Ep 70: "The Geek Way" — Andrew McAfee's New Book on the Radical Mindset that Transformed the World

0:00It's called the planning fallacy. We are just super overconfident in our ability to upfront map out what's going to happen. We think we can do it. Everything from how long your home renovation. is going to take to what the customer wants. We're just overconfident on our ability to know the future and plot a course through it. And we're not. We're actually fairly lousy at that. Peter's insight, the insight of a lot of geeks that I work with, is that you need to do some planning. You need a minimum viable plan. After that, the only way you're going to learn about reality and learn about what the customer wants or learn about how long your kitchen's going to take is to start doing stuff.

0:43I talked to Andrew today about his new book, The Geek Way. Andrew's a bestselling author. He's written a lot of great books about how technology influences the world. Helped a lot of us think about what we're doing here in Silicon Valley better. There's a lot of big industries that Silicon Valley has run circles around. There's some really key principles around how these companies work. And there's a lot of ways the rest of the world can learn from the very best organizations in Silicon Valley. The government has a lot to learn here. I'm starting a new university. The new universities that we build have a lot to learn here.

1:10I'm going to give this book to all of our CEOs, and I'm excited for you all to learn from Andrew today. Andrew, you're a principal research scientist at MIT. You're a bestselling author, and your new book is The Geek Way, The Radical Mindset That Drives Extraordinary Results. Thanks for joining us, Andrew. Joe, thanks for having me. So start by telling our audience, please, a little bit about yourself and your work. Broadly speaking, you study how technology impacts our world. Can you unpack that? Yeah. Yeah. So all the books that I've written and the research that I've done has been completing a sentence, how tech progress changes X and X changes.

1:41Maybe it can be jobs and wages. Maybe it's business models. The last book I wrote was our relationship with the environment. And Joe, if you had told me at any point in the past 20 years that I would ever write a book about culture, about corporate culture, I would have laughed in your face because it wasn't that there's no good work done on culture. There's some beautiful work done. But in general, I found the discussions kind of self-congratulatory and fluffy. And there's no shortage of CEOs willing to tell you how they made the tough call and pulled and turned things. And they were just, I didn't get anything out of it.

2:15But I spent my career with one foot in the old economy and one foot in the new economy. And I hate those labels, but let's go with them. And I just kept trying to pattern match and try to understand what was so different about the two because they were deeply different. And those differences mattered because the upstarts were just mopping the floor with the incumbents over and over and over again. This was not a fair fight at all. And I tried to figure out what these upstarts, the new economy, the Silicon Valley, whatever phrase you want to use, what were they doing that was so different? And the book, The Geek Way, is my answer to that question.

2:51I think they've upgraded the technology that is the company. You've captured it really well. You're describing a lot of what I've learned from my mentors and done over the last 20 years. Before we jump into the book, I want to ask, in 2020, you wrote a Wired article titled, Why Degrowth is the Worst Idea on the Planet, talking a lot about the technology world. And I'm curious, why do you think so many people distrust technology and even fair technology today? What's going on there? Joe, I don't know. It's weird. Your podcast is called The American Optimist, right? And we're in this period of pessimism about a lot of the things that I think we should be incredibly optimistic about.

3:27And it is true that every time a technology, a super powerful technology comes along, and maybe I'll lapse into economist speak and call these things general purpose technologies. These are like the steam engine and the internal combustion engine and electrification. These are the things that accelerate economic progress and not coincidentally make our lives better. Every time one of these things comes along, some problems come along with it. This is not new news anymore. But we've lapsed into this mode that I find weird and troubling, where especially the more elite you are, the fancier you are, you signal that you're elite and fancy, not by analyzing the amazing new technologies that we have now, but by problematizing them.

4:13And those are not the same exercise. It's like we're back to the Luddites again, but it's a sophisticated version of it. Yeah, but let's just list all the things that could possibly go wrong. That's not that hard to do. You and I could get it on a whiteboard and make a long list of all the bad things about the hammer. I find it a shallow exercise overall. Analyzing, talking about costs and benefits or trying to understand what's going on in a holistic way, that's a really difficult exercise. I think in general, when we analyze these very powerful technologies, we conclude that nothing improves the human condition and the state of nature more than technological progress.

