not necessarily production ready versions, but what would this idea look like? What would this idea look like? Let’s compare them, see what I want to take from both of those, and then maybe throw one away entirely or create something totally new.

Matthew Sinclair (20:29) In intent, I’ve written a bunch of subagents in intent that fit into Claude and one of them is called Socrates. And the Socrates subagent, what Socrates does is I give it a…

I set up some context for it and I give it to Socrates and Socrates and Plato. So Socrates is the CTO and it’s the context there and Plato is the lead engineer with the problem and I get it to go off and do a Socratic dialogue on a problem. And so it’s got my input context and then it goes away and thinks about it and the output, like this seems insane.

Like even five years ago, if you’d have told me this was a thing, I would have gone, you are smoking crack. There’s no way this could possibly even be a thing ever. And however, the results of this is extraordinary. Like it’ll crack the hardest architectural problems, or even optimization problems, things going wrong. It goes off, it spends 10 minutes banging away, and then it comes back and says, right, here’s this 10 page Socratic dialogue.

between Socrates and Plato where they’re evaluating the code and the problem. And as you said, actually, the really interesting thing is it’s amazing when you say, me options and evaluate those options. It seems to lift a gear when it’s evaluating options against itself, when it’s arguing against itself. It’s super powerful. But anyway, that’s in the intent as one of the sub-agents, the Socrates sub-agent. I use it every day.

Erika (21:46) It kind of

reminds me of all the stories of creating DeepMind who played Go and ⁓ the way they trained it was actually replicating a second version and playing against itself for a month straight. I think the numbers is like it played more games of Go than…

Matthew Sinclair (21:54) yeah.

Erika (22:10) a single human code in like a hundred lifetimes or something. It’s like, yeah.

Matthew Sinclair (22:14) If I

remember correctly, I vaguely recalled it. They ran out of games. So they had historical game data and they ran out of historical games for it to play. And so they got it doing exactly what you said, playing it. And in some very short period of time, it had played more games than have ever been played by humans in Go, ever.

Erika (22:32) Mm-hmm.

Matthew Sinclair (22:33) Right, so it’s

Erika (22:33) Yeah.

Matthew Sinclair (22:35) exactly the same thing. For whatever reason, that really, and obviously Go goes way more complicated than chess. And what was his name? Lee Sedol, who got beaten by AlphaGo whenever it was. so like, okay, game playing is a solved problem for computers now.

Erika (22:51) Yeah.

Yeah. Super interesting. Well, thank you for sharing.

Bethany (22:57) Yeah, to pivot a little from the engineering side to the more leadership side. So as I understand, you’ve scaled a lot of teams from very minimal sizes to large sizes. And so I’m curious if you have any core philosophies on technical leadership or any mistakes that a lot of new managers or new leaders make when they’re

coming into a role like that.

Matthew Sinclair (23:24) Yes, I know a lot of mistakes because I made them and learned the lessons way too late. Look, the first one I would, I mean, I say this to everyone I’m coaching or working with is this basic why, what, how breakdown. And it’s a little bit different to the classic sales start with why, because I think they say why, how, what, but I spin this around, for leadership and management, I spin this around a little bit. And so my basic pitch to people is that,

Humans are storytelling apes, okay? And so in the absence of a good story, they’ll make up their own narrative, okay? And so what you have to do, and this is increasingly true, sadly, in a post-truth world which we seem to live in, there’s lots of narratives out there and people can pick, they’ve got a lot to choose from in terms of the stories. And like I said, if they don’t have a good one, they’ll make one up.

So the job of leadership, the first job of leadership is to explain why smart people that they’re working with should get out of bed in the morning and come and play their game, whatever their game may be, right? That’s the first job, tell a good story. And then the second, and if you can’t do that, the smart people have got plenty of options, they’re go and do something else. And so then the second thing is peers should collaborate to work out what to do, not how, but what. And then the trick of good management is to get out of the way.

I think and let smart people work out how to do the things that they’ve hired them to do. So that why, what, how I I use that, I use that everywhere. The other thing that I say to people about leadership and management is that management is not a promotion. It’s a career change. Okay. I didn’t coin that phrase, someone else coined it. But the interesting thing that they also don’t tell you is that leadership is a career change from management. Okay. And that, that one is like, that’s a footnote at the back of the, you know,

the advanced textbook, you don’t really get that one. so a lot of people, when they go into leadership positions, treat it as management, and it’s not the same thing.

