Pod Insights Y Combinator Podcast #153 - Cory Doctorow and Joe Betts-LaCroix About this episode Cory Doctorow https://twitter.com/doctorow is a blogger, journalist, and science fiction author. He’s also the co-editor of Boing Boing https://boingboing.net/.
Joe Betts-Lacroix https://twitter.com/bettslacroix is the CTO of VIUM https://www.vium.com/ and an Expert at YC.
Y Combinator invests a small amount of money ($150k) in a large number of startups (recently 200), twice a year.
Learn more about YC and apply for funding here: https://www.ycombinator.com/apply/ https://www.ycombinator.com/apply/ 
***
Topics
00:00 - Intro
00:32 - Adversarial interoperability
7:55 - So what should founders do?
13:55 - Copyright
19:55 - Remixing
27:00 - AT&T
29:50 - Firms and state-like duties
31:25 - Medium-term suggestions for founders
39:25 - Conspiracy theories
51:50 - Science fiction and predictions
56:55 - Peak indifference model
Transcript 13,452 words · 3 speakers
0:00 Speaker 00 Hey, how's it going? This is Craig Cannon and you're listening to Y Combinator's podcast. Today's episode is with Cory Doctorow and Joe Betts-Lacroix. Corey is a blogger, journalist, and science fiction author. He's also the co-editor of Boing Boing. Joe is the CTO of Viium and an expert at YC. You can find Corey on Twitter at @Doctorow and Joe at Betts-Lacroix. All right, here we go. All right, guys, welcome to the podcast.
0:27 Speaker 01 Excellent, thank you.
0:28 Speaker 00 So today we have Cory Doctorow and Joe Betts-Lacroix. Joe, could you start it off?
0:33 Speaker 01 Sure. So Corey, when I saw your talk at Burning Man this last time, and I heard you mention adversarial interoperability, my nerd buzzer went off and I thought, wow, this is going to be a really cool concept for the founders at Y Combinator to think about.
0:53 Speaker 00 Yeah, so I think most people are familiar with the idea of interoperability per se. You know, one thing plugs into another thing. And we've all encountered a kind of indifferent interoperability where, you know, people who make your fridge don't care what kind of magnets you put on the door. We've encountered like cooperative interoperability where the people who make your car actually go and adhere to a standard for the 12-volt lighter receptacle to make sure that when you go to the gas station and pay 50 cents for that, you know, USB charger thing, that it always fits and works every time. And those are great and they're important and they're kind of how we build systems. But there's another kind of interoperability that's really important to the history of tech, which is adversarial interoperability.
1:35 Speaker 00 And that's when I plug something into something you made without your permission, even if you don't like it. And maybe especially if you would prefer that I didn't in order to make your customer my customer. And you can understand why companies that are very established don't like adversarial interoperability. But people who are trying to start a company, that's kind of where you'd want to start. And it used to be completely standard, right? If you remember, there were a couple of moments in, say, Mac OS's history where adversarial interoperability was totally pivotal.
2:09 Speaker 00 At one point, Microsoft had not just dominance in the operating system market and the applications market, but it was establishing dominance in the local networking market, intranets. They had a product called SMB that was a proprietary networking protocol. And although they made clients for SMB for other platforms, they were deliberately very poor so that those systems would always be second-class citizens on the office network. It was really bad for Mac systems. It was really bad for GNU/Linux and other Unixes. And it really sold the hell out of a lot of Microsoft systems. And so, very slowly but surely, everything that wasn't a Microsoft OS was being squeezed out of the office LAN.
2:52 Speaker 00 And this Australian grad student used a protocol analyzer to capture enough SMB traffic that he was able to replicate SMB. He released it as free open-source software. He called it Samba. It attracted a developer community and was, became kind of a gold standard, where now it's bundled in with, with everything you use. It's just part of, you know, the standard Linux distro and, and it's, it's in Mac OS and, and even in mobile platforms. And they did that even though Microsoft didn't want them to. Apple then internally produced another set of adversarial interoperable tools that were really important in the, in the mid 2000s, when Microsoft's dominance of the office market, of the office suite
3:37 Speaker 00 market, was really threatening Mac's ability to collaborate within those environments. The, their, Microsoft's versions for the Mac always lagged the Microsoft versions, and so the documents weren't really compatible. Excuse me. So, Apple reverse engineered Microsoft's file formats, and they cloned them, and they made a new suite called the iWork Suite. And they did this again without Microsoft's cooperation and in defiance of Microsoft's marketing strategy. And in so doing, they were able to make Macs first-class citizens on the, on the LAN and in the office environment. So in both cases, they were able to do this because there were very few legal
4:22 Speaker 00 recourses available to dominant firms that had taken over their field. Software patents were rare and thin. The Computer Fraud and Abuse Act, which is this 1986 anti-hacking law that Ronald Reagan signed into law in a panic after he watched Matthew Broderick in War Games, I'm not making that up, had not yet been extended to cover terms of service violations, as it would be eventually, thanks to legalization by Blizzard and Facebook, where they were able to effectively say that violating terms of service was hacking and potentially a felony. We didn't yet have theories of tortious interference for inducing users to break terms of service,
5:07 Speaker 00 to break contracts. The Digital Millennium Copyright Act and its prohibition on reverse engineering hadn't become widely applied yet. So today, if you were to try and do this, like say you were to try and make Android devices first-class citizens in an iOS environment, and in order to do so, you wanted to reverse engineer iOS and allow yourself to sideload apps into iOS and also to support the iOS API under Android, you would find yourself running up against software patents. Digital Millennium Copyright Act claims for circumvention. You'd find yourself running up against terms of service violations and probably tortious interference claims.
5:51 Speaker 00 This thicket has grown up around adversarial interoperability, you know, like every pirate wants to be an admiral, everyone who goes up the ladder wants to pull the ladder up behind them, and it's not unusual that they wanted to do that. I mean, IBM wanted to do that when they were dominating the mainframe market. The difference is that we didn't let them, we didn't equip them with these legal tools. And in the years since, we have created these legal tools that allow firms not just to dominate, but to maintain their dominance, not by making better products, but by establishing a kill zone around their products where it becomes so legally fraught to make compatible products that nobody even tries.
6:27 Speaker 01 Well, we have YC, this group of the little people who are...
6:34 Speaker 01 Striking out and wanting to create these these huge companies rather than letting the big companies do everything. And and one of our flagship publications or forums is Hacker News. There's this ethos of we are hackers and we get we can get there by hacking. So I I would like to think that we can still hack and not get our heads hacked off is one of the so a friend of mine, Aaron Patzer, created something called Mint and ended up taking a bunch of eyeballs from big banks and other financial companies by hacking, essentially going doing adversarial interoperability with their websites by scraping, which is sort of marginally legal.
