The Exchanges

Q a question about that because one of the things that I've always thought about when it comes to this tension between academia and industry, you might hear about needs and requirements from industry, but when you're in research environment, you have unlimited open-ended, you know, vision and an algorithm and a research paper is very different than something in production at industry scale. So how do you guys navigate that?

A First of all, students, they do internships, they understand It's a problem. They develop the first solutions. And then with that understanding, you can start more principled designs to build these systems. But the truth is that in every successful project, we had at least one or two partners. For instance, uh, for, um, we work very closely initially with Facebook. When we started working with Facebook, Facebook, you know, has an entire, uh, cluster, big cluster for big data. It was 80 nodes. And their big data team was like three people. Then, for instance, in the case of Mesos, we work very closely with Twitter. And Hinman went to Twitter and worked very closely with Twitter engineer to deploy Mesos in production. And actually, the feedback from Twitter has a big impact on the Mesos evolution. From just supporting big data, cluster computing frameworks like Hadoop, it went to support These long running services.

AI assessment note: “in every successful project, we had at least one or two partners.”

Q We're talking about the manufacturing and the productivity slowdown, but where does this idea of people being able to buy more because of the cheap products that are coming into the market fit in? Because while people may have had a hard time finding jobs, they can buy more with less.

A We do that by trying to adjust for inflation. So if we have very slow inflation, prices aren't going up that much relative to wages. What we'll see is What we call real wage gains. If you get more money for your buck, then your real wage is going up. It turns out that real wages in the United States have been pretty close to flat. Now they've risen the last couple of years, but in the 2000, real wage gains were very, very anemic. Here's where it gets tricky. There's many different measures you can use for inflation. So how much should the price of healthcare weigh in that? How much should the price of rent weigh in that versus the price of a TV? It's very tricky, and you can use some inflation adjustments that show a pretty strongly rising real wage, or that show essentially no wage slowdown in the 2000. You can use PCE.

AI assessment note: “We do that by trying to adjust for inflation.”

Q Did you guys tell me the truth? Did you guys have like a magic number in your head before you started those pricing discussions? Like, did you think in your head, you know what, when I go to sleep at night, I want 25 dollars when this thing goes on the market.

A No, we, we, we, we, you didn't. Part of what our strategy was, we're going to do a teeny little IPO. And then if it went well, we're going to do a Pretty big secondary. And so the company was much more focused on make the secondary successful than it was make the IPO successful. Part of making the secondary offers successful is you need a couple deep pocket people in the IPO, even though it was a teeny little IPO. So our, one of our leading shareholders ended up being Will Danoff of Fidelity. And so we say, we'll invest out of your ten billion dollar, whatever it is, fund four million dollars. And he's like, I don't have the time To read your earnings release at that level. But we, we convinced him to come in because then in the secondary, he was able to back up the truck and he got what he wanted, which was a large ownership allocation. His IPO allocation was what? Five percent of 70 to four million dollars. Some number like that.

AI assessment note: “No, we, we, we, we, you didn't. Part of what our strategy was”

Q What are we actually comparing to as we're thinking about these? Well, this is not good enough. This is in our, what are we holding up as what we want, you know, something that we'd prefer it to be?

A The end of the post-war boom and the return of the ordinary economy. The story I'm telling is that the quarter century after the war was an unusual period of very rapid economic growth. The period from To 1973 was probably the period of the fastest economic growth in the history of the world. GDP around the world grew at more than five percent a year. Now at five percent a year, something doubles in 14 years, quadruples in 28 years. So even with some population growth, people's incomes were growing very rapidly. People's living standards were rising in a way that was Was visible to them. They were able to buy houses for the first time, and cars for the first time, and send their kids to high school, and maybe even college, and, and we had all kinds of, of very rapid advances in living standards.

AI assessment note: “The end of the post-war boom and the return of the ordinary economy.”

Q internet where you have this infinite shelf space to, to find them. But the second half of this, the fact that it went beyond a niche to a little bit more mainstream interest is I think really interesting. And besides nine 11, like what are some of the other epics in policy and political history that Have kind of led us there. Like, would Snowden be added to that list?

