字幕表 動画を再生する 英語字幕をプリント I was I was out to dinner with a friend of mine. Um, and we're having a conversation about him becoming a manager in my current group. And, um and he was you know, dude, I just totally don't want to do it like I'm not gonna be doing anything. I'm just gonna be like a manager from now on. Uh, it was one of those moments where you blurt something out. You didn't realize you thought it until you said it on. Guy said out loud without consciously thinking about I said, I said, Come on, man, Like this is an amazing opportunity. It's your job to protect this incredibly fragile bacteriological colony inside this body called intel. And if you let until have its way with it, intel will kill it. But But if you don't, then just like in a human body, you know, the flora and fauna and our gut right, it benefits the body, right? It works out well. And I realized at that moment that that was sort of how I think about building these things. So plucky rebels being agile in an unnatural place. What is pluckiness? Eso? Flakiness is a combination of determination, but it is not grim. Nobody wants to follow the guy who says, Let's all die going up that hill, right? It is also cheerful, but nobody wants to follow somebody who is purely cheerful because purely cheerful is Well, well, I don't really care. I'm not gonna keep trying. So pluckiness is a combination of cheerful and determined. Why rebels? Well, um, if Intel already knew how to do the thing that you're thinking about trying to do inside Intel than Intel would have a p n l. And there'd be a Z B B, an appeal war, and you'd be on a road map and you wouldn't be a start up. It'd be a multibillion dollar business and like, that's not a start up. So, um, a little bit of insurrection is good for the soul. Uh, so that's why rebels and another point, how much agile is the right amount on? They're probably people in this room know more about agile than I do. I am, you know, scrums and sprints. And, um, the right amount of agile is right up to the body, rejecting the bacteria everything up until there is. You want as much as you can possibly make fit, but you need a veneer. You need a veneer of predictability until is gonna want to see quarterly things. Maybe you're gonna have to have finance help you do an NPV. So how do you do this? Rule number one Be right about the future. How can you be right about the future? The best way to do that is to create the future. Because if you create it, can you get it right? Um, I've gotten this wrong in my career. In the mid nineties, I was working in the Windows Operating system group, and it was my responsibility to figure out what the right strategy was for CDR. Anybody remembers recordable, right? One CD media, Uh, and we were thinking, should we put this in the Windows operating system or not? And we thought, Oh, yeah, the general consensus was users don't understand, right? Once the operating system can't do right, once we're gonna exhaust the media. If we do backups to the auto saves and word documents, all they're not gonna work. So the net result of that was we chose not to make a future in which you could easily manage rewriteable or right ones, media. And if anybody remembers how much of a pain it was in 2000 5 years after we ran into this problem toe actually put anything on a CD? It's because in 1995 we didn't decide to make a future in which it would be easy. In 2000 I will take credit for having screwed that up. Um, I've gotten this right in my career. I'll talk about bit locker in a second, But even if you're wrong, deciding you're gonna make a future can actually create profound impact. There was a guy named Jack. He was writing software toe help. People type in messages on a computer in order to send an SMS message to a taxicab to do a dispatch system in New York City. The contract failed. The technology couldn't ship. Have you ever used Twitter because that was Twitter? So even if you're wrong about the future you're creating, if you've got forward momentum, that can really help. So that's sort of this notion of leading from the front. Keep it secret. Keep it safe from Gandalf, one of my favorite Wiseman. There are lots of people inside large companies who want to help you. They're very well, meaning their heart is in the right place. It is just that some kinds of help with the kinds of help we all could do without, they're gonna want you to give them a 10 year NPV. They're gonna want to show a 31% return on the investment from start to finish. They're gonna want you to be able to explain exactly what it is that you're gonna be doing to take advantage of a hardware feature that doesn't ship for two years. Um, don't be afraid to go dark for a while when you're doing a start up your super proud. And this is one difference between doing a start up in a big company and a small company, and I've done both. When you're in a small company, you're starved for attention. You will tell absolutely anybody about what you're doing in a big company. There's a lot to be said for going dark until you have something to show. We had this philosophy. Shipping software beats power point. Don't explain what is going to do as soon as you come out of being dark handed to someone, then you don't have to explain what it's gonna do. They can hold it. They can see it. They can feel it. They interact with it. They know what it does. They may ask, Why did you build this? You say, Well, you