7:18 Speaker 02 Was it through like a virtual browser that he would just like take your password information and log you in?
7:24 Speaker 01 Absolutely.
7:24 Speaker 02 Okay.
7:25 Speaker 01 He gets the passwords from the users voluntarily and then and then everything just magically works in the background.
7:31 Speaker 02 Because there was no API at the time.
7:33 Speaker 01 Correct.
7:33 Speaker 02 Okay.
7:34 Speaker 01 Yeah. And that allowed him to build a big startup that got sold for, you know, gazillions of dollars by one of those giants, financial corporations, which is now in one of these monopolies that is trying to keep other people from doing similar things. He also sold a bit early, like it wasn't a billion dollar company yet. And do you think that if Y C founders are using some of these tools that are kind of illegalish or strictly illegal right now, that they're going to be forced into selling earlier because otherwise they won't be able to defend themselves?
8:07 Speaker 00 That's a great question. So let me answer it with a kind of theory of change because I don't think that the thicket is sustainable. I think that nobody really likes having a web that's just five giant websites filled with screenshots from the other four. And and, you know, you see it in the rhetoric of the big platforms when they describe their own failings, they say, Well, what do you expect from us? We're giant, right? So Zuckerberg's so-called free speech talk this week, you know, one of the things that he said was, Well, How can we possibly make speech adjudications in 100 languages for 2.3 billion people
8:51 Speaker 00 in hundreds of countries? You know, once you build a thing on the scale of Facebook, you're going to end up with some, you know, unfortunate ethnic cleansings, but you know, you can't make an omelet with a without breaking eggs. And you know, there is an obvious answer to this, which is, well, if you can't do it at that scale, maybe you need a smaller scale. And given that we have gone from these big platforms and their inability to provide for use cases that are at their peripheries, like women who are harassed by terrible men
9:31 Speaker 00 or whatever, is intimidating and promotes inequality and makes people's lives miserable to these platforms and their inability to address these corner cases result in in acts of genocide. You might imagine that we are headed for some kind of crisis in these firms. So Lawrence Lessig, the cyber lawyer, who was the special master in the Microsoft antitrust case and co-founded Creative Commons and so on, he says that the world is regulated by four forces. One, you're definitely familiar with, it's code, right? Things that are technologically possible happen sometimes, things that are technologically impossible never happen, but obviously, right?
10:16 Speaker 00 There's markets, right? Things that are profitable happen more often. And there's law, things that are lawful are more likely to occur than things that are illegal. And then there's norms, things that are socially acceptable are easier to do than things that are socially beyond the pale. You know, the the question of whether or not someone like Jeffrey Epstein is welcome in your community has become one of enormous moment, right? And, you know, you can see that even though Jeffrey Epstein had billions at his disposal, the fact that he was normatively beyond the pale really irked him, right? This this Sociopathic monster of a man made it his life's work not just to pursue
10:59 Speaker 00 the terrible acts that he pursued with vulnerable young women, but also to do so while not being socially unacceptable. His whole goal in life was to be in company of wonderful people. And you know, you sometimes meet founders who cash out big businesses and decide they don't want to pay taxes and they give up their citizenship and they move to the Grand Caymans or whatever. And they discover after months or years that hanging out with like substance addicted sociopaths on yachts for you know, every hour that God sends is a bit tedious, right? And that eventually you wanna, you wanna put down your daiquiri, get on a plane, go back to a place where the action is happening. So norms are also very important.
11:44 Speaker 00 And so what I would say is that if you have a business that is doing something that is legally fraught, but that business is doing something that people love, that you are creating a constituency for a normative shift in how we think about this stuff. And that you are also creating surplus capital that can be used to fund legal adventures to make this stuff more legal. And that in the absence of such a constituency, that change is going to be longer in coming and also much more chaotic when it arrives, right? Change that occurs without any kind of market constituency,
12:26 Speaker 00 where it's just, it's purely normative generally involves like guillotines, right?
12:33 Speaker 00 Firms, when their interests are aligned with a new normative view in the world can ride a wave, right? They can, they can be part of a legislative change that is also incredibly profitable for them and that they support for self-serving reasons, but that has broad popular support because it is doing away with something that was manifestly unfair. Now, they may find themselves on the wrong side of it later, right? And that's, that's often the cycle, the pirates become the admiral. You know, the, the, the people who made phone, uh, radio and who played phonograms on the air and who faced down the collecting societies who said, no music should ever
13:16 Speaker 00 be played on the radio. That's piracy. They were on the right side of history, but when they sued cable operators 30 years later for putting broadcasts on a wire, they were on the wrong side of history. But you can be on the right side of history if you were involved in this kind of firm. And, you know, the big, big upsides are not creating another Facebook because there won't be another Facebook. The big, big upside is creating the thing that happens once Facebook has been torn down. Uh, and, um, you know, there is no future in mimicking the past.
13:50 Speaker 02 So I'd be interested in talking about specific examples like Airbnb, like writing a change. But I'm also interested in talking about this as it pertains to your career. So as a creator, copyright becomes really important for your livelihood. So how have your thoughts changed over the course of your life as a creator who sells their content?
14:10 Speaker 00 Yeah. So I should mention, I'm a science fiction novelist. I'm a reasonably successful one, 20 something books, New York Times bestseller, six-figure advances, all of that stuff, right? So I'm definitely in the 1% of literary creators. And, um, I like the idea of copyright in as much as I like the idea of being in an industry that has a framework that describes how stuff works, right? What's, what's saleable, under what terms, what is unconscionable? What are our enforcement mechanisms? What are our remedies? You know, all that stuff? Having rules of the road is really important in an industry. The specific rules that we've
14:54 Speaker 00 created, I'm very skeptical of. Um, not least because we keep trying to apply them to non-industrial contexts. So every, every industrial rule has to figure out who's in the industry and who isn't, right? If you create finance rules, again, something that I'm broadly in favor of, and you say, things that are bank-like need to account for themselves with the SEC or with the FTC or with some other regulator to make sure that bad things aren't happening. And we say, you are doing something bank-like when at least a million dollars changes hands. And then you need to, then you need to fill in some paperwork. That might be a reasonable test of whether you are bank like or not.
15:35 Speaker 00 But if then we have hyperinflation and a burger costs a million dollars, buying lunch for a friend does not make you a bank.
15:44 Speaker 02 Well, I mean, flip it around and be like, if you have a million dollars in net worth, something flips in your brain and now you are an accredited investor and can make investments.