A So Snowden, Snowden is actually the, the second to last of these. So first, a huge percentage of the site was about the law of detention. The issue at hand was Guantanamo and the sort of developing law of detention, and should you close Guantanamo? Under what law should you hold people there? Under what law should you hold people somewhere else? Over time, that set of issues faded. The salience of the issue diminished. The parameters of the dispute narrowed, and the capacity to get the final thing done was reduced, and reduced to the point that it wasn't even clear how meaningful it would be if you got it done. You close Guantanamo, you move 40 detainees to the United States, you're still holding the same 40 detainees under the same legal authority. So just as that set of issues sort of started to fade away, it was replaced In the public's, in the, in the debate with issues around drones. And by the way, there were multiple aspects of that debate. There was a targeting debate, right? You know, when is it legal to kill somebody, like a different continent? When that person is in a war zone fighting you, but also when that person is not in a war zone, maybe not fighting you, but you think he's a terrorist who's going to. That's a very complicated set of legal targeting questions. But then there's also the, you know, the flying scary robot question, right, which is, you know, does…

AI assessment note: “Snowden, Snowden is actually the, the second to last of these.”

Q How'd you know that Opsware was where it would be at? Like that there was something there. I mean, was that just like a guess in the dark kind of thing?

A So we had some signs at the time all this was going on. I'd actually moved to New York to, um, run our field operations and implementation team. And we w we would repeatedly have these conversations with, uh, with prospective customers and we'd pitch them on this whole Big, beautiful cloud services vision about how they were going to get more efficient and save costs and move faster. And every single one of those meetings would end the same way, which they, they would say, I like your people. I like your services. I like the other customers that you guys already have, but everybody else in your, in your segment has already gone bankrupt. So I'm sure you guys are going to too. And then they, they, they'd pause for a second and they'd say, but You have this Opsware thing that you talk about that you use to run our, your data center. It's like, that sounds really interesting. And so for, there was probably a, about a six month period where we would just kind of laugh at them and say, no, we're, we're a cloud services company. We're not a software company. And that, that's, that's part of where Oxide was born from was the realization that maybe there was actually a gem hidden in, in what we were doing.

AI assessment note: “every single one of those meetings would end the same way”

Q So given that shift, what does that mean for the current and future of designing these new cars? Essentially, these cars that are no more software than they are people caring about the mechanics.

A Well, one of the big changes, especially if we're going to switch to fleet ownership, is the car manufacturers will start being a lot more like Boeing and Airbus than they will be like Ford and BMW today. So if you think about Ford and BMW, they spend a lot of To establish a brand persona in the buyer's mind, right? BMW is the ultimate driving machine you hear on the radios. You've got billboards. Boeing and Airbus don't do that because they know that who buys the plane is the operator, the fleet operator. It's United and Lufthansa. And so they market to them, right? So the things that matter to people in terms of differentiating your car will no longer matter. Basically, they need to win Lyft and Uber.

AI assessment note: “car manufacturers will start being a lot more like Boeing and Airbus”

Q So before we break down some of the verticals, what does that mean when you say deregulatory? Like, what does that mean for startups? Is it a good thing or a bad thing? I mean, I don't mean to be like so black and white about it in moralistic terms, but how should an entrepreneur think about that?

A Speaking as the Republican, I think it's almost always a good thing to have a deregulatory environment. It gives more freedom to the entrepreneur to, to build his business or her business in the way that, that they see fit. And To experiment with, with new forms of new products, new forms of reaching consumers, new forms of giving disclosures, alternatives to traditional government. A lot of regulations stems from a desire to promote health, safety, general welfare, present and prevent disasters. Many tech companies figure, well, we've got new ways of preventing that. The great example is ride sharing, of course, where you say, We could have a taxi commission that prescribes a bunch of rules that, that, that, and puts a number in the back of the taxi cab that you can call if you have a complaint, or you can just do this app and have crowd, you know, have ratings by drivers and users. And if you think you got ripped off on the way from LaGuardia into the city, then you can, instead of trying to remember, if you don't know the area, instead of trying to remember what route you took, Lyft and Uber have it on the app and they can retrieve it and say, yeah, you were.

AI assessment note: “I think it's almost always a good thing to have a deregulatory environment.”

Q So before we break down some of the verticals, what does that mean when you say deregulatory? Like, what does that mean for startups? Is it a good thing or a bad thing? I mean, I don't mean to be like so black and white about it in moralistic terms, but how should an entrepreneur think about that?