4:50And somehow we've lapsed into problematizing instead of analyzing. I think that's a shame. It's completely true. I think you have to actually be slightly smarter to understand systems and be optimistic about the new possibilities we don't see versus it's like a slightly easier problem to see the obvious problems with changing things. You know, I think it's not maybe not just smarter, but also a little bit braver, because I think there's an old John Stuart Mill quote who says it's the person who despairs and everybody else is in a good mood that we think is smart. I just I completely butchered that quote.

5:17But he nailed this negativity bias that we still have and that I think is really rearing its head today. And I think you and I should say up front, we are not I'm not and I know you're not. We're not technology Pollyannas. There are challenges that come along with very powerful technologies. Amen. We need to spend time on them and deal with them. Amen. Do we need to sit around and problematize the technology all the time? No. Well, you know, I woke up every morning this summer with notes from friends who journalists had reached out to to try to find some dirt on me. So I think this is a culture right now.

5:50This is a to attack technologists. But let's jump into analyzing them instead and hearing about them. Your book's called The Geek Way. How do you define a geek? You grew up obsessed with computers. Do you count as a geek? Yes, but that's being arrogant because geek is a term of praise for me. And I think more broadly in our culture, the definition of geek has morphed and it's now tossed around more as a term of praise or admiration than an insult. And it's not just somebody who likes computers way too much. These days, I think a geek is anyone with two characteristics. Number one, they get obsessed with something and they're tenacious and they can't let it go.

6:24And whether that's wine or Star Wars trivia or artificial intelligence doesn't really matter. They just get obsessed and they want to solve the problem that that is obsessing them. And if the solution that they come up with is unconventional, that's OK. They're not constrained by conventional opinion of their mainstream or status quo. They will go where their inquiries take them, even if it's to an unusual place. So these days, my shorthand for a geek is an obsessive maverick. I love it. And your core thesis of the book is that a bunch of geeks have figured out a better way to run a company. Joe, a present company very much included, I think, of you as part of this cohort of geeks that for a while now, but I think really accelerating over the course of the 21st century said, like, wait a minute, we don't have to do things the way companies did them in the industrial era.

7:14We are not bound by the status quo. And we're going to keep tinkering and iterating and experimenting until we find ways to solve these dysfunctions that keep creeping in, like bureaucracy and sclerosis and big projects that go so badly off track and these corporate cultures of silence and different kinds of toxicity. I think what you and your cohort did is say, can we do better here? And my answer is yes, hence the book. Who were the group of people, though, responsible for the culture? How did this start in Silicon Valley? How did they get into the water there? I asked people, some folk that you know, I asked Eric Schmidt this question a while.

7:50I asked Carl Bass this question. These are guys who have been around the valley for a long, long time. And they say, look, you know, there was always in Northern California, this combination of academia, of industry, and sometimes even defense industry, and hippiness and freakiness and letting your freak flag fly. John Markoff, who's the longtime technology reporter for the New York Times, wrote a great book called What the Dormouse Saw or Said. I can't get the line right. about this early confluence of freaky hippie culture and academia and companies trying to invent these things called computers in the 70s and 80s in California that helped create the personal computer movement.

8:35But I think more fundamentally, this particular cocktail that is shaken up in Northern California, bringing this stuff together, I think it was a really important catalyst. And then in the 21st century, as the web expanded, as the smartphone revolution expanded, as capital became more abundant, I think there was the second wave of geekdom entrepreneurship that went on. Again, headquartered out there, but not exclusive to there. Yeah, no, I mean, I definitely wouldn't have known how to do the things I did if I hadn't gotten to spend some time at PayPal while I was at Stanford. I think it's just seeing what a functional culture looks like.

9:10And I love how you break it down into four norms. So the four norms are speed, ownership, science, and openness. On the speed one to start with, I want to ask you to describe it. One of the things that really stood out to me is plan less and iterate more is one of the things you said. And, you know, working with Peter Thiel, there'd be times where I want to like talk about something we were going to do or talk about something we should analyze. He was never interested in that. He was like, just do it and then come back to me with it. It was this culture of let's just go fast and like, let's talk about this again in an hour after you guys have done that work.