15:53 Speaker 00 Sure, sure. Yes, yes. But imagine that suddenly a million dollars is worth a fraction of what it was. Imagine that something that used to only happen every now and again is now happening a million times before breakfast.
16:05 Speaker 02 And so you're back now.
16:06 Speaker 00 And so not only are you a bank, but when you give your kid allowance, he's a bank too. Right? So now we're, now we have to say, okay, well, this regulation, whether or not the regulation was good, the test that we're using to figure out whether you're in the regulation or outside of it needs to be revised. With copyright, the test that we use to find out whether or not you were in the entertainment industry is whether you were making or handling a copy of a work that involved creative labor. And that made sense when every book had a printing press in its history and it made sense when every record came from a, from a record press and every movie went through a film lab. And there would be some edge cases that it would capture that weren't actually part of the entertainment industry. And sometimes those would go to court or they would just have a gentleman's agreement to ignore it or whatever.
16:50 Speaker 00 But the main body, the first couple, three sigmas, was all people in the entertainment industry. And if you can afford a printing press, you can afford a lawyer to make sure that you're using your printing press well. But the internet only works by making copies. You know, we have this like polite fiction that there's this way of getting data without copying it called streaming. It's like a consensus hallucination, like the internet is made out of like a cunning network of cardboard tubes with mirrors at the end that somehow get the picture from here to there.
17:18 Speaker 02 Not only that, it's with like 10 million concurrents.
17:20 Speaker 00 Right. Streaming is just downloading without a save button. Yeah. Right. Streaming is I'm sending, I'm letting you download this thing. And I think the client you're using doesn't have a save.
17:28 Speaker 02 And you don't know how to.
17:29 Speaker 00 Yeah. Yeah. And you don't know how to hack one in. Right? So. Everything we do on the internet involves copying. And you copy a million times before breakfast. Yeah. That doesn't make you part of the entertainment industry. And the idea that we're going to take copyright law, which is foundationally about regulating entities in the supply chain of the entertainment industry, and use it as our first line of regulatory coherence for people's romantic lives and people's family lives and people's employment lives and people's education and their civic and political discourse. Right? That when someone uses a song in a political ad that our first question is not a question about elections and election law, but about entertainment law.
18:15 Speaker 00 We've clearly put the horse before the cart. Not only that, when we ask our fans to use copyright law to navigate their interactions with each other and with us, we create the circumstance under which either our fans are always going to be on the wrong side of the law or the law will have to be so simplified and streamlined so that fans whose job is not to be in the entertainment industry can understand it, that it will cease to be functional as like the technical framework for our work. So, you know, I live in Burbank, California. I can walk from my house to Universal, Warner and Disney. And there's a very good Harry Potter theme park at Universal Studios that was made by licensing a bunch of IP from Warner.
18:59 Speaker 00 I also live on a block with a bunch of kids who go to middle school with my kid. And some of them write Harry Potter fanfiction. When the lawyer at Universal called the lawyer at Warner and said, "What are the terms on which we are going to conduct this extremely technical, multimillion dollar, high stakes deal?" They did not have to use a framework that was so comprehensible that the 11-year-old on my block writing Harry Potter fanfic could make sure that she was on the right side of it. And if they had to, it would have been a nonsense. And so, there are a bunch of rules that I would like for copyright. But the most important copyright reform, the thing that gets lost whenever we talk about copyright reform, is that we should just stop asking people who aren't in the entertainment industry to give a shit about copyright. If you have to be a copyright lawyer to read a book,
19:43 Speaker 00 then either no one will read books or everyone who reads books will be wrong. And so, how do you think about remixing and republishing? Yeah, I think that when it's cultural, it's one thing, and when it's industrial, it's something else. Can you define that? Sure. So what I was gonna say is that like some industrial use is very obviously industrial, right? Like if I write a book and publish it. Yeah. It's obviously industrial. So perhaps another way is like for profit? Not all industry is for profit. So if I write a book and publish it for the Church of Scientology and don't charge any money for it, it still might be industrial. Okay. Right? On the other end of the scale, if you're writing fanfic and sharing it on your classroom's blog
20:24 Speaker 00 or, you know, Google Classroom site, that's clearly not industrial. And then there's some corner cases. And crisply defining those is it's the work of years and you would still miss stuff. So that's why we have judges who apply principles when the corner cases matter. Most corner cases don't matter, right? Most corner cases like, oh, well, that guy's doing something. Maybe it's industrial, but it's like, if there's a loss to me, it's measured in fractional sense. And if there's a gain to him, it's measured in a fractional sense. People have blogs that have ads from ad brokers who have a few hundred readers a day. Those people are industrial-ish. No one is ever going to adjudicate whether or not they're industrial because the stakes are so low that nobody cares. They don't care if they're industrial.
21:09 Speaker 00 The people whose work they're using don't care if they're industrial. It's just, it's just a wash, right? So I like the idea of there being rules for industrial use, but I think that those rules should err on the side of free expression. So a good example of where that's worked is there was an African-American writer who wrote an unauthorized sequel to Gone with the Wind called The Wind Done Gone. And it was told from the perspective of the slaves, right? And it was about revealing the hidden white supremacist subtext of Margaret Mitchell's famous novel, beloved novel. Now, unsurprisingly, the Mitchell estate was not happy that Margaret Mitchell's book
21:50 Speaker 00 was being brought into disrepute by this work that was derived from it. But I think we can understand that artistically, if you wanted to puncture The, you know, white supremacist subtext of the wind gone, gone, the best way to do, of gone with the wind, rather, the best way to do it is to make close reference to that work, especially if you're going to do it in fiction, is to have the same character, same characters, but a different point of view to understand what's missing, what's omitted from the frame. The Supreme Court agreed. The Supreme Court said this is fair use and we're good, right? Even though it was commercial, it came out from a publisher, they charged money for it, it was sold in bookstores, it satisfied all the criteria for being commercial.
22:33 Speaker 00 So those rules, a robust practice of fair use, I think is, is great. I think we could combine the European fair dealing approach where they have enumerated exceptions with the American fair use approach where they have principles that are applied to get a kind of best of both worlds where we might say, okay, here are a bunch of things that are always fair use, right? These, these, these kinds of things like if you take X seconds of a song, it's always fair. If you take X lines from a poem, it's always fair. But that's the floor. The ceiling is if you are doing something that has some free expression nexus, that is critical and transformative and so on, these things that are harder, more nebulous and might require a judge, then we can future proof fair use because some fair uses are
23:17 Speaker 00 intensely technological, right? Like, we never had a question about whether it was fair use for like Dizzy Gillespie to blow eight bars of Night in Tunisia in the middle of a solo. But we had that question come up when it was Terminator X taking a couple of bars out of a classic jazz or R and B composition and mixing it into a Public Enemy album. And those fair use questions are going to evolve as the technology evolves and the way that we address them should be in the hands of both judges and lawmakers. So we should have judges adjudicate them. If the answers become incoherent because we have different circuits making decisions or if they fall afoul of public sentiment, then lawmakers might be enticed to change the law so the judges interpret the law differently.
