Transcript

Rich Ziade Congratulations to you, it’s 2027, you’re havin’ a drink and it’s—it’s a nice sunny day. You know, there’s climate change but we won’t get into that. So, good luck to you in 2027 [music plays alone for 18 seconds, ramps down]. Paul, I have a question for you. 

Paul Ford Alright, ask your—ask your question. 

RZ How come in some cases when you add more people you get more done? 

PF Umm—

RZ Meaning like, “I need to plow an extra 20 acres, just go [music fades out] get me 300 more people to, like, plow 20 more acres.” 

PF Well—

RZ But how come sometimes, most of the time, especially in our world, you add more people, and everything comes to a screeching halt. And it gets harder to do anything. 

PF Ok, what you’re asking about is actually the foundational problem in software development. It’s not code. People can learn to code and most code that needs to get created can be created by a pretty good programmer. It doesn’t actually—you don’t need everyone to be super genius. It’s all coordination costs. That’s the hardest part. Start with your plow example, right? That is an embarrassingly parallelizable problem. You can simply plow a section. You could say, “This is your section, please go plow it.” And an individual can learn that, go plow it, and return the yield or the harvest after the plowing is done and the field has grown. 

RZ What you’re saying is there’s no dependency. 

PF Well, there’s no depen—that’s right, you can just split it up. And actually, I mean, we can skip a few steps, right? This is the goal of the API, of the purely distributed software which is [mm], “I’m gonna give you everything you need to get your work done and you will never—you don’t need to coordinate with me. Go plow your field.” And a perfect example here is Google Analytics. You need to analyze parts of your website. You will never speak to Google. And it will analyze parts of your website. You will put a little Javascript on, and it will talk back to Google [mm hmm], and you’ll never have to deal with Google except through that interface. 

RZ Which leads to a form of parallelization. 

PF Well, what that mea—that’s—that’s your field. Google has said, like, “Your website is a little field, and here is how you plow it to get analytics into the analytics bucket and plant the seeds of growth.” So off you go. That part of the work is done. 

[2:35]

RZ Yeah [stammers] and I’m hearing you and this is why I kind of set this question up this way but, you know, before we hit record we said we’re not gonna talk about the pandemic. Whoever you are you may be listening to this podcast in 2027 [mm hmm] and you may not know what we’re talking about—[laughs]—

PF You get to go outside. 

RZ [Laughing] You’re havin’ a margarita. 

PF Yeah, God. 

RZ But I guess you applied it to software which is wise because it’s one of the most dependent—heavily dependent disciplines, right? It’s very hard to parallelize but the API is a secret weapon to parallelize software, right? 

PF Well, the metaphor that gets used a lot is that nine people can’t have a baby in one month. 

RZ Yeah. And it’s real. Right? I mean that’s—that’s—

PF It is real except that it actually is reductive, right? Because you can look for patterns of success or tools that already exist and apply them and so, I mean, the reality is that people come to us and say, “I need to have a baby in, you know, usually three months.” And we say, “We will but we’re not gonna start from scratch.” 

RZ So, I wanna get out of software for a second though. I think—

PF I don’t know how to talk about anything . . . [both chuckling] This is it, Rich. This is all I got anymore. 

RZ No, but when you add people, things slow down. I mean this is, look: we’re in the middle of the pandemic. We’re right now having a real hard time doing anything at scale. Right? It’s hard to watch—

PF Oh as like a nation, as a culture. Sure. 

RZ Well, no, as a like, “I need glass vials.” Or, “I need Q-tips.” 

PF Well, or, “I need to vote.” Like all—the existing processes and supply chains we have don’t scale to this situation. They’re actually not robust. 

RZ They don’t. They’re not robust and, you know, I think back to—you know [stammers] it probably was really messy during World War II. Like, it probably was terrible, everybody’s—all we have is footage of people building bombers in car factories, right? And it’s just the most inspiring thing—

[4:30]

PF It was four years of—

RZ It was probably a disaster! 

PF —profound global despair mixed with—Oh yeah, all the stories of like individual soldiers and stuff it wasn’t—it wasn’t that heroic. Everybody hated everybody else and nobody wanted to die, right? 