A Speaking as the Republican, I think it's almost always a good thing to have a deregulatory environment. It gives more freedom to the entrepreneur to, to build his business or her business in the way that, that they see fit. And To experiment with, with new forms of new products, new forms of reaching consumers, new forms of giving disclosures, alternatives to traditional government. A lot of regulations stems from a desire to promote health, safety, general welfare, present and prevent disasters. Many tech companies figure, well, we've got new ways of preventing that. The great example is ride sharing, of course, where you say, We could have a taxi commission that prescribes a bunch of rules that, that, that, and puts a number in the back of the taxi cab that you can call if you have a complaint, or you can just do this app and have crowd, you know, have ratings by drivers and users. And if you think you got ripped off on the way from LaGuardia into the city, then you can, instead of trying to remember, if you don't know the area, instead of trying to remember what route you took, Lyft and Uber have it on the app and they can retrieve it and say, yeah, you were.

AI assessment note: “I think it's almost always a good thing to have a deregulatory environment.”

Q was in Uganda and they were kicked out During the seventies with Idi Amin, my mom's sister went to Sweden and my grandparents went to the UK. And I think that's the other side of this discussion. You mentioned the impact on receiving countries and communities. Can we talk a little bit about that? Like what some of the fears and concerns are and what's top of mind for folks?

A If you visit Lebanon, Jordan, Turkey, these are the countries that have received the most Syrian refugees because, as Lina said, they're, they're the neighboring countries. And put this in perspective, in Lebanon today, somewhere between a quarter and a third of the entire population is a Syrian refugee. There are more Syrian children in Lebanese public schools than there are Lebanese children in these schools. Imagine the burden that this puts on these countries. Transpose it to the United States. The number of refugees in Lebanon, if you transpose that to the United States, it would be the equivalent of us taking seventy million people. And you know the debates we're having about taking 10,000 Syrian refugees. So these countries are bearing a very significant burden. And Lena's exactly right. Ultimately, the way to end this challenge and problem is to end the civil war in Syria and to stop the bombing. And we're working 25 hours a day Trying to do that, even just getting a cessation of hostilities to take that pressure off and get humanitarian assistance flowing. But as that's happening, what's driving them when they get to Jordan, Lebanon, and Turkey are two things. One is access to education. They want their kids to be in school. And second is the ability to work. What you're seeing is so many people and so many middle-class people who've left take their savings with them. …

AI assessment note: “somewhere between a quarter and a third of the entire population is a Syrian refugee.”

Q of doing things. Part of the contestability and skepticism is about Criticizing and not taking what's given as a given, but there is also an element when you're building a company or if you're pushing forward an idea back then to now where there's also a collaborative element where there's trust that goes with that skepticism. So what is the role of trust in pushing forward this engine of creativity?

A That's a really good question. We should realize that much of the knowledge created at that time is created by single individuals. There's not that much Collaboration, you know, co-authorship, and these are multiple names on papers and books that we see today, that wasn't very much, that was quite uncommon in those days. But what happens is that, you know, knowledge has to be circulated, it has to be distributed, and what is created in Europe is a virtual network, and that plays a very important role in this book, which is known as the Republic of Letters. The Respublica Literaria, as they call it in Latin. And what it is, it's a sort of a virtual network. It's not a formal organization. There is no bricks and mortar institutions involved, but it's based on a network of people who write letters to each other, who correspond. They publish books and they read each other's books. And the reason that trust emerges is because that everybody knows that if you publish something or you, some discovery, some, Mathematical theorem that you have proven, or some planet you've discovered, or some new species that you found, that others will take, will look at it, because they know that other, this has been vetted by other experts, and so if it survives, that means you must got it right. Trust is created By the fact that experts and specialists and learned people talk to each other, communic…

AI assessment note: “Trust is created By the fact that experts and specialists and learned people talk”

Q So what are some of the differences when you don't have two Fresh new startup candidates or products, and you have like, say, a David versus Goliath type of situation. What's some of the differences? I mean, both politically and in business.

A I've been on both sides of this. I mean, I was on the, on the incumbent side with President Bush, and I was on the other side of this with Governor Romney. The parallels between oh four and 12 are pretty stark. I mean, it is a huge advantage to be the incumbent. Um, the president has a ton of money and has been building up organization for four years, has been road testing that organization in a smart way. One of the biggest criticisms, for example, of, of the 12 race of, of mid is, is, is, uh, digital and the investment in infrastructure. Well, the problem is we didn't have the money to invest in infrastructure in both field and on the data side because we were broke, you know, and we had four months over the summer to try to redo all that sort of stuff, to build out a field, full field program, to invest in data and technology. Well, four months, When the president and his team had been spending four years, in fact, eight years running for president, and they'd done a great job, and they used that to their advantage. Data and technology and field operations mattered. The Obama campaign had plenty of lead time to build that organization out, to road test it, and try different things, and one of the things they did really smartly was they learned about this whole concept of virtual precincts, and generally, precincts are neighborhoods, right? You've organized, you have one pers…

AI assessment note: “it is a huge advantage to be the incumbent. Um, the president has a ton”

Q guys have definitely answered why the shift is happening, like a more systems approach, things are more complex. That's kind of obvious to me. The complexity pretty much invites computer science as a, as a point of Doing things that humans cannot calculate. But what are some of the manual versus automated things that you guys described as then and now, like what's changed in the lab and people's practice?