24:02 Speaker 00 So for example, with deep fakes, we're seeing a lot of opportunity for legitimate free expression, things that are artistically or politically important that, that are.
24:14 Speaker 00 You know, very exciting from an artistic perspective. And yet, we also see a lot of risky things that are happening with it, including things that might be harmful to free expression, that might be harmful to artists and their livelihood and so on. And if we had a set of principles that could be interpreted, you know, like we do now with fair use, where we have the four factors, then we can start to build up a set of norms around when a deep fake is good and when a deep fake is bad. And we'll have some liminal stuff in the middle that judges will continue to adjudicate. And maybe eventually we'll get a law that will clarify matters. Although, the idea would be that the law would, would again, form the floor and not the ceiling, that we'd maintain the principles that could be used as the basis for the next set of law
24:57 Speaker 00 as deep fakes evolve and other kinds of technologically assisted creativity evolve.
25:03 Speaker 02 But you have to allow some freedom there, right? Because if you write it all down, the future happens and then it's not relevant.
25:09 Speaker 00 I mean, you know, I alluded before to what happened when the phonograms.
25:12 Speaker 02 Yeah, exactly.
25:13 Speaker 00 Broadcasters, it actually goes before the phonograms. The phonograms exist because they successfully fought off an effort by the people who made sheet music to prohibit recorded music. John Philip Sousa, you know, the composer, who was in the sheet music business, went to Congress and said, "If the infernal talking machine is allowed to continue, we will lose our voices as we lost our tails when we came down out of the trees." Right? So, you know, many points to Sousa for being an early advocate of the sadly still controversial theory of natural selection. But, but, you know, he clearly missed the boat here. And so, yeah, so the sheet music people sue the record people, who sue the broadcasters, who sue the cable operators, who sue the VCR, who sue Napster.
25:52 Speaker 02 Right.
25:53 Speaker 00 Right? At each turn, you have someone who, who claims, "When I did it, it was legitimate artistic expression." When you do it, it's piracy, which brings us back to adversarial interoperability, where, you know, you have these firms that routinely engaged in adversarial interoperability, who describe what they did as part of a kind of heroic effort to wrest control from dinosaurs who didn't understand what the future needed, but who now describes themselves as guardians of propriety, who, who mustn't ever be disrupted. You know, Zuckerberg and that recording that leaked to the verge of his all-hands meeting describing, you know, Facebook's future and rubbishing Elizabeth Warren and her breakup plans. She, he says of Warren and the breakup plan, "If you're worried about political disinformation
26:38 Speaker 00 and other harms that arise from, from Facebook..." You must allow us to be as big as possible for who can generate the surplus capital needed to hire the army of programmers necessary to deal with the dysfunctions of our scale, but for us, right? And you know, this is an argument that's been made by monopolists since the year dot. AT&T, when it was formalized as the Bell system and given a license to operate as a monopoly, with that came all kinds of special privileges. Like, they had the right to decide what you could plug into the phone network, all the way down to mechanical couplings. They sued a company called Hushaphone that made a cup that fit over the handset of your Bell rental phone, because you could only get a phone from Bell, so that when you had
27:21 Speaker 00 a conversation into it, it would be muffled and other people couldn't read your lips. And they said mechanically connecting something to the handset of a Bell phone endangers the long-term security of the Bell system. And after all, the Bell system is now part of public safety, it's part of security, it's how we administer wiretaps, it's how we connect the police with their constituents, you know, the people they're supposed to be protecting and so on. How are we to provide all this service unless you give us the monopoly windfall rents that we collect from monopolizing who can plug what into the system? And you know, it held back modems, it held back the internet, it held back, you know, early Usenet. It was only the dismantling of the Bell system that enabled the next generation to emerge.
28:05 Speaker 00 And you know, Facebook is positioning itself to become a state monopoly like the Bell system was, as is Google, as is Apple, right? All of these companies, they want to be, when they say, "Please regulate us," right? When they say, "Please make rules to tell us when we're on the right side of the law." What they're really saying is, "Please fashion a kind of throne for us, and when we sit upon it, we will allow ourselves to be draped with golden chains, right? We will become kind of the divine rite of kings." And you aristocrats in the regulatory world, who are, after all, when you've only got five companies in an industry, everyone qualified to regulate them is formally an executive at one or more of them, right?
28:49 Speaker 00 These aristocrats who are drawn from our executive suites will tell us how best to exercise our noblesse obligée to the little people so that we can go on running it. And you see this kind of state monopoly framework emerging again and again when people talk about regulating big tech. What very few people are yet talking about is just breaking them up.
29:09 Speaker 02 Right. Well, some, but I think it's often, I mean, you saw it a couple of years ago with the hearing. Many lawmakers have no idea of the scale, how they work. I mean, I think WhatsApp's a great example.
29:21 Speaker 01 Or, or Amazon is now involved in video surveillance on the, at the behest, as you were saying, of police forces all over the country. And that's, that's a situation where they have this creeping, growing monopoly power, but it's going to be likely enshrined or ennobled, as you say, by these bastions of legitimacy in the country, which is people who fight the crime and keep us safe.
29:46 Speaker 00 They're recruiting constituents, right, to their, to their cause. If they are part, if, once you ascribe state-like duties to large firms, you make it impossible to make them smaller because then they can't perform the state-like duties.
30:00 Speaker 02 Right.
30:01 Speaker 00 We can, I think that we are at this point where we can decide that we're going to fix tech, or that we're going to fix the tech companies. But we can't do both, right? If we fix the tech companies, if we say, in order to operate a social network, you have to make sure that you are policing hate speech, copyright infringement, terrorism, and every other kind of bad act, then what you're really saying is that you can only have a social network if you're so big that you can afford the moderators to do all that work. The reality is, they're not going to do the work very well. But also, the reality is that if you give them 10 years without having to look over their shoulder at a competitor, that they're going to be even more intransigent, and it's going to be even harder to uproot
30:44 Speaker 00 them. You know, how, if you look at the AT&T antitrust case that finally broke them up in 1982, it was a 30-year effort to break up AT&T. And it arrived at the moment in which Ronald Reagan was dismantling antitrust enforcement, which is why immediately after they were broken up, they started to re-merge. Because that, they were like the last hurrah of antitrust enforcement in America.