9:39And there was no like waiting like a month or a week or anything. It was just go, go, go, go, go, you know? Yeah. I didn't know that about Peter. I didn't know that he was wired so much that way, but he overcomes this very deep seated human tendency that we have. I think it's so deep seated that Danny Kahneman gave it its own label. It's called the planning fallacy. We are just super overconfident in our ability to upfront map out what's going to happen. We think we can do it. Everything from how long your home renovation is going to take to what the customer wants. we're just overconfident on our ability to know the future and plot a course through it.

10:13And we're not, we're like, we're actually fairly lousy at that. Peter's insight, the insight of a lot of geeks that I work with is that you need to do some planning. You need a minimum viable plan. After that, the only way you're going to learn about reality and learn about what the customer wants or learn about how long your kitchen's going to take is to start doing stuff and getting valid feedback on the fastest possible cadence. So for me, the great geek norm of speed, velocity is not the right synonym cadence or clock speed or pace of iteration is the right synonym just get good feedback and then try it again 100 you know i love this because right now we're working a lot on all these ai possibilities and i have these young geniuses on my team and they come out of you know academia and phd programs their whole lives and there's this instinct to like kind of like over plan everything and talk about it for so long i'm like you know what guys actually let's just hire a few of these people and let's try building this and seeing how it works Then let's get feedback from the industry and let's try again.

11:07And like, but Joe, how could you start something when we don't even know for sure what's going to work? And I'm like, well, that's how you do it. You start trying, right? You start iterating, you start going. And that's, it's a Silicon Valley mindset to start building and trying things. I think it is very different than how the rest of the world approaches things. I agree. And I think it's one of the keys to your success. When I talk to industrial era companies about this, the immediate pushback I get is that sounds fine for building a consumer web application. But there are mission critical things out there and there are things that we cannot get wrong.

11:39And I say I completely agree with that. But you're making the mental error of smashing together the way you develop a thing versus whether or not the thing can fail. And to make that concrete, SpaceX is the only American company that has built a capsule that is authorized by NASA to take human beings up to the ISS and eventually onto the moon and maybe Mars, right? They're the only ones that did that. There's nothing more mission critical, I think, than keeping people safe in space. SpaceX is the only one that can do that these days. And the way they did it is not by planning a ton. And as you probably know really well, that company are fanatics about iteration and getting feedback.

12:23Meanwhile, NASA, which got a bigger contract at the same time that SpaceX did to build the capsule, they have not yet had their first crude test. Yikes. This iterative approach is the way to learn quickly and improve, even if what you're trying to do is very mission critical and failure intolerant. I love that. Let's move on to the second one. Write a lot on ownership. It means a lot of different things to you. Tell me what you mean by ownership. For me, it's kind of the opposite of bureaucracy. And everyone would nod their heads to that. Here's the tougher part. It's also, in some sense, the opposite of cross-functional coordination and process and structure and hierarchy.

13:01Now, I'm not saying that ownership equals chaos or equals anarchy, but it involves a lot less of all of those things than we became fond of in the industrial era. Joe, I don't know if you're old enough to remember the era of business process reengineering that really got kicked off in the early to mid 90s. I read a lot about it. It sounds terrible. But none of the fact that you just said that is interesting because to your ears, it sounds terrible. I cannot tell you how good it sounded to us, me included, when we first heard about it in the mid-90s. Well, you had to fix the processes, I guess, huh?

13:34There are too many processes. We're going to define in advance all the things this company is going to do. We're going to embed them in software, and everybody has the same sheet of music to play from, and we're going to make beautiful music together. And we did that. We did that for a long, long time. And we wound up with these, again, these very top-heavy, sclerotic kind of bureaucracy, barnacle-encrusted companies that found it very hard to move quickly or to kind of liberate their people at all. And what you and your geeky cohort are doing is figuring out how to build companies that are still very good at executing, right?

14:08But don't have that same kind of crazy overhead associated with it. Amazon is one of my favorite examples because they're a gigantic company now, but they still have this philosophy of like, kind of what's the minimum amount of production? They got rid of their NPI as you talked about, right? Because they used to do this horrible thing in the nineties. Oh yeah. Yeah. Yeah. This is, I didn't know this at all. Amazon was actually well on its way to being exactly this kind of top heavy bureaucratic mess. If you wanted to innovate at Amazon in the late nineties, you submitted your idea to this innovation bureaucracy that would pass judgment on it.