A So a lot of my PhD work was in a field called computational biology, which is in some sense a halfway house between completely generalizable machine learning approach to answering these questions and just looking at single genes. Computational biology is really a field that's designed to use computational tools aided by the biological knowledge of the scientist that's wielding those tools. So some of the things that you can do in computational biology, for example, is take the DNA data that's coming from an Illumina sequencer, for example, tremendous amounts of data. You're going to have to go through essentially prepare that data before you can do anything with it. That step is called DNA alignment, and that used to take days. Now there are tools out there that can do the alignment process in five minutes. So that's some of the things that that field has contributed to in the last 15 years. Where I think computational biology could use help from things like machine learning is making sense of what that amount of data is actually saying about how we understand disease and how we understand human health. And that's something that manually is very hard to do because let's say you can ask what is the RNA expression level across those 20,000 genes. That doesn't answer the question of which genes are relevant For determining whether somebody has a certain disease or not. It only tel…

AI assessment note: “DNA alignment, and that used to take days. Now there are tools”

Q now been acquired by Snapchat. Like we have stickers everywhere in these different ecosystems. And one of the arguments that we made about emoji is that because it sort of translates interoperably, you know, unless of course, like Apple and Twitter have their own proprietary emoji sets, that's not necessarily true of stickers. So how do you smooth this communication across all these different platforms and have these custom folders?

A Yeah. Well, at least on WeChat, there is this ability to import any image from your photo album on your phone and make it into a sticker. And they have these automatic functions that strip out The background to make it look as close to a sticker format as possible. So you are able to actually import images and graphics from all over the web. So for example, a lot of the line characters, which is include this very popular brown bear called Brown and a cute little white bunny called Kony. A lot of those characters, I've seen people convert those images into WeChat stickers and actually use it on the WeChat platform. So you can really take an image from anywhere.

AI assessment note: “there is this ability to import any image from your photo album on your phone”

Q There are so many interesting threads to pull on that one. So just starting with, first of all, the question of why celebrities? Like what, what does that give the person who's doing that? Like what does it provide them?

A I kind of think that it, there's this funny interplay, right? Between like, I want to show you that I'm expressing this emotion. And so, um, I'm going to show you this very exaggerated form of the emotion via this celebrity expressing that at sort of The peak. It's the same way we use reaction GIFs, right? Where you see people using GIFs of, like, Nicki Minaj giving a side eye, um, like, you can't give a side eye as good as Nicki Minaj can, so you want to express that. It's like a pure form of the emotion, but at the same time, I think those faces and those expressions then become an emotion that you now have, if that makes sense, right? So you're like, oh, I feel totally Nicki Minaj side eye dot GIF about this, um, and so It actually shapes the way people talk to each other in a way, and eventually it becomes codified in a really interesting way.

AI assessment note: “I'm going to show you this very exaggerated form of the emotion via this celebrity”

Q Do all marketplaces have network effects by definition?

A So I don't, I don't think that all marketplaces have network effects by definition. They have the potential, but different marketplaces are different at different stages of development in their evolution. So a marketplace that is new and just begun and say it's six weeks in or 10 weeks in, they don't have sufficient liquidity. Someone could start that down the line or even a few weeks from now and they may have Stiff competition. A classic example, and we've talked about this before as well, is Airbnb. The first three years of Airbnb was a real slog, because it's a global marketplace, um, and they were trying to build supply. At the same time, they were also trying to build demand, and they needed to sign up the homes. They needed to make sure that people were, people trusted the marketplace and felt secure to make a booking. But three years into, until then, the growth was really sluggish, and you can almost see it in their graphs that they share. The Room Nights book just started growing enormously after the third year because they had sufficient liquidity, both supply and demand, and it started working.

AI assessment note: “I don't think that all marketplaces have network effects by definition.”

Q and getting help because he's so engaged in the game. And even this lovely Moving story about an animal shelter being out of dogs because people were using them to walk. So they don't look like idiots playing the game. So they had an excuse to walk their dog and actually catch Pokemon. It's just amazing. And what's, what's happened now to make this happen? Like what went into this?