31:08 Speaker 02 So, thinking about this... Kind of in the medium term for founders, like strategically, how would you advise someone?
31:17 Speaker 00 So, nobody knows when these anti, these anti adversarial interoperability laws can be invoked because the jurisprudence on them is very thin. What you have is a whole ton of extremely risk averse counsel and risk averse investors who, who don't, you know, they claim that they're swaggering lords of creation and they're, as we were saying before we turned the mics on, you know, they're perfectly willing to say, well, I've got this idea. I'm going to disrupt everything by turning every apartment in the city into an unlicensed hotel room. But as soon as you say, well, I've got another idea. Facebook doesn't have users, it has hostages. People hate it.
32:00 Speaker 00 They lost 15 million Americans aged 13 to 34 last year, largest ever exodus of Americans from Facebook. All those people landed on Instagram, right? I know what I'm going to do. I'm going to offer a tool that just like Mint, logs into Facebook, pretends to be you, grabs all of the messages waiting for you on Facebook, cleans out all the waiting queue in your favorite groups, and then puts them into a new context for you where Facebook can't watch you and allows you to reply to them. And, you know, Facebook, if you believe that they have 2.3 billion users, which may or may not be true, if they have 2.3 billion users, they have to support 2,301 in a million use cases every single day, which means that their bot detection and their ability to distinguish a bot from someone who's just doing a one in a million thing, it's really, really hard.
32:45 Speaker 00 You know, I have a friend who, Raph Koster, who designed a game called EverQuest. And in EverQuest, there was this thing where they, they set up an equilibrium for crafting, where if you crafted a shirt, a sheep would disappear. And then if you wore the shirt until it wore, it wore out, a sheep would appear in the game. And so there was an equilibrium between sheep and shirts.
33:10 Speaker 02 Okay.
33:10 Speaker 00 What they didn't imagine was that there would be someone who would find making shirts meditative and who would just fill their house with, with shirts until there were no sheep left in the world. Because while not wearing the shirt, it didn't wear out? Yeah. Okay. So that person is a one in a million very special player, right? But the world is full of one in a million people. Yeah. Right? A world of seven billion people has, uh, check my math here. Seven thousand. Seven thousand, one in a million users. Yeah. Yeah. So, uh, every day. Yeah. And so what if you said, okay, Facebook has double digit year on year growth. They're worth billions of dollars. I just want to capture 2% of that market, right?
33:50 Speaker 00 That's still bigger than any payout you get from making a company be aqua hired by Facebook. And you get to be your own boss, right? You wouldn't have to work for the dysfunctional hornet's nest that is Facebook. And, uh, and you'd also operate in a much lower stakes environment, right? You could, you could create a thing that just peeled off all the anime fandom or all the, you know, even all the medical support groups, right? That's a, that's an area where Facebook is like really taking a lot of hostages. I, I know someone who started a medical pre-viver, cancer pre-viver group, which is people who have a cancer gene, but don't have cancer yet. Oh. Uh, and Facebook actively courted them. And so she brought her group there. She established this group on Facebook. And then she discovered a bunch of defects in the way that Facebook managed groups
34:35 Speaker 00 with highly sensitive information, like having a, a gene that puts you at a high risk of cancer. She discovered that you could enumerate the membership of any Facebook group, even if you weren't a member of it. And when she reported it to Facebook, they said, that is a feature request, not a bug report. We will send it off to the product team. And when she made a big deal out of it, they said, okay, we'll meet you halfway. Now you can only enumerate the full membership of any group you're a member of, which is still potentially very, very risky. Yeah. Right? So imagine that this person who's, you know, currently suing Facebook over this. Imagine if they could have a Facebook alike. Yeah. That, uh, didn't rely on everybody who used that group deciding to give up Facebook en masse, which is a collective action problem that's very hard to solve.
35:20 Speaker 00 And instead said, okay, here is the place where we now conduct our cancer pre-viver thing. We understand that everyone is over on the other one, so it's got a bot. It logs in as you. It grabs your waiting messages from that group. It brings them over here. It lets you push them back over there. It allows you to block off sections of messages that are encrypted to a shared key that can't be read if you're on Facebook. So parts of it are just off limits to Facebook scrutiny. And every message comes with a footer that says, sent from this rival.
35:50 Speaker 00 Thirty-three percent of the members of this group are now using it. Once it reaches 51%, we're going to shut it off. Don't miss, you know, don't miss the boat, right? That's a product I would use. I pay to use that product if I were a cancer previver and worried about Facebook taking all my data. The thing is that right now, people have this, this all or nothing offer from Facebook. Either they get the support that they badly need to go on living their lives, or they walk away from it. And the quid pro quo is you have to let Facebook kind of plunder every, every.
36:22 Speaker 02 I mean, there, there are still some fringe communities online.
36:25 Speaker 00 Sure.
36:26 Speaker 02 That have their own sites. They host their own stuff.
36:28 Speaker 00 Yeah.
36:28 Speaker 02 Maybe they host it on Amazon, but.
36:30 Speaker 00 The ones that got courted by Facebook and slurped into Facebook.
36:33 Speaker 02 Yeah.
36:34 Speaker 00 Are now stuck there, right? They're now, they.
36:36 Speaker 02 Yeah, for the most part, yeah.
36:37 Speaker 00 And Facebook is, you know, what Facebook is good for, and, you know, this kind of dovetails into another important topic, I think, which is whether or not Facebook is a persuasion engine. What, what Facebook is mostly good at is finding people. You know, if you're a refrigerator salesman, you know, you've had this conundrum all your, all your life, as have all your forbears since the dawn of refrigeration. Cause the median person buys one or fewer refrigerators in their life. This is why refrigerator ads appear beside, you know, freeways, because you've just got to blitz them and hope someone's buying a refrigerator. You know, the, the conversion rate on a refrigerator ad through normal means is probably like 0.0000001%.
37:16 Speaker 00 Facebook lets you find people who have correlates of, of refrigerator purchasing, and they don't have to be very good to improve a conversion rate that's that bad.
37:23 Speaker 02 Yeah.
37:23 Speaker 00 Right? You can target anyone who's looked for a new kitchen from Ikea. You can target everyone who's read a refrigerator ad. You can target anyone who's read refrigerator reviews, any of those things, right? All those correlates, bought a new house, whatever. And even if you multiply the conversion rate by a thousand, which maybe you could.
37:41 Speaker 02 Yeah.