RZ No, no, but I’m talking about the production. Like the production machine was like, “We must mobilize and inspire. You know, bring the real might of American manufacturing to bear and here we go.” And the next thing we know we’re building so many weapons and bombers and bombs and it’s like, “Whoa, you know what? We could do this all day long. We might not even have a good battle plan [laughs], we could just keep making these.” 

PF No, no, it was an utterly reproducible process. It’s the same as, “Plow the field.” “You go over here, you’re making cockpits.” 

RZ Alright, man, but it’s 70 years later and I can’t get Q-tips! What happened?

PF Oh, this is—ok. This is my theory: thanks to the immense productivity gains brought by the digital revolution and the way that we approach things through globalization and the way that we have structured our businesses, we have created incredible efficiencies just in time, delivery. You know [mm hmm], no one holds any inventory. None of that, right? 

RZ Mm hmm. 

PF And so, the days of everything being vertically integrated. Like, “Oh, you know, we own our own cocoa farm and the beans come in a boat that we own and goes—” Those are all gone. And instead you have this mutual set of middlemen with unbelievable efficiencies, except that all of those efficiencies are utterly based on long-term stability in the markets, the climate, and the world in general. 

RZ We could just end the podcast right now. The observation you’re making here is profound. 

PF Well, I’m stealing it from one of our clients who’s got a global warming charity, and his point is that we have counted on stability for all of history. The last thousands of years and that [right]— you can’t really see stability. Like, stability is—it’s like seeing economics, it’s the air, right? And so you take it for granted, of course it’s stable. You have a different point of view about this. This is one of the things I’ve learned from you, and this is literally your heritage and you coming over in a very unstable situation and having an unstable situation. Growing up, you see a lot more randomness in the world than many entrepreneurs do. I see chaos—Until we started Postlight I was never really great at translating into business but I could see the chaos and I’d be like, “Ah, that’s a lot of chaos. I don’t know why people think that’s normal.” [Rich laughs] [Chuckles] I’ll give you an example: the stock market is crazy cakes. It is a banana cakes set of human wishes and dreams that everybo—When people talk about it, they talk about it like it’s a large, rational animal that knows what it wants. 

RZ Mm. Yeah, yeah. 

[7:07]

PF It’s like, “Oh, stock market wants to eat more bananas.” Right? 

RZ Yeah. 

PF And then blockchain’s even worse. And all of these things, all of these assumptions, all of the way we talk about this stuff is just built on this understanding that it’s gonna stay roughly the same tomorrow. 

RZ Yeah, well, look: there is value in those markets. I believe in those markets. You know, I’m pro AirBnB and pro Uber and pro finding efficiencies. Again, pro in principle. I mean, you know, some of their practices are gross but in principle I like that, you know, we can use technology to find those efficiencies, right? That’s very cool stuff to me. 

PF Also, no matter how politically different you and I are, we both essentially believe in a concept of growth. Otherwise we wouldn’t start this business. Not even a software business. Just a business. 

RZ Right. 

PF That growth is possible that growth can have, you know, productive effects, and that giving people jobs and asking them to do things for other people is meaningful, right? 

RZ Yup. 

PF Like, that’s—It’s part of the DNA. It’s very American. 

RZ Yeah. But I wanna follow your thought through cuz I think it’s a really important one. Ok, so, here we are: we rely on stability for what I would call all of these sort of—you know, this hyper-microtransaction world that we live in, right? Like if I—I broke my dumbbells like two months ago. 

PF Just workin’ out so hard. Just a beast. 

[8:23]

RZ I dropped it. I dropped it. 

PF That’s disappointing. Would’ve been [chuckles]—

RZ Yeah, I crushed it through the ceiling. Yeah, exactly. [Paul laughs] No, and so I needed a part. And I was like, “Oh my God, this is impossible, they’re 10 years old.” They are these adjustable dumbbells, so they have the like click system so you can get different weight.There’s no way! I go on eBay [mm hmm] and I find an individual. Clearly it was not a company or a shop, just an individual who happens to have—has amassed these parts for the gear systems inside of the dumbbell. 

PF Mm hmm. 

RZ And it was—he charged me 18 bucks for a piece of plastic but, frankly, I would’ve had to spend 300 dollars to get a new dumbbell system, right? 

PF It’s worth 18 dollars. That’s fine. 

RZ Worth 18 dollars! But the thing was the fact that I was able to even go do that, and he was able to find a market where he was like, “You know what? I’m gonna load up on these cuz I think people are gonna break ‘em all around the country.” The dumbbells I have aren’t made anymore.