A I think what's super interesting is that Ingress launched several years ago by the same, same team, and it was relatively successful for compared to most apps and services, but not nearly as successful as Pokemon. And part of that is just the branding itself. And Pokemon has been around for a few decades. It has this nostalgic factor that I grew up with and so many people around my age grew up with. And it's also something that's, that's extremely accessible. Uh, you know, you see kids playing it to even, uh, you know, adults and you'll see these people running around on the streets now holding their phone in awkward ways and you could tell they're playing Pokemon. And so it's had this interesting, uh, nostalgic factor and just accessibility that's reached such a wide audience. Whereas with most, uh, Big social platforms. Let's take Snapchat as an example. They started off really focused on a younger teenage audience, and while they're growing over time to a wider audience, initially, like, my parents would never use Snapchat on day one. Now they are, but it took several years to get there. So I think that's super interesting mechanic of, of Pokemon Go.

AI assessment note: “part of that is just the branding itself. And Pokemon has been around for a few decades”

Q So what's the difference, again, between AR and VR? Could you define it for the audience?

A Virtual reality is, you know, what we're seeing coming out of Oculus, which was acquired by Facebook, Valve, um, with the HTC Vive, so it's this idea of headsets that you wear on top of your, on your face, where it blocks out the real world and puts you inside in a highly immersive virtual world, so it looks like a game in that it's rendered content running, you know, on your graphics card, and you're kind of in that experience, uh, and they try to play up the immersion with, you know, positional audio, so sound comes From different parts of the environment around you, just as it does in the real world, and you usually have, like, hand tracking, so you can interact directly with things in this virtual space. Augmented reality is very similar in that, you know, what we get excited about when we think of AR are these headsets where, you know, like Microsoft HoloLens, where you still see the world around you, but it layers on bits in addition to the atoms that you're seeing. So, you know, HoloLens, they call the things that you see holograms. That's not technically accurate, but it's close enough for what People's conceptualization of the concept or of the idea. So basically it's you go around and you're seeing Pokemon on your phone screen, and when you encounter it, you lift your phone up and see a Pokemon through the screen projected on top of the street that you're walking arou…

AI assessment note: “Virtual reality is... blocks out the real world... Augmented reality... layers on bits”

Q Yeah. Well, what are some of the other ways that you've seen people stumble on this? Because sometimes people discover the hack by accident, or is it deliberate? I mean, how do they really, I mean, do you actually wake up one day and say, okay, I'm going to figure out the hack for this marketplace?

A Oh yes, you do. No, no. When we're talking entrepreneurs, you're looking for the theory. What's your theory on how you're going to bootstrap? How are you going to get the flywheel spinning? How are you going to solve the chicken, the egg problem? Which side are you going to start on? So OpenTable ended up starting completely on the restaurant side because there was zero utility to the consumer until you had a selection of restaurants. And so how do you get a restaurant to adopt it when no consumers are using it? So you, in OpenTable's case is what Chris Dixon, you know, come for the tools, stay for the network. They build a suite of tools that they charge 200 dollars for and laboriously rolled out one restaurant at a Time, uh, throughout, you know, the country that had enough utility that a restaurant was saying, okay, I'll adopt that in the absence of a network. And as they slowly built that base of restaurants, and I mean, slowly, I mean, they, they, a good salesperson would do three or four new restaurants a month. And, you know, there's, you know, there's 2000 restaurants in San Francisco. That is slow. Um, and then, uh, there was a point where there was enough restaurants on the system. It was a subset of the restaurants in the system. But the utility of making an online order was so great that they'd say, I ignore the fact that you don't have any percent of restaurants be…

AI assessment note: “Oh yes, you do. No, no. When we're talking entrepreneurs, you're looking for the theory.”

Q because if I'm in those founder's shoes, how do you know to hang in there? Like, what are the signs, the metrics? Like, okay, uh, I'm gonna just stick around for another 10 years, or I have some sign here. So how do we tell? Let's walk through some case studies, just to have the data, because we use a lot of words. I mean, let's look at the numbers.