37:42 Speaker 00 You're still at 0.00001% conversion rate.
37:44 Speaker 02 That's worth a lot.
37:45 Speaker 00 And it's worth a lot.
37:46 Speaker 02 Yeah.
37:46 Speaker 00 But it's not mind control.
37:49 Speaker 02 Yeah.
37:49 Speaker 00 And so we hear people say, well, what Facebook does is it lets.
37:55 Speaker 00 Cambridge Analytica find normal people and turn them into Trump voters. Right? And certainly Cambridge Analytica made that claim. They cited all this junk science about the big five personality types and sentiment analysis and all this stuff that when you actually read the literature from people who don't work for companies peddling this crap, they say, oh, you know, it's thin to nonexistent, small sample sizes, small effect sizes, dodgy methodologies, P-hacking like crazy. You know, it's just phrenology 2.0, right?
38:24 Speaker 02 I haven't read it. I'm not sure.
38:26 Speaker 00 You know, it's, people have been claiming to have mind control rays for a thousand years. No one has yet built a mind control ray. If Cambridge Analytica had a mind control ray, they could have made trillions just getting people to like, lay off the carbs and lose 20 pounds, which right now is, you know, one of the biggest market opportunities we have. And they would have found willing subjects as opposed to unwilling subjects. Cambridge Analytica's failure to exploit any of the things that you could actually get super rich from if you could control people's minds tells you that probably what they were doing was finding racists and convincing them that Trump was someone to vote for, not finding normal people and convincing them to be racist.
39:03 Speaker 02 So what do you say about YouTube then? Because YouTube finds things that are like, everything is, it becomes a hub, right? And there's a little spoke. And maybe this thing is a little bit more tantalizing. And then maybe the next thing's a little bit more tantalizing.
39:17 Speaker 00 So this, this raises this important question. Why is it that people are vulnerable to conspiracy theories? Right? If, if what these algorithms do is find people based on traits, why is the trait of vulnerability to conspiratorial thinking seemingly on the rise? Right? Or, you know, I guess there's an alternate explanation, which is that they found a previously untapped seam that we never knew how to find conspiratorial thinkers because they're all in bunkers, you know.
39:46 Speaker 02 Well, just think about it. It's like one of your siblings, but like they don't meet each other.
39:49 Speaker 00 Right, right. So maybe they're just so thinly dispersed that, that they, that we didn't know that maybe they were always there. I don't think that's the truth, though. I think that we have some well-known correlates of conspiratorial thinking. And the biggest one is conspiracies, right? When you live in a world in which there are a lot of conspir- Conspiracies, it's hard to know what's true. So you know Putin's head of media strategy in his cabinet cheerfully admits that some opposition groups to Putin are secretly funded by him out of a slush fund. But he doesn't say which ones. And so some of the groups that are agitating to overturn the Russian state are inauthentic. And you never know which one you're in. And so when you're in one of those groups and someone says, I actually think that you're a Jean Provocateur.
40:34 Speaker 00 You're like, maybe they're in a Jean Provocateur, right? Because we know that these groups are shot through with them. Yeah. When you scratch a ufologist, you find someone who knows an awful lot about military coverups, not fake imaginary military coverups, actual aerospace and military coverups, of which there is no shortage. When you talk to people, primarily African American people who believe that during Katrina, they blew the levees to flood the black neighborhoods and spare the white neighborhoods. Inevitably, they will know an awful lot about Tupelo, Mississippi, where forty, fifty years earlier, that actually did happen, right? And when you scratch an anti-vaxxer, you will find someone who will tell you chapter and verse about the opioid crisis.
41:17 Speaker 00 And about the fact that the Sackler family weaponized some junk science and a few very vague early indicators about the addictiveness of opioids and their safety and whether or not we were undertreating pain. And they suborned the regulatory process and the research process to create the opioid epidemic, which claimed more American lives in the Vietnam War, and made them richer than the Rockefellers in the process. That's true, right? So that rise in conspiracies that we're experiencing now, where there are, there's tons of industry collusion with regulators.
41:56 Speaker 00 There's lots of regulatory capture. Industries converge on sets of policies that if they were legitimately competitors, you would expect that one of them would defect from their consensus, but they actually all line up behind the same thing. Look at the cable industry and its universal condemnation of net neutrality, right? If we had a competitive cable industry, you'd expect to Find a cable operator would say, sign up with us, we're the neutral one.
42:18 Speaker 02 Yeah.
42:19 Speaker 00 And you know, you have that here, you have Sonic, but Sonic is in single digits.
42:22 Speaker 02 Fraction.
42:23 Speaker 00 And they don't, they don't, they don't play on the national stage. Ting is another one that buys mom and pop cable operators and rolls out fiber loops. But most, most people don't have a competitive market.
42:34 Speaker 02 Yeah.
42:35 Speaker 00 You know, you look at that photo, that infamous photo of the tech leaders around the table with Donald Trump and Trump Tower after the inauguration. So the first impulse you might have is to say, how can they sell us out like that? But the second impulse should be, how is it that everyone who runs tech fits around a single table? Right? And, and once they do, how hard is it for them to converge without an explicit conspiracy? On a set of mutually beneficial policies, like the policies that all of tech arrived at. Along with the big entertainment companies to do no poaching in California. So California has a legislative prohibition on non competes. It's why we have Silicon Valley, you know, Silicon Valley was founded by a Nobel Prize winner.
43:16 Speaker 00 Who later went crazy, right? The guy who, who, who figured out how to make transistors out of silicon instead of gallium arsenide shockly. After he got his Nobel money and started the first semiconductor company here in Silicon Valley. Then went nuts, became paranoid, became an ardent eugenicist, spent his Nobel money paying black and brown people to be sterilized. And started wiretapping his colleagues and his family. And it got so bad that his eight top hires, these Hungarian mathematicians quit and started a company called Intel. And if it hadn't been for the non enforceability of non competes here in Silicon Valley in California.
43:55 Speaker 00 That would have been the end of semiconductors, right? They would have been held hostage to the mad fancies of a eugenicist. And so in California, we have no enforceable non competes. So the big tech companies all got together in secret with the big entertainment companies and agreed that they wouldn't hire each other's employees, top employees.
44:13 Speaker 02 A lot of people don't talk about this. Yeah.