Okay, and so that little lesson alone, just knowing that that’s a thing, that when I’m becoming a manager and moving out of IC to management, and our podcast that I’ve mentioned before, What Next? is actually based on this premise that interesting things happen at boundary conditions, on the boundaries, right? So when you change from something to something else, that’s when interesting stuff happens. And the palliative that I give to the flailing new manager or new leader is that your worst day in the job is today.

Right, because today you’ve learned something and tomorrow you’ll know a little bit more than you knew today, so it can only get better. And the worst time you’ll ever have as a manager or a leader is the first day, your first day in the job,

And again, it’s one of these little things about just knowing that is very helpful because it allows you to lift your head a little bit and go, okay, right, today was terrible, but I learned something, so now I’m gonna be better tomorrow. So those two or three things are pretty interesting. The other one is, the last one was a bit more general, not really management-related, but I’m a big believer in values, and what I mean by that is that if you’ve got shared values, this works across the board, socially, family relationships and so on.

If you’ve got shared values, you can kind of get through just about anything. And if you don’t have shared values, it’s hard, much, much harder. And so it’s not that everyone has to have the same values, but when you’re selecting for things, it’s probably a good idea to try and pull people together that have similar values. So if you need to…

do exploit something or go really to the hustle, then you want people who have a kind hustling mentality. If you want to do something bit more nurturing and a bit more, know, whatever, holistic, then you probably want people who share those values. And it’s when you mix that up, you tend to get a bit of friction. So, yeah, that’s another bit of advice that I’d give to them. It’s very easy to say, of course, from here. But it’s just one of those things, just knowing it is quite handy.

Erika (27:02) Do you have like a reference of values that you go from or are they usually like self-described?

Matthew Sinclair (27:08) I have my own that I stick with, are more or less what I’ve said to you so far. Why, what, how. Actually, the first one I have is be intentional about your values. This is one of things I say, particularly people in startups where you’re growing your business and there’s no such thing as no culture.

there will always be a culture and you can have an intentional culture or an unintentional culture. And so as a CEO of a small business in trying to grow something, it’s pretty important that you’re intentional about the culture. Otherwise you’ll get a culture that you may not have wanted. Or worse, you end up a culture that’s just based on the bad behaviors of the most senior people. All right, which I’m sure we’ve all seen.

we’ve all experienced that and it’s pretty toxic. So the first one is be intentional about having some values and communicate them. There’s a bit of a test that you can use with this, is, actually if I make it to a slight diversion, I think the role of the CTO, just to bring back to something practical, the role of the CTO is what I call the…

Your role is actually Chief Vocabulary Officer. And what I mean by that is it’s your job to come up with the words that people use to communicate because as the CTO, you’ve got to…

foot in both camps, right? You should know the technology, obviously, that’s self-evident, but you also should know the business and you should know the product and you should know the people, okay? And so it’s your job as the CTO to admit the words that people can use to communicate with each other without feeling uncomfortable and they can talk about progress and status and so on.

So yeah, have some values, be intentional about it. There’s no such thing as no culture. You are, as a founder and CEO of a business, you are the culture ultimately and you have to demonstrate it. As a CTO, you need your jobs to come up with the vocabulary that people can use. Why, what, how, it’s very effective. Always remember that boundary conditions are really interesting, so they can be perilous, but they’re also opportunities.

And again, there’s a guy that used to work for me, we had a really deep conversation one day and people had been telling him all this career that he needed to change this particular behavior that he had. it occurred to at the time that I think sometimes it’s really bad advice because the way you are is really hard to change. And so what’s better is just knowing how you are. And then if you know how you are, then you can maybe grab a mask on a…

from time to time and put a mask on to fit into a certain context or situation. But changing is exhausting or even impossible for many people. And so just knowing, this is one of those situations where I’m gonna lose it and so I’m going to write a one-pager for the meeting, submit it, and shut up, right? Or I’m…

I’m one of those people who just talks in every empty space in a conversation. So I know that I’m gonna go into the room, into this meeting, and I’m gonna find the most junior person in the room, and I’m gonna go and ask them what they’re doing or how they’re feeling, or what they’re working on. These are just sort of hacks that you come up with to get through it. I don’t know, I’m not sure there’s any real recipe for any of this. If there was, we could hand it over to machines.

Brittany Ellich (30:15) It sounds like you’ve sort of been on a journey to find out these things about you. Do you have any recommendations for somebody who wants to know these things about themselves?