14:43And then they would tell you if you got to proceed, if you didn't get to proceed, but you didn't have to support anybody else, or if you didn't get to proceed with your idea, and you had to support a bunch of other ideas, this was just this absolute misery inside Amazon. Bezos realized that it wasn't working, that it was jamming the company up. And so in this great book, Working Backward, written by a couple of ex-Amazonians who were there, they said, okay, And Bezos said, to his great credit, my first preferred approach for managing innovation at this company is wrong. And we're going to go 180 degrees in the opposite direction.

15:18We're going to build a company where you don't need to ask anyone's permission and you don't need to resource seek from anybody if you want to try something new. I love it. And that involved completely modularizing and decoupling the technology stack. That's the birth of Amazon Web Services and to a large extent, the cloud industry now. And also this organizational decoupling. And what the authors of Working Backwards said is we learned over time that when we start, when we set this up and they had two pizza teams initially, the teams that worked the hardest to reduce their dependencies were the ones that were most successful later on.

15:52they quote bezos as saying if amazon is a place where builders get to build we have to cut down on the communication and sebastian thrun who's an amazing alpha geek said the same thing he said look i tell my teams it's not that don't ever talk to anybody but if you run if you run your idea up the flagpole and then back down the flagpole it's going to be almost unrecognizable by the time it comes back to you just go do stuff this is how steve jobs at innovation as well he'd create these groups that were totally separate and he wouldn't let anyone else reach in an airfare i think he i think the quote was that like, uh, if you try to, you have a committee design a horse, you're going to get a camel.

16:24And so you need to have, you need to have these, the groups be totally separate. And he'd have the ones only he could talk to. So it's, it's very interesting. It's a similar framework, you know, for all these guys. Yeah. And, and you know, Apple is a very secretive company in a lot of ways, but part of what they're trying to do is I think what you said, they're just trying to encapsulate innovative teams so that they don't get bogged down in all of the crossfire that goes on in a company, which was done with good intentions, but the end result wasn't great. So the business geek said, all right, we're not doing that anymore.

16:53I love how you talk about the bureaucracy mass index BMI. These things are fat for their bureaucracy. And the autonomy point really is interesting. You do need single threaded leaders who could focus and have people stay out of their way. One of the areas you talk a lot about for ownership and bureaucracy is about status. You talk about humankind, obviously all of us have this bias towards status. It's tied to our health, it's tied to everything about how we see ourselves. And apparently status helps create bureaucracy. Can you talk about that a little bit? Yeah, you bring up this really interesting point.

17:21So far, I've been dumping on the bosses, right? They're the ones who create these really elaborate processes and approval loops and all that. That's half the story. And I'm not even sure it's the bigger half of the story. Because one of the things I learned when I was researching the book is how important status is for us human beings. And I was actually surprised by it. I shouldn't have been. Because once you realize what we are, we're the most social species on the planet. We're the most social creatures on the planet. Of course, status is important to us. Social status is super tied up with our evolutionary, our fitness, our success.

17:55It matters to us like crazy. And one way you get status inside this thing called the company where you spend so many of your waking hours is to be in the loop. And so subconsciously, almost all the time, we humans, we want to be part of the process. We want to be in the loop. We want to be consulted either in a hard way or a soft way. We want to gatekeep. It's part of who we are. And what the geeks have done is they've channeled that desire for status and they channel it in ways that are aligned with the goals of the company. And they work super hard not to, not to allow these kinds of bureaucracies to encroach so that you can gain status in a way that's not aligned with the goals of the company.

18:37Just through doing books and podcasts, right? Yeah, exactly. Fine. Yeah. Or, you know, with the best of intentions, you say, I think our company which makes software for a living needs to emphasize sustainability. That's a perfectly reasonable thing to say. However, if you're spending five hours a day promoting sustainability inside your software company and that's not your job, are you acting in ways that are aligned with the goals of the company? Maybe not. And so some of the leaders that I learned from who I really respect are the ones that pare back the ways to gain status that aren't aligned with the goals of the company.