44:15 Speaker 00 Yeah. So how did they get there? Right? So how is it that we have all these conspiracies? Well, those conspiracies are necessarily a function of market concentration. Because the collective action problem becomes insoluble. When you have too many people who have to be inside the conspiracy. If there are a hundred cable operators, it's going to be really hard for them to arrive at a single lobbying position. When there's one per district, when they've carved up the country, like the Pope carving up the new world between Spain and Portugal, and they never have to compete with each other, it's very easy for them to agree that they'll scratch each other's backs. And so I believe that the rising conspiratorial thinking is part of a, an epistemological crisis, right? A crisis in how we know whether something
45:00 Speaker 00 is true. Because, you know, it is hacker-ishes as anything to go, okay, well, I'm going to find out what's true by researching it myself. But even the most ardent hacker can't do that for everything.
45:10 Speaker 02 You can't know everything about everything.
45:11 Speaker 00 Yeah.
45:11 Speaker 02 But there's a basic human trait, though, to like, once you figure out a little thing, you just apply it all the time, right?
45:17 Speaker 00 Sure.
45:18 Speaker 02 And you're, so you're arguing that has changed, or is like a step function higher now?
45:22 Speaker 00 Yeah, I think that we historically, we used, instead of understanding how things worked, we understood how the process by which things were known.
45:31 Speaker 02 Okay.
45:31 Speaker 00 We would say, okay, well, we have these truth-seeking exercises, a regulator and expert body, a blue-ribbon panel, gathered together, taken competing claims from different people, different firms, people with self-interest, academics, whatever.
45:42 Speaker 02 Yeah.
45:42 Speaker 00 Sort through those claims, show their work about how they arrived at a conclusion, publish that conclusion along with their work, along with registers of their interests, you can see whether or not they might have been suborned, they are bound by strict ethical guidelines. And then there's a whole rule of law set of fallback positions when new facts come into evidence, if we need to revise our understanding of the truth. But the truth-seeking exercise has become an auction. A good example of this is in West Virginia. So, West Virginia, we think of as coal country, it's actually chemical processing country, it's the biggest industry. And because it's highly concentrated, like every industry, Dow Chemical is, basically swamps all of its competitors there. So, in the chemical industry lobbies in West Virginia, it's Dow. And Dow, with through its lobbying arm, just intervened in a regulatory process
46:27 Speaker 00 in West Virginia that was asking the question, should we relax the standards for how much chemical waste can go into drinking water? And there's a national level that's set by the FDA, and they were seeking to determine whether or not they could have a variance that would allow for. For more chemical processing. Now, on its face, that's not a terrible question to ask, right? There's certainly levels of of waste waste product that can go in your water without making you sick or harming you in any way. And it would be silly to set that number at zero, because then it would make the useful things that we get from chemicals much more expensive. But Dow's answer to this was, of course, we can afford to have higher levels of toxic
47:09 Speaker 00 waste in our drinking water. Those national levels were based on the national BMI. And here in West Virginia, we are so fat that the poison will be diffused in our tissues. And besides, West Virginians hardly drink water, right? So you remember when we only had medical marijuana in California, not recreational marijuana. And if you want a marijuana, you had to write a reason for it. And you could just write, like, you know, trouble sleeping. Yeah, right? This is what you write in the box if you're Dow Chemicals, and you know, you get whatever you want. You know that you bought the truth, right? Well, if you don't know whether your drinking water is going to kill you, and you don't know whether the reinforced steel joists in your roof are sufficient to keep it from falling on your head, and you don't know whether the tailpipe emissions from your diesel are going to poison you and your family, and you
47:54 Speaker 00 don't know whether your anti-lock braking system is secure, or whether or not the Credit Bureau that has compiled a non-consensual dossier on you and all your financial information is about to breach it. And you can't trust the process by which we determine whether those things are true or not. Then what you end up with is instead of the rule of law, you end up with the rule of man, where you find someone who says a thing that you know to be true that turns out to be true, and you follow them, right? It's like the Hitchhiker's Guide to the Galaxy. If you have the towel, you probably have the toothbrush and the backpack and the rail pass. So you follow people where they lead you once they show you that they know one true thing. Donald Trump says the system is rigged. The system is rigged. He is going to rig it more. But he is the only one in the presidential debates who will say the system is rigged.
48:39 Speaker 00 You know, Joe Biden doesn't say the system is rigged. Joe Biden says, let us return to the time in which we no longer we know we didn't talk about whether the system is rigged.
48:46 Speaker 02 It's crazy. Like, so, well, you're Canadian, but I just got married, and my wife is Dutch.
48:53 Speaker 02 Came into the country, we're doing all the green card stuff. You know, Trump said like, we only want rich people coming in. And I was like, that's a terrible thing to say. But having just gone through the process, it's always been this way.
49:02 Speaker 00 No shit, right?
49:03 Speaker 02 Exactly. I have a green card. I've been through the process. I can tell you, I know exactly where you're coming from. Yep.
49:09 Speaker 01 I saw a recent study, academic study about conspiracy theories. And because people have been saying, wow, there's so many more conspiracy theories now. And they used a method for analyzing that, which is hard to measure over time. Have conspiracy theories gone up or down?
49:25 Speaker 02 Really?
49:26 Speaker 01 Their method was to look at thousands and thousands of letters to the editor that are all well documented in the New York Times, going back, you know, 75 year, 100 years or whatever. And it's actually gone down. Like the number of letters to the editor saying, I think the system is rigged. I think the, you know, that there's a secret cobble of people determining the price of wheat or whatever. And I don't remember the exact percentages, but from, you know, 36% of the letters down to 20% of letters today or something.
49:56 Speaker 00 That's what reflects an editorial change in what letters the New York Times publishes?
50:01 Speaker 01 It absolutely could be.
50:02 Speaker 00 Right. I mean, that, that might be a very different.
50:05 Speaker 01 And it also may be like, especially super recently, a letter to the editor isn't the way that people communicate about conspiracy theories anymore either.
50:13 Speaker 00 Sure. Yeah, I might just tell you something about the conspiratorial beliefs of people who write letters to the editor as opposed to conspiratorial beliefs per se. I mean, it is very interesting to look at things like the resurgence of the flat earth, which I think, you know, it's, although I don't have an empirical basis for asserting it, it certainly seems qualitatively like the, like the belief in the flat earth is more central than it was.
50:36 Speaker 02 Yeah.
50:37 Speaker 00 10 years ago. Now, and on its face, there's never been a worse time to promote the flat earth, right? Because people have seen the curvature of the earth. They've been in airplanes, right? They've, they've like attached a GoPro to a balloon at their local hackerspace and brought it up and brought it down again, or even at their elementary school. And so. This belief in conspiracy, when you, when you ask those people to explain why they believe that there would be a conspiracy about the flat earth, they will regurgitate a, a kind of blend of real conspiracies and conspiracy theories that are predicated on the existence of real conspiracies.