PF Yeah. 

RZ That’s another thing worth pointing out. So it was this incredible sort of ability to a) to find someone; to seek that buyer—that seller, I should say . . . is incredible. Right? Like, just let’s pause for a moment there and just point out that I found this guy in the corner of Idaho who was selling this piece of plastic. 

PF And, let’s also point out: it took you about probably a minute. 

RZ It took me about a minute! It took—I mean I didn’t know if I should—First off, I thought it was—I didn’t know if it was right or if he was legit and all these questions came up. And one of the things I’ve noticed with the pandemic is that everybody went to eBay kind of. Everybody thought, “You know what? We’ve got this—these incredible marketplaces,” but what you found was you were sending 6,000 masks . . . to a hospital of 80,000 people that needed new ones everyday. 

PF Right. 

[10:07]

RZ And the scale challenge that kicked in was just utterly profound, and we found ourselves back at like, government provisioning. Right? Like when the government orders cheese. Like, we’re back there. 

PF You know what I’ve noticed about our world? The world of tech, which is that we’re very good at creating one-to-one markets and there’s some good tools for communities but [yeah] cultural change is hard and requires large state actors and the systems that we’ve built don’t really support that sort of thing, right? Like, they just don’t. 

RZ No, they don’t. 

PF And here’s an example. I’ll ask this as a question. Let’s see if you think the same thing I do. How—when—governors have formed blocks throughout our nation right now, right? 

RZ Yeah, yeah. 

PF Five, six states. How did the governors coordinate and work together? 

RZ Probably call each other. 

PF Conference calls. 

RZ Well, first off, conference calls and probably geographic proximity cuz it made sense to them. They probably know each other because of interaction they’ve [oh, they do] had around the border, right? New York and New Jersey talks to one another a lot. They share budgets, there’s all kinds of interaction. 

PF Absolutely and then, you know, they call each other. You have a conference call and you talk about what you need and then you turn that into action because basically the governor is the middleman slash—the governor is the marketplace at that point. Like they are the arbiter and they define [yes] what goes what and who goes what. And there’s no system that you could automate with current technology, or maybe not even with general artificial intelligence that could deal with that. 

RZ No. There isn’t. There isn’t and it’s talking and coordinating and frankly, it was actually a little infuriat—Imagine if the governors had to coordinate to create those bombers in World War II. Disaster! You wouldn’t have had a single airplane. Right? Like I mean that would’ve been—I found it—

[12:02]

PF The entire economy was mobilized . . . in order to generate the effect you’re describing which is, “I’m gonna put this person in charge. They have a certain amount of time. They will deliver this. [Mm hmm] And then if they don’t, I will have to get someone else.” And it just fanned out, like it was optimal—you know, you have . . . that sort of McNamara type brain. 

RZ Yeah. 

PF Around—around production and like, what do we need to do to systemize? Humans didn’t factor in a lot. It was literally like humans were units of labor that required calories and sleep, right? 

RZ Yeah. I think what’s interesting here is that, you know, they advertise efficiency: these sort of micro transaction markets, like Uber. I mean Uber is incredibly efficient, right? Like it’s constantly pricing cars and telling you what’s nearby so you can get one at the—And it lets the drivers compete and the more drivers, the price goes down. It’s almost beautiful. 

PF Well, it’s—I think because it’s Economics One-oh-One [101]. Right? The only problem—

RZ  It’s Economics One-oh-One [101]. No, but I think what happens is when you get to scale those markets, a) can’t meet that level of demand, and b) they’re actually incredibly inefficient. What we found is that I can’t cobble together 6,000 orders of Q-tip boxes. I need one giant order! Right? 

PF Like, that’s—Uber isn’t the real economy, right? It’s heavily subsidized through stockholder money, right? So that part’s really tricky—is that the bill still hasn’t fully come due. I think AirBnB is a more efficient business overall but like Uber—

RZ No, but I think Uber can stop the spigot and be profitable. Like they could do that anytime. They’re just continuing to grow—

PF Not at the scale that they’re at. It’s like Amazon, right? They’ve got that Amazon strategy where it’s like, “We will continue to grow, at all costs, with all money available.” Except Amazon tends to use money, what you’d usually give out as a dividend. Whereas Uber is just sort of—it feels like it lives on borrowed time. But I basically agree with you. Like I just like—It’s worth noting that that one is at more risk than many others because of the way that it counted on the demand for people driving in cars to continue to go up. 