19:14One of the points you mentioned a lot is also, and I think it's related to this, is taking people's biases and their ultra-sociality and these biases and directing the cultural evolution of the company in the desired direction. So I guess you'd want them to gain status by doing things that are culturally relevant, right, is what you'd want to shape for. And aligned with what you as the person running the company want. Benioff had this great insight in one of his books. He said when he was a pretty senior person at Oracle, he didn't know how his activities aligned with the overall goals of the company.

19:44And he kind of asked his colleagues and bosses about it. And he said, Mark, don't worry about it. Just hit your numbers. And when he co-founded Salesforce, he said, we're not doing it that way. And you've probably heard of the V2 mom process at Salesforce, which everybody there does and participates in and kind of believes in. All it does is this cascade where you and your team have to say how what you're doing aligns one level up, one level up, aligns up right up to the top of the company with the overall goals of the organization. And my friends who work at Salesforce say, before you go into a meeting with somebody, it is very common to look up their V2 mom.

20:19So you know what their priorities are. That's really lightweight. It's just an example of what, you know, John Doerr would call an OKR process and objectives and key results process. It's simple, but it works. And that's how to keep everybody this atomized kind of swarm of teams at the company, how to keep them aligned and moving in the right direction. And on the science part, you kind of hinted at that, but I love your focus on quantifiable like things have to be things you can actually measure which is you know i wish the government was more that way too i think you can learn from this too you know we wish so many parts of our society were more like that and you can get too obsessed with numbers and if you're dumb about them then your metric turns into something that everybody tries to game uh but that's that's not a reason not to try to measure things and for me the heart of the the geek norm of science is something you said a little while ago it's channeling people's innate tendencies i think what you and the other geeks have realized is let's not try to change how we human beings are wired.

21:15That's really, really tough homework. Let's work with it. Let's work with our wiring. One of the aspects of our wiring, maybe the easiest cognitive bias to elicit, you can see it in any intro to psych course, is overconfidence. We are chronically overconfident creatures, mainly because, and this was a great insight from Robert Trivers, the most important evolutionary biologist since Darwin, Trevor said, the reason we are overconfident is because confidence is valuable for individual members of a very social species. If you believe that you're really, really good, you will appear that way to other people and they will want to have you as part of their coalitions.

21:53So we are overconfident. It seems like a bug. It's a feature. It's a feature for a species as social as ours. However, you will make lots of bad decisions if you're overconfident, if nobody checks you. And the point of the great geek norm of science is to have your peers, peer review in science, to have your peers argue with you and make you produce evidence. The channeling is that you then become an overconfident evidence producer. You're like, okay, I can do science. I'm a good scientist. I'm going to go do the incredibly hard, laborious work of generating evidence to convince the scientific community that my version of reality is the right one.

22:28The business geeks have imported exactly that insight. We're overconfident, make us overconfident in our ability to convince other people with evidence. Man, that's beautiful. It's interesting. Thinking about the evolutionary background so that I would assume that it makes more sense for men to be overconfident based on where it ends up for them in a society, which could be both good and bad, I guess. I believe there are. I haven't gone super deep into the gender differences. The biases that I'm talking about are pretty consistent across human beings. I think you're right. In general, men tend to be more overconfident and that might manifest in somewhat different ways.

23:05But any of the classic demonstrations of overconfidence, if you do them, you will get an effect with any gender mix in the room. One of my all-time favorites, I'm going to do this with a group someday. One of my all-time favorite demonstrations of this was they sent out a questionnaire to a bunch of people. They said, okay, how moral or ethical are you along these different dimensions? And of course, everybody says they're way above average. Then they sent those in. Six weeks later, they sent a follow-up and they said, okay, here are the averages across the whole group of how moral and ethical people thought they were.

23:37Where are you compared to the group? And they're like, yeah, well, I'm, you know, five to 10 % better all across the board, right? They just rated themselves as better. The fiendish trick was what they were reacting to was not the average from the group. It was their own answer from six weeks earlier. So people said, yeah, I'm much more ethical and moral than I am. We're just wired that way. Don't ignore it. Don't try to assume it away. It's very hard to train away. Maybe you could do it. What you should do is work with it. It's a good self-awareness insight. I want to ask about tech today. Obviously, a lot of what you've described is how we created Silicon Valley.