Relentless
01:00:25 · Brian Armstrong
[01:00:25] S0: Yeah. Yeah. So it'd be game over. Right? And you're always just trying to, like, how do you not die and make it to the next step of the game? So if you just don't die sometimes in business, like, that's enough to succeed because everyone else dies. And so this was a moment where I I didn't really know how to build. I mean, I I studied computer science. I had studied... Like, taken a class in computer security, but I was not like an expert in cryptography or anything like that. Right? And we didn't have that many people in the company yet. It was either. It was pretty... Still a relatively small team, relatively junior team. So as I often do when I'm trying to, like, figure it out, I think, like, who who would I know in the world that could help with this? Right? So I called two people, and they were both cryptography experts. Actually, one of them was Zooko, who later created Zcash. One of them was this guy, Ryan Lackey, who's, a security professional, really interesting dude. And I called both of them, and I said, we need to build this next generation cold storage system to store a ton of crypto. Number one, what's the right architecture to build it? And they gave me their opinion on that. And then I said, number two, how long do you think it would take to build it? And both of them said something like, this would take a team of 10 people about two years to build it. And I was like, well, we have about sixty days. And by the way, it's like really just like me and this one other guy. Like, we don't have like a big team. And and so they were like, oh, that sounds, really challenging, you know, but, like, let me know if I can help. Good luck. And so this was one of those kind of crisis moments where it could have ended the company, but it actually turned into a defining moment because what we did was we sat down, and this was like myself and Charlie Lee and Craig and a couple other early employees, and we said, we're not going to be able to do like a perfect system that a giant company would be able to do, but we can do a pretty good system. And a lot of engineering is trade offs, right? It's like, okay, what would get us 80% of the solution for 20% of the work here? The remaining 20% that's not there, can we live with that? Can we mitigate that risk? How might we mitigate it?
01:02:24 · Brian Armstrong, Ty Morse
And so we actually sat down within like three or four or five days, we had designed like what we thought was their next architecture. We started to go and code it up. We then had the challenge of trying to transfer all of the old system into the new system. And I remember as we were racing the clock towards this sixty day mark, we were all, like, pretty sleep deprived. And if you study how accidents happen in history, like the Exxon Valdez spill and even the Challenger explosion, a lot of these things can be traced back to sleep deprivation as one of the [01:02:52] S1: contributing factors. Mean, I Brian Johnson literally has this graph where it's like decision making quality over the course of a day, and it's just... Yeah. Yeah.
01:02:59 · Brian Armstrong
[01:02:59] S0: So I remember, you know, we were sitting there, and we were about to like, okay. We need to do the initial transfer into this new system, and let's do a small amount. Like, let's transfer, like, $1 or something. And I remember I was like, we need to pair... We're gonna sit side by side and pair program every single keystroke because we're both sleep deprived and we need to check each other's work. It's kind of like copilots in the airplane. You know, you like double check the work. And so we transferred like a dollar, and we waited to confirm that. Okay, nothing exploded. Let's do $100 Let's do $1,000 Let's do $1,000,000 You know? And I remember we started this migration, and we're like, okay, let's go home and, like, sleep for a little bit and make sure and come back. And we started transferring, like, really big numbers. And every one of those was scary because, like, if one of them had gone wrong... And there were a bunch of hackers trying to break into our systems at this time. You have to remember, like, we didn't know could one of our laptops have malware on it. So we were we were doing some of the key signing, like, totally offline, disconnected from the Internet with these, like, brand new laptops we had just brought, you know, brought out of the bag that we were pretty sure had not been compromised, you know, like this kind of stuff. And so it was nerve racking, and we got it all transferred like nothing blew up, and I was just... I just remember being, like, so relieved, and we just went home and finally slept like twelve hours. The main lesson of that from my point of view was that, number one, I look back on that as one of the most important moments in my career, right? It showed me that we were capable of more than we thought. On shorter timelines, we were able to do great work. And, you know, I'm really proud of it too. Like, it's something... And so I... Now there's been many stories like that inside Coinbase where different teams have risen to the challenge. And, you know, I often find, like, sometimes teams are really scared if you go in there and you're like, okay, we need a solution to this within this amount of time. And they're like, why are... You know, why is everything so urgent? Why are cracking the whip, Brian? You know, like, like, not everything is just like a made up deadline. You know, this is like... So...
01:04:57 · Brian Armstrong
And and look, there is an element of truth to that. Like, some things require a great craft, and you don't wanna just be like, if you force everybody to do artificial deadlines, you'll get, like, really shoddy product coming out sometime. So there's a lot of truth to that. But it's also true that people are capable of more than they think, and the problem will expand to the amount of time that you give, right? And so it's great if you, like, come up with a goal. And even if you don't hit it exactly, the fact that you had that goal means you're gonna be a lot closer to it than you would have otherwise. So, yeah, there's moments like that that we're defining that required us to make it a defining moment of the company, our our finest hour instead of our end of the company.
01:05:35 · Ty Morse
[01:05:35] S1: With podcasting, it's kind of this magical thing where for this shoot, this is a huge thing that we have to pull off, and we have a deadline, and it's very clear because your schedule is... It's either you're here and there's something or there's nothing. And it's the same thing when I was interviewing Peter Beck, he said, If rocket launches didn't have the payoff of seeing the rocket actually go up and work, no one would work on it because it just sucks most of the time. How do you think about trying to find those critical milestones almost where you can rally the entire company? Can't have... Even if you manage $500,000,000,000 eventually it gets boring. And so you have to kind of create, here's the next here's the next goal. Here's where we're marching next. Yeah. How do you constantly try to find that next marching post?
01:06:18 · Brian Armstrong, Ty Morse
[01:06:18] S0: Yeah. Well, that's part of what was so great about, like, the moon landing is that it had such a clear moment defined in everyone's mind visually. It's like, did the guy walk on the moon or not? Like, you know if you did it, and it was a big one. Most companies don't have that. Right? And so most companies are on some kind of cadence, like they're putting out a new car every two years or, like, a piece of software is, like, shipping continuous updates every week now every day. And so one thing we try to do is have, like, an operating cadence to the company where, for instance, now twice a year, we're doing these product showcase events, and so those are great moments to, like, you know, ask everyone, like, okay. That thing is six months away. Let's write the headline. Like, what do we want to be able to announce on that day for our customers? And then it's like everyone can rally behind that. And even things which launched a few months earlier, we still haven't told the world about it in an adequate way. We need to package it up and like... So those those tend to be, like, great milestones. You know, you can do, like, annual planning processes and OKRs and q... You know, quarterly business reviews and things like that. But I think, like, having those public moments helps rally teams. [01:07:18] S1: Part of the reason why I was initially thinking doing a crazy set and costume design and all this stuff is because of some of the things that you've done publicly to communicate better with your shareholders and stakeholders. [01:07:28] S0: Yeah.