A So this is, so the first example we're gonna share is actually Facebook, and the reason we used Facebook was to tease out, ah, the difference between viral growth and network effect, right? So this is the chart usually startups show at the start, like, hey, we're growing really fast, ah, you can see that we reached, you know, the 808 hundred million plus MAUs, ah, in a very short time frame, so the chart looks great, but this is what we, and, and Facebook did this with pretty much zero Customer acquisition costs, right? They didn't spend a dime. But this is what we would call a speed of adoption, which is growth. This does not tell us whether there is a network effect. What tells us whether there is a network effect is actually this, which is even as they kept growing their retention. So this chart shows daily actives divided by monthly actives. And you can see that it kept growing from 45% to 57%. And Facebook was actually one of the first platforms to actually do that. There's, because usually you've never heard where, oh, you increase users. Do you actually increase usage? That was not the norm. And so that's the sign of network effect. And so, and this metric is different depending on what type of company you are. For a marketplace, like for Airbnb, you would actually look at the number of guests that booked rooms and how that trended over time, right? That would be the sig…

AI assessment note: “What tells us whether there is a network effect is actually this, which is even as they kept growing their retention.”

Q original network effect. So there is a certain density there. Well, speaking of the trend side of things, we talked about, The sharing economy. So let's talk about examples like ride sharing and crowdfunding, and those are new technologies. I mean, they haven't been around before, or maybe they've happened, and now smartphones are here to sort of give us new behaviors around them. Do those trends have network effects?

A Yeah, so I think I would combine, you know, I would call ride sharing, um, even food delivery, all of those, I would put them, like the on-demand category in general. I think it's a new phenomenon after mobile, which has made it really easy for consumers. So the way to think of it is, um, so if you look at ride sharing services, right, it's, you know, 10 times better than taking a taxi cab. The product is way better, and consumers use it, but it's a city by city rollout, to a large extent, like OpenTable. However, the difference, I would say, in a ride sharing service, it's a point to point service. In OpenTable, you offered the tools to the restaurants, and then, you know, that sort of, you know, the restaurants used it for the tool, and then you brought in the network, which is the diners, and You know, you match sort of the two sides. In ride sharing, it's really point to point. So I would say the general ride sharing, we think it's more supply side economies of scale. And what that means is the more drivers you have on the platform, you know, you can make sure that you get a good quality driver within five minutes, but that's where it stands, right? So it's like, it's almost like the Amazon first party, if you wanted to think of it, which is like, it's still a good business. We just call it as different, which is You know, supply set economies of scale, which is scale econo…

AI assessment note: “supply set economies of scale, which is scale economy versus a network effect”

Q of the GPU as part of this next platform ship. I mean, I think the biggest surprise people have had is that this is the graphical processor unit, which is something that was developed in the gaming industry for really high resolution graphics processing and is now finding, I guess, unexpected. Is it a surprise to us that it's finding uses in these new platforms like VR, AR, deep learning?

A It's actually, interestingly, it's a new application of an old idea. Back when I was getting started, 30 years ago, working in physics labs, if you wanted to run, um, just a normal program, you just, You just buy a normal computer and run the program, but if you wanted to do, uh, run a program, many physics simulations had this, had this, uh, property where you, you would want to run a very large number of calculations in parallel, right? And so you could, you could basically divide up a problem of simulating anything from a black hole or to different kinds of biological simulations. Um, you could basically write these algorithms in a way that they could run, you could basically parcel the problem into many different pieces, uh, and then run them all in parallel. And there was actually in the old days, uh, there was actually a whole industry of what were called Vector processors, which were literally these kind of sidecar computers that you would buy, and you would hook up to your main computer, and they would let you run these parallel problems much faster. And so literally, 30 years later, the GPU is a, it's basically a vector processor. It's basically a sidecar processor that sits along the CPU and runs these parallel problems much faster. And it, graphics are a natural application of that, but as it turns out, graphics aren't the only application.

AI assessment note: “It's actually, interestingly, it's a new application of an old idea.”

Q Right. Well, let's spend a few minutes talking about the economic effects of the sharing economy, because, you know, for better or worse, you, you are an economist by training. You've done a lot of research on the sharing economy in traditional economic terms. What, how do you think about the impact of what it brings?

A So, um, I have, I categorize the traditional economic impacts of the sharing economy or crowd-based capitalism into four key impacts. Um, the first is how we are Increasing capital of assets of money through more efficient use of it. So you might think of this as leading in the long run to an increase in productivity because you're using stuff more efficiently, and so that's one impact. A second impact has to do with an increase in consumption that comes from greater variety. So let's contrast, for example, the level of variety that you have on Airbnb With the level of variety you have in the hotel industry. There's just so much more choice of so many different configurations and kinds of accommodation on a platform like Airbnb. Economists disagree about most things, but there are two things that economists agree about. I mean, one is that when you increase efficiency and you increase productivity, this leads to long run economic growth. And the second thing they agree about is that when you increase variety, You increase the amount that people consume, and increased consumption also leads to economic growth. So those two factors, increased capital impact and increased variety, have a decidedly positive prognosis for how crowd-based capital is going to alter the economy.