Matthew Sinclair (30:25) Gosh, that is a very deep question. I’m not sure that I do know, to be honest. I think I can kind of guess at it a little bit and I’ve fumbled around in the dark a little bit. one thing that really helped me was I started writing down how I thought about stuff. And I…

I don’t have, like I’m not a successful blogger anything like this, right? But what I started doing was every Friday at Digital Ventures, when I was at Digital Ventures, like something would happen through the week of, you know, it pretty intense environment. Something would happen and I would, I could be like a one-liner and I would use that as the title of the blog post and then I would put it out to the team. And in the end, I started putting it on Medium, like it was internal, an email I did internally and then I did it as a blog post on Medium. And what,

In the back of my mind, what I was thinking I was going to do is I was going to try and write a book about how to be an engineering leader. And it turned out that Will Larson did a way better job of that. And The Elegant Puzzle is the book that I wanted to write, and so I don’t need to write that book anymore. But what I do know about writing, like a lot of creative activities, is that there’s craft and there’s art. Okay, so when most people look at something creative, they see the art, but they don’t see the craft.

So if you take a painting on a wall in an art gallery, you see the art, but what you don’t see is someone understanding how light falls on paint and how you stretch your canvas or how you mix paints or find ingredients for your novel colors and so on. But no one sees that unless they are themselves are in that domain. And I think leadership’s a little bit like that as well, right? You see that.

When you see a good leader, you sort of see the art of it, but you don’t see the craft. And so I’ve tried to come up with these, I call craft moves, that I’ve collected over the years, which are just little lessons, and like the one about going into the meeting and talking to the most junior person and finding out what they’re doing, and shining a light on them, as it were, and just listening to what they say. It’s a very good signal to give to a team, to let someone have a voice. And it’s not hard to do, right? It’s a really, really easy thing to do. You just have to put your ego out of the way.

You know, the, I don’t know, so right, back to the thread. Writing really helped me. What I found I was doing was, if you, like a lot of these things that I’m saying are actually blog posts that I’ve written, and what happened was in the writing, in writing it down, I went from just mush to a coherent couple of paragraphs that made some sense, and that was super helpful. So in doing that, it turns an idea into, like just a really rough idea into something that you can

And so I had this, like one of things is this dolphin’s not whales, right? In a conversation one day at work, I was trying to explain to a client why we work the way we did. And I said, it’s a bit like a whale. If you think about a whale jumps out of the water, sorry, comes out, breaches out of the water and falls back in. If it’s going the wrong way, it’s got a huge arc that it has to turn around to come back to get on the right direction. And if you think of a dolphin, a dolphin’s, you know,

really quick and agile and they come up out of the water all the time and if they ever get it wrong, they’re never wrong by very much, they can course correct. And so this whole dolphins not whales thing became a mantra in TV, this is how we work, dolphins not whales. Anyway, and it was an accident, it was just a conversational accident, but I turned it into a blog post and then it became a thing and it was like, Matt, you have to come and do the dolphins not whales presentation for this client. yeah, so write stuff down. It’s, you know, an audience of you.

I mean it really was audience of me just writing down. It helped me turn rubbish, messy ideas into something that I could repeat.

Brittany Ellich (33:53) that yeah I think there’s a lot of science behind that too like the whole mindfulness and journaling and how it’s just so good to like get all of your ideas out. This has been awesome I don’t think I’ve ever been excited about like checking out Elixir but now I am you’ve been a great evangelist for that. You have so many great ideas and I’m so excited to dig into all of these links after this

Matthew Sinclair (34:14) the Elixir community is very welcoming, the Slack’s great, the Discord’s really good, there’s a bunch of really smart people in there, definitely check it out, it’s well worth your time.

Brittany Ellich (34:21) That’s awesome. Yeah, well, we’ll include links to that in the show notes and I will also check it out because it sounds cool. I love Bethany’s idea. Advent of Kodak is coming up. So maybe maybe I’ll also take that challenge to do, Alexa. We are coming up on time, so I want to transition a little bit to our fun segments. This week we are going to be doing Two Truths and a Lie. Matt.

We know you have a few unique interests like owning an electric mountain bike and a finger lime orchard in Australia. We want you to give us three statements about your life or career and we have to guess which one is the lie.

Matthew Sinclair (34:59) Okay, we have to test my poker face now.

And tell me see if I see you can tell what I’m Okay, so in no particular order to choose to align I can say my alphabet backwards very very quickly and there’s very few people in the world I know who can do that my son I spent some time teaching me how to do it. My grandma taught me how to do it You’ll know my surname Sinclair. It’s the second one. So I’m related to the Sinclair’s of the Knights Templar a second the second founding family of the Knights Templar and the Rosicrucians

if you’ve been to the Rosalind Church in Edinburgh, which was featured in what’s it called Dan Brown’s, what was that book called? Da Vinci Code. And last one, I have a reconstructed right ankle that looks like a bike chain under X-ray and does set off airport security when you go through.