[14:13]

RZ Yeah, I—I don’t know. I think their growth was geographic rather than like, “What can I suck out of New York City?” In any case I think it fails! At this scale, I think there is a point in time—

PF Wait, clarify: fails to do what? What does it fail to do? 

RZ It fails at a massive scale. eBay doesn’t work. 

PF Well, it works if you wanna order an old book or get a part for your dumbbells but it doesn’t work to get—

RZ I guess let’s go back to software. Let’s bring this full circle. The coordination around the testing and I’m not a scientist. There are parts. There are different chemicals and parts and pieces that you need. So everybody’s like, “Oh if I could only get the machine.” It’s not about the machine. It’s about the materials you need, it’s about—And if you divvy out all of those pieces, right? That’s where plowing the acres breaks down, right? Because there are dependencies now. There’s the Q-tips, there’s the chemicals that you need for the tests to be done, there’s the sensitive machinery that has to get produced, and even the sensitive machinery probably has tens to hundreds of suppliers that are supplying parts just to the machine itself: the LED screen that is on the machine is coming from someone else, right? That coordination, that mobilization that you need, there is a point where you now have left the realm of, I think, capital markets and into the realm of, frankly, overarching mandate. 

PF But what can work at that—There’s only like three or four organizations that are non-governmental that are integrated at that scale. Like Apple is one. Apple can—knows it’s—

RZ Apple is one. 

PF Yeah. Tim Cook understands the supply chain top to bottom and he thinks about each component at some abstract level. He’s like, you know, “This little tiny WiFi antenna that goes into an iPhone, how can we shave five cents off of that and get it made and delivered [mm hmm] in one less day?” And God knows if somebody can pull that off inside of Apple, I bet you get a big promotion. 

RZ Yeah. 

PF Right? So that’s—

RZ I think that’s right. 

PF That’s the scale of organization necessary to do this kind of thinking and who the hell has that experience? 

RZ Nobody. Very, very, very few companies do. I mean even a Boeing has probably thousands of suppliers. 

[16:26]

PF So does Apple. Everybody does. So it’s like how do you—

RZ Everybody does. 

PF Yeah. And so like, taking that knowledge and skill, grafting into government . . . which at a [chuckling] federal level right now just isn’t gonna happen. It’s not—we’re not a government that respects practice very well, very—Like, [yeah] at this point, like, the person—That is, to me, when I look at Apple, that’s what I see. I actually don’t see an enormous amount of product innovation cuz it feels like they’re on their roadmap. I see the best possible supply chain optimization at scale—

RZ Efficiency, yeah. 

PF Yeah. 

RZ Yeah, that’s right. 

PF With a kind of innovation backbone driving it forward, right? Which is why Tim Cook got put in charge of this giant ass company instead of Jony Ive. 

RZ Yeah, I think that’s right. And to close this out, I wanna just share an observation that I have and again, it’s six years from now and you’re listening to this podcast and God bless you but let me tell you something that happened in 2020. Like, to your point earlier, like eight governors got together to essentially say, “We are a single purchaser.” 

PF Yup!

RZ Which was a very antimarket thing to do. And I found it actually kind of depressing and infuriating. And I’m as—I’m as pro-capitalist as you can be. But the fact that that even needed to happen isn’t because they were countering evil. Because commercial business—which are under enormous stress and pain right now wanna make money and put money in the bank, right? What the governors were saying is, “Look, we can’t go into a marketplace right now and compete with each other, that’s bananas.” Right? But the fact that that even needed to happen across eight out of the 50 states, is sad to me. 

PF Well, don’t forget the side effect of all of your World War II manufacturing was vigorous price controls throughout the entire system. 

RZ Right. 

[18:09]

PF You know, FDR on down, like, you know, you could only charge this much for steel. I mean it was just—That was—

RZ Correct, correct. 

PF So that entire system at a federal level, they basically put a hold on capitalism. And that’s one of the ways you got that efficiency. 