AI assessment note: “I categorize the traditional economic impacts of the sharing economy or crowd-based capitalism into four”

Q because it's a topic that comes up a lot, and whether you agree with the academic nature of the discussion or not, it does impact Where people put funding, how people allocate resources, how they think about policy and everything related to the topic of the future of work, just at a very basic level, like what's the main problem with GDP as a measure for software in the world?

A I think we can break down the trouble with GDP into at least three buckets. Um, the first is that there's a lot of value that's created by technological progress. Like, you know, the value you get from Google searches, the value you get from sort of connecting with other people on Facebook. There's sort of the surplus that, you know, consumers enjoy that aren't captured. This, this isn't captured by GDP. And so if that's relatively small, it doesn't matter. But, Technological progress, especially progress with digital technologies, seems to make this consumer surplus piece pretty big, and GDP is not measuring that. It's just measuring the flows. It's measuring the prices that we pay and the revenues that firms get. So that's one problem. A second problem has to do with distribution. GDP is a total measure, meaning it's adding up all of the dollars, and it's not giving us a sense for how equally or unequally they're distributed. So, for example, um, about 80% of India now has access to a mobile phone. Um, a couple of decades ago that was just 10%, and so we've not just seen an increase in consumer surplus here, but we've seen, like, a tremendous reduction in inequality of access that just doesn't get picked up by GDP.

AI assessment note: “I think we can break down the trouble with GDP into at least three buckets.”

Q You know, the book is called The Age of Ambition. Um, certainly China's a very dynamic place, shall we say. Are we Going into a different phase, and how would you describe that?

A Yeah, that's a good important point, because something is changing at the moment in China. Something is, and I don't say that with great pleasure. I did a long profile in the New Yorker last year of the president of China, Xi Jinping, and he came into office at the end of and he's turned out to be a much more important figure, because in simplest terms, he's taken what was this consensus model of Governance, which was you had this group of guys who were almost studiously bland and nonspecific. They wanted them to be that way. They were looking for the opposite of Chairman Mao. And instead, now we have, once again, a very dominant single personality. And as that can be a good thing or that can be not a good thing. And in some ways, it has set back the process of bringing China into the world, the process that China itself initiated. So in practical terms, if you're sitting in Beijing or Shanghai right now, you're not able to get on a half a dozen information sources, news sources, data sources that you could have three years ago. And from my perspective, how many countries in the world today do you go to where the internet connectivity is worse than it was three years ago? So it's that instinct that worries me, because what that suggests is politics over pragmatism, and pragmatism has helped China so much over the last 40 years that I want that to be in the foreground as much as…

AI assessment note: “something is changing at the moment in China”

Q Well, tell us more about those other entrepreneurs you met. I mean, what was like, what was unique about them? What were the things they were working on?

A The reality was that in some cases, these are still somewhat derivative of successful American products. That's a dynamic we've seen for a long time. But in other cases, They'd begun to tweak things in some important ways. I highlight one example in my book. There's a woman that I met when I moved to China whose name was Gong Haiyan. She'd been very successful. She'd gotten a PhD. She had grown up in a little village in the middle of nowhere at the foot of a mountain and had made her way to the city. So she got into her late twenties and her parents said, okay, now it's time for you to move home and get married. And this is the way that we have always done it. We will introduce you to somebody in the village. You'll be very happy. And, you know, traditionally in Chinese marriages, there really isn't much room for individual choice on the part of the participants. And in this case, she said, that's not going to work for me because the people that you know are nothing like the people that I know. And so, and she started a company, essentially the Chinese equivalent of match.com. And it was very early. In fact, it was so early that people didn't even have enough, uh, connectivity. They were mailing in their photos.

AI assessment note: “she started a company, essentially the Chinese equivalent of match.com”

Q Well, tell us more about those other entrepreneurs you met. I mean, what was like, what was unique about them? What were the things they were working on?