Brittany Ellich (35:42) Wow, you really took this assignment in and made it really hard. That was great.

Erika (35:47) I

was trying to do my alphabet backwards when you were saying that and I got like four letters in.

Matthew Sinclair (35:54) Sid by x, w,

t, r, q, e, o, n, l, k, j, h, g, f, e, d, c, a.

Erika (35:58) Alright, so that was true. So now we only have two to pick from.

Brittany Ellich (35:59) So it’s not that one.

Bethany (36:00) There was too much detail with that one.

Matthew Sinclair (36:02) Now you’ve got a…

Now you’ve got… I tried to put unnecessary detail into all of them. I mean, I may have just learned my alphabet. I may have just learned my alphabet backwards for this question. Well, you’ve got a 50-50 chance now.

Brittany Ellich (36:05) Mm-hmm.

Erika (36:14) I think the second

Bethany (36:14) Bye.

Erika (36:16) one is the lie.

Bethany (36:18) That’s what I was gonna guess, yeah. While there was detail in all of them, there wasn’t that much personal detail in the second one.

Matthew Sinclair (36:25) Yeah, you’re right. I put that one in because I would like it to be true. I keep waiting for the phone call to be admitted to the Knights Templar. It’d be quite cool. And the Illuminati, but it hasn’t happened yet. So may never happen. So no, that is the lie. Well done. My poker career is over. My poker career is over.

Brittany Ellich (36:39) Well, I guess we’ll never know, too. Yeah.

Erika (36:42) It

was very good lie.

Brittany Ellich (36:45) It was good. It was good. Yeah, I was very convinced. This has been, this has been awesome. Matt, welcome. Where can people find you if they want to hear more from you on the internet?

Matthew Sinclair (36:58) A very easy spot to start is matthewsinclair.com and you’ll have to bear with me because I have this long-term project to bring down surveillance capitalism, which is probably a topic for another podcast. But I am slowly putting all of my content onto my own stack. So I was using Medium and so on. rather than just simply do a simple static website, I decided to make my own…

my own tech stack for the CMS, because I’ve got a bunch of websites. So matthewsinclair.com is my personal website, but it may be a little bit patchy at the moment. It’s all written in Elixir, and it’s running on, yeah, it doesn’t matter what it’s running. It’s written in Elixir, and it’s all part of a thing that I’m building called Luxa, which if you know what a Luxa is, it’s a sort really cool, very spicy soup from Singapore or Malaysia. And Luxa is like all the flavors.

And so my idea is to put sort of all of the CMS functionality into this thing called Luxa and I can very easily publish like a static site. Everyone, when they learn a new language, right, the first thing they do is write a static site publisher. So I did that, but I sort of did it on steroids with a bunch of other stuff. So that’s gonna, I’m gonna release that as not too distant future. I was working on it here this evening. But yeah, mattysingler.com, that’s the best place to go.

Brittany Ellich (38:10) I love that. Do have RSS? Speaking of, on your own? Okay, great. Yeah.

Matthew Sinclair (38:13) It does on, yeah,

well, it does on, in fact, Luxor has, by default, you can just automatically RSS, it puts the RSS on the end of the URL and it’ll give you an RSS for the blocks. It should, test it out, let me know.

Brittany Ellich (38:25) Love that.

I will, I will. RSS is the future. It’s the past, but it’s also the future, I think. I agree with that.

Matthew Sinclair (38:32) Honestly,

I’m so with you, right? This is part of my bring down some ounce capitalism thing. I think we need to get back to web rings and completely federated non data monopolist and networks of small, special interest groups, not this algorithm, what do call it? Brain rot, crap. yeah, I’m gonna fight this campaign on my own. So join me if you wanna come along.

Brittany Ellich (38:38) Mm-hmm.

Love that. Yeah.

Yes,

yes, we’re, I think we’re all there with you for sure. This is great. Well, thank you so much for tuning in to Overcommitted. If you like what you hear, please do follow, subscribe, or do whatever it is you like to do on the podcast app of your choice. Check us out on Blue Sky, especially if you are also interested in stepping away from surveillance capitalism and share with your friends. Until next week, goodbye.

Matthew Sinclair (39:00) Yeah, Good.