RZ But you know what? I—I say this partly because it’s sad that, you know, this is about money. But what’s sadder to me is this hard! To your point. They had to get on conference calls, they could only get eight of the states cuz the ninth governor was busy or whatever [Paul chuckling]. Like this is—We’re back to square one, we’re back to software development. We’re back to all those dependencies. Who—you know how many project managers are probably coordinating across those eight states?! Probably like 50 people are just talking to each other with flow charts and workflows and all kinds of things to even coordinate! 

PF Well, and also, governors don’t manage by what’s possible, they just tell you what needs to be real. 

RZ Yeah, absolutely. Right? They wanna have the press conference and tell the story, right? 

PF Those are the scariest clients for us. Like, [yeah] the ones who come in and are like, “On next Thursday, I will have this.” And you’re like [Rich laughs], “Ok? But you just described a rocket to Mars.” They’re like, “I don’t care what I described.” 

RZ So, look: I think—I think a lesson learned from building a lot of software is that sometimes you have to force things to align and not wait for markets. And not wait for—in the case of software—a lot of independent actors with their own agendas and more or less levels of urgency. Sometimes you just have to come in top down and line it all up. Right? And that, to me, is how you force things to click together versus “Hey, can you get that thing to Tom, please? He’s been waiting for four days.” 

PF Listen: everybody likes a autocratic leader in times of terrible crisis. 

RZ So, you’ve got two great quotes on this podcast, Paul. 

PF One of the best things you can do if you are someone who is an autocratic leader [chuckles], is create a lot of crises, right? 

RZ Yeah, that’s right. 

[20:05]

PF Cuz you get to solve it. In this case, we have a real crisis but you see that pattern and then when things calm down, everyone’s like, “We’re all equal and we’re all working together using—” 

RZ Very good point.

PF You know, interrogating those realistically, right? The thing with—Bringing it back to software development, the autocratic, top down approach can really work. The key thing that you miss with that is everybody gets into a state of like, “I have—I have to get the leader what they want,” and they stop looking around for things that could truly excelerate. You know, it’s just like, “Let’s just get our—Put your hat on, we’re goin’.” As opposed to like, “Well, could we just use this thing over here?” And so—

RZ Right. 

PF Don’t ever forget that. Like, the reality of software development in 2020 is that you actually can have—nine people can have a baby in a month if they know what the baby’s supposed to look like. 

RZ Right, right. 

PF So you gotta—it’s changing. That part of our world is changing. It is fluid. Like, what’s interesting in the paradox of everything we’re talking about here is that . . . there’s not just that one way. There’s a lot of ways, they have to all be on the table, and it’s very fluid. That the very things that you’re saying, like . . . you know, these tools that we’ve built aren’t really serving the needs that we have at a large macroeconomic scale. It doesn’t mean that two months from now somebody might not say, “Hey, I hooked my Excel spreadsheet up to this and we can run this part of the economy now.” Like, it’s—I guess the way I would close this out is we’re seeing new needs that we didn’t see before. 

RZ Yeah! 

PF And we should address those and think about them. And you don’t have to think about them ideologically, you don’t have to be like, “Well, if it doesn’t [music fades in] serve the stock market, why bother?” Like, there are just needs that humans have and we should all be thinking about them. 

[21:43]

RZ Agreed. 

PF Alright, Richard, well, you know, if anybody wants to get in touch with us, what would—

RZ If you wanna build great software efficiently—

PF But, here, hold on—

RZ With near zero dependency! Call Postlight! 

PF Listen, we’re your partner. We will work with you to get your new thing built and launched and out and it will be a big technology thing, and people will love it. And if you wanna get in touch with us just send an email to hello@postlight.com. That email address goes straight to Rich, me, and other folks, and we wanna hear from you and we wanna learn about what you’re doing and we would love to be workin’ with you in a room for three months, six months, a year. We’re ready. Just get in touch. 

RZ Well put, Paul. 

PF Thanks. 

RZ You’re on a roll today. 

PF Well, we’re just tryna—We are your partner. We have to start sayin’ that more and more cuz that’s important. 

RZ Yes. 

PF Alright, friends, see you next week. 

RZ Have a lovely week! And hey, guy in 2027, enjoy your dinner. 

PF [Chuckling] Enjoy—enjoy your cricket protein burger. 

RZ [Laughs] Have a good week. 

PF Bye, everybody [music ramps up, plays alone for three seconds, fades out to end].