A The reality was that in some cases, these are still somewhat derivative of successful American products. That's a dynamic we've seen for a long time. But in other cases, They'd begun to tweak things in some important ways. I highlight one example in my book. There's a woman that I met when I moved to China whose name was Gong Haiyan. She'd been very successful. She'd gotten a PhD. She had grown up in a little village in the middle of nowhere at the foot of a mountain and had made her way to the city. So she got into her late twenties and her parents said, okay, now it's time for you to move home and get married. And this is the way that we have always done it. We will introduce you to somebody in the village. You'll be very happy. And, you know, traditionally in Chinese marriages, there really isn't much room for individual choice on the part of the participants. And in this case, she said, that's not going to work for me because the people that you know are nothing like the people that I know. And so, and she started a company, essentially the Chinese equivalent of match.com. And it was very early. In fact, it was so early that people didn't even have enough, uh, connectivity. They were mailing in their photos.

AI assessment note: “There's a woman that I met when I moved to China whose name was Gong Haiyan.”

Q We did. So for granted that it's shocking that it's like eye opening. Actually, that brings me to an interesting question. Are you seeing government, you know, their government clients do startup like things now because of this?

A We are, we are. And it's really cool. So the, a couple of examples I'll say is like, Which is a new group that started out of GSA. 18 F is based on the street it's on, 18 F. There's sort of this little internal SWAT team of techies that are going in, and they're helping government with their projects, number one projects that weren't doing as well, or new projects they're trying to get off the ground. And they're trying to use innovative technologies, and AWS, of course, is core to what they go to, because they're like, okay, let's get your infrastructure up, let's get going, let's move fast. And then also the digital services office. I think they're a really great example. Those kind of things are happening, and I see that around the world. Liam Maxwell, who's the CTO of the UK government, sort of started the same kind of trend in the UK where he said, we are spending way too much money on tech and not getting the benefits. They've created their own digital service, and they're starting to put also policies and mandates in place that says, wait a minute, don't go out and buy Uh, like a bunch of servers when you can go to cloud. Show me why you can't use cloud. New procurements have to go through this group to say, why not cloud?

AI assessment note: “We are, we are. And it's really cool. So the, a couple of examples”

Q He's setting up his game too. Okay, so one last question then to wrap up. Speaking of media and your role now as editor-in-chief of Vox, and we started talking about this in the beginning, coming full circle, like what do you think is happening next, not just for you personally, but in this space?

A So I'm pretty optimistic about the business, but it is changing. One thing I did not expect was actually how fast some parts of it would change. I think the thing that's going to be very, it might be great, but it's going to be, I think, difficult. Is how much we are going off platform. When we started Vox, a lot of the theories about Vox were about, you know, Vox.com, frankly. When we talked about card stacks and created card stacks, card stacks are something, and those are a product we have to attach contextual information to our news story. So if you're reading something about, say, um, the, the Trans-Pacific Partnership trade deal, we can cover the new news in that, and then also let you see this Set of cards you can swipe through that give you all the background information. So if you've not been following that, you can catch up very quickly, but card stacks cannot be imported to our Snapchat discover product. They can't be put on at the moment, Facebook instant articles. Uh, they're not relevant for a lot of the places people are now discovering our content. And as that happens, as we have less control over the platforms on which we publish, I actually do worry about the way it will choke off innovation. Because all of a sudden, I mean, let's say we create a great data interactive, which is something we do a lot at Vox, and that interactive, it can't be in a Facebook Inst…

AI assessment note: “how much we are going off platform”

Q But I will say that when I first saw what was happening, everyone was kind of mocking as if it was a joke, and I was like, why is anyone not taking this more seriously?

A What I think you're seeing with Trump, so one, the faction of the Republican Party Trump represents is by no means new. It is the faction that was voting for Pat Buchanan, It's very similar to the pro coalition, which drew not exclusively, but quite heavily from the Republican party. And I think many of the Democrats who were attracted to pro are now Republicans or part of that shift over in the South. Uh, so this is not new. What is interesting about what Trump is doing is that he is able to completely divorce himself from the party apparatus while running within the party. Usually the party would, the party has traditionally suppressed this part of itself, right? The fact that you had a lot of Downscale white voters in the Republican Party who like Medicare, who like Social Security, who are happy to tax rich people, who are skeptical of trade, but who are also socially conservative, heavily nationalistic, and very, very upset about immigration. Trump didn't discover that. We knew that. The Republican Party suppressed those kinds of candidates by its control of money, its control of information, its control of who gets taken seriously, of who gets good jobs, of who gets good positions. Trump For unusual reasons, his media celebrity, his money, et cetera, has been able to subvert all of that. He has both been able to be, uh, a candidate on the stage in Republican debates, and …

AI assessment note: “the faction of the Republican Party Trump represents is by no means new”