know, because it's really awesome. Um, but but modern software is all about people wanting a return right here. Right now. That's one of the things that service is a really profoundly changing the way we do software. So, um, keep it secret, keep it safe, and then when it's time to talk about it, have something to show. Don't don't rely on power point to tell your story. Ah, always cheat. Always win another wise man, Clint Smith. But Peter, it says, right here on my badge. You know, Intel, Code of conduct. That's dishonorable cheating. It's your job. Is the leader of a start up in a big company or a small company Thio To make things unfair on behalf of your people, it is your job to optimize the system so that you get greater returns than everybody else. That's your job. So a little bit of optimizing is a very Good thing. Um, this is kind of like the bacteria analogy again, right? You you you want you want, you gotta carve out enough space for that bacteria to actually do its job. Digest food, right? If the bacteria is gone, you can't digest it anymore. But if it's too successful in it exist the body and comes back in somewhere else, it kills you. So you want to You want to contain that, But at the same time, let it do its job. Um, we do this one cheating thing we do. We like to make up names so people can assume they know what you're doing when you say cloud service is. But if you say affiliated service's, they have absolutely no idea what you're talking about, which means you get to explain it. And it means that especially in a large company, you're not inheriting the debt of defying term when you try to talk about it because people are, like, totally no cloud services are. We tried those three years ago. Don't you remember? Insert name here. Name here. Right. Um, well, but these air affiliated service's they're totally different. It's cheating. They're exactly the same but But it's very effective, so I recommend that. So you do this to help you on Dhe. This is my number one rule. If you forget everything else about anything I've said, find some users and make them happy. There will be projects in the future where you see the most profoundly awesome graph that says, If we do these 19 things, we will have a $1.7 billion business in the next five years. You can always do that right. There's always a way to to take data and help support a business unless it's truly, catastrophically bad, in which case I shouldn't. But if you have people who are happy with what you've done, if you have found some users and you have made them happy, you have got a superpower, right? I mean, you have something now where you could be like Bob. Bob likes it. Bob wants more. Bob told me he wants the next 19 features. Yeah, but you know, but Bob wants 19 more features like it becomes this incredibly compelling conversation. If you have some users and you've made them happy and just the act of going out and finding people in deciding. I want to make some people happy today. First of all, it's kind of nice from a societal perspective, but also it's a you're going straight to the solving the problem and, um uh, an example bit locker in tooth. Roughly 2001. I was having a conversation with a security officer from a bank, and he said, We lose on average, one laptop in a New York City taxicab every day. While later I'd like to say it was like pitching like this lightning bolt, right? Oh, inspiration, Genius. Probably like a year later, it suddenly dawned on me. There is an entire office in that company and they're an only job is to deal with the misery of a loss laptop in New York City. That's all they dio. Nobody ever told me that. But you have to have those people there. Have to be. People are like, Okay, let me get this straight. It was the e V p of mergers and acquisitions. Was was the hard driving scripted like where did go? I mean, it's two o'clock in the morning. Do we have to call the Wall Street Journal? Do we need to get the analysts in, You know it's a nightmare. So we decided, Let's build a product that is exclusively dedicated to making those people happy. We're gonna build a product that makes risk managers and incident Response Team's inside large corporations happier about having to deal with lost in solar laptops. The product of that was bit locker last. I knew it was over half $1,000,000,000 revenue line, but it was exclusive like we did not think about I t workers. We did not sit. Think about the I t department. Like we just said, All we want to do is say that those people go from from seven hours of sleep a night to eight hours of sleep and were successful on git was a huge, uh, success for my perspective, so make an attractive corpse. All right, so we're planning to fail. No, we're not planning to fail. You should never plan to fail. But here's the thing. Projects get canceled, the world changes. Cos change. This happens in big companies. This happens in small companies. Um, you never know when the music is going to stop. So this is a thing that agile can really help with because the cadence of agile in the agile process delivers more frequent leases of smaller things that s'more little victories. Uh, the more little victories you could have, the more likely it is that if you do, God forbid, wind up being a corpse. You are an attractive corpse because you have given the people who were with you on this journey a better resume, an opportunity for the next job. Ah, thing they can hold.