51:20 Speaker 02 Yeah.
51:20 Speaker 00 And, and so the credibility of the system has been eroded through market concentration and, and corruption, right? That's what corruption is, right? Is it's when, it's when truth-seeking exercises are suborned by parochial interests, rather than regulating the public interest, we regulate in a parochial interest.
51:40 Speaker 02 So we have yet to talk about your work at all. And I'm, I'm related to this. I'm interested in your predictions of the future.
51:47 Speaker 00 Yeah.
51:48 Speaker 02 And especially in, you know, 50-year future, because Joe and I talked about this before we started recording and we're like, man, so many of these essays and like, like speculative fiction works are like very near future at this point.
52:01 Speaker 01 Yeah, well, in the, you know, so like in 1992, Neal Stephenson wrote Snow Crash. And it was brilliantly prescient of a lot of things that would come. And you could say, even though those are kind of self-referential, like Second Life was probably explicitly based on having.
52:16 Speaker 00 Oh, no, totally. When you got hired at Second Life, they gave you a copy of Snow Crash and said, you start Monday, read this over the weekend.
52:23 Speaker 01 So you and your fellow cyberpunk authors are not just predicting the future, but also creating the future, which is a fabulously creative position to be in. And in the past, it seems like the things that came out of, of those novels took 10 years to happen, you know? Whereas now it feels like, it feels like they're all becoming closer term. Well, you know, like Neal Stephenson's latest book, Fall, he was talking about how the internet is going to just fall apart and be discarded. And he calls it the miasma and, and, and that there's so much fake news and disinformation and bots creating news that nobody trusts any of it anymore.
53:08 Speaker 01 And they're just going back to, to just like these little subnets where people are following other people they really, really trust. And similarly, like in Radicalize, your most recent book.
53:20 Speaker 01 A lot of it seems very near term, like writing about things that are happening right now or are just on the verge of happening. Is that a trend, do you think? Because the world's changing faster?
53:32 Speaker 00 I've always been pretty near future in the work that I've written. And it's not all of it, but, but, you know, in the main thrust of my work is very near future. I mean, I was once nominated for a Hugo Award for a novel set, a novella set like 500,000 years in the future, among dueling solar system scale AIs that were fighting over, you know, where their, their major limitations were things like light speed lags and whether their light cones intersected and stuff. But, but for the most part, I've been, I've, I've kept things pretty close to here. I don't think science fiction does predict the future. I, I think it can make the future, which is exciting. But I think that the idea that the future is predictable is itself horrifically fatalistic, right? Like, I think, you know,
54:16 Speaker 00 whether it's the hacker mentality or the activist mentality, you have to believe that the future is up for grabs, that, that something different can happen. There are some things that are like parameters that we probably can't change. If a million people are born this year, then in 18 years, there won't be more than a million 18-year-olds. That's, you know, that's, that's definitely true. But the lower number is zero, right?
54:35 Speaker 01 Yeah.
54:36 Speaker 00 They could all die tomorrow, right? We could, that, you know, I've, I've, I've been to a Passover seder. I know that, I know that the firstborn can all be snuffed out in, in one night. So, you know, we, we, there are some things that we, we don't, that we can say with a certain degree of certainty, but they're actually pretty trivial. And I like to think that what you get when you read a science fiction story is insight into what that writer hopes for and fears about the way technology is interacting with human civilization. And that when you find a science fiction book that is doing well in the world, you find out what the world hopes and fears about technology and their
55:20 Speaker 00 civilization. And that's hugely influential.
55:23 Speaker 01 I, as a founder, look at science fiction as a menu of choices. Like, which future do I want to bring about? You know, there are these, in Snow Crash, there are these tiny computers instead of the big laptops that were around, just barely happening in those days, or desktops mostly. There are these little ones that, you know, that, that were put on the chest and so on. And that was part of my inspiration for creating the world's smallest laptop in the early 2000s.
55:47 Speaker 00 Right.
55:49 Speaker 01 Which is now in the Guinness Records, I guess.
55:54 Speaker 01 The the menu of options is something that I, you know, it's great. It's it's part of the creative process. And and I would I would love that if for the Y C. Community to avail itself of science fiction as part of the idea generating process. Because a lot of founders come into the Y C. With terrible ideas, but they're like, sometimes when they're fantastic founders, they're like, let's just accept them anyway, because I know that the when they're exposed to the richness of this network and the community here, they'll realize, oh, there's a much cooler thing to do.
56:30 Speaker 02 Well, I mean, that's the incredible downside of being in an ecosystem like Silicon Valley, right? Because there's the group think. And then all of a sudden, you just come out with the same derivative ideas, which is why, like, taking the full menu of science fiction and could be so creative and fun.
56:42 Speaker 00 So I I think that in terms of like what I see in the kind of near to midterm. So first of all, we are at a tipping point for our understanding of of the climate risk. I have a model I call the the peak indifference model, which is when you have a problem that's a long way off, whose causal relationships are not crisply understood or crisply demonstrated. It can be hard to even convince people that there is a problem, right? Like, what's wrong with smoking? Smoking is fine. I don't have cancer, right? Now, that if the problem is real and you don't address it, if people don't believe that it's it's a problem, then you build up debt, right?
57:23 Speaker 00 That over time, you will experience more mutations and your likelihood of getting a tumor goes up. Over time, the planet will get warmer. Over time, the lack of encryption in the web will invite more surveillance, man in the middle attacks and so on. And at a certain point, that debt will be so onerous.
57:43 Speaker 00 That the number of people who are indifferent to the problem will monotonically decrease without any outside action, that day to day and week to week, there'll just be more people who have cancer, right? Or whose houses are on fire, or whose lives were ruined by identity thieves, right? And so they now are no longer in need of being convinced that there's a problem, but then they need to be convinced that there's a potential for a solution, right? That denialism can slide into nihilism very quickly.
58:12 Speaker 02 Absolutely.
58:13 Speaker 00 If I spend 30 years unsuccessfully convincing you that we need to care about rhino populations, and then one day you wake up and you're like, okay, well, there's only one left, the logical next thing to say is, I might as well find out what he tastes like.
58:25 Speaker 02 Yeah.
58:25 Speaker 00 Right? So there's only one left.
58:27 Speaker 02 I mean, there's like, going back to what you said before, there's no reason you can't use persuasion for positive.
58:32 Speaker 00 Sure, sure. And so, you know, back to this idea that what the platforms actually deliver is the ability to find people who share beliefs, that that's critical. It's also dangerous, right? It's what created the Movement for Black Lives. It's what created non-binary gender identities, right? That's, you know, these things that are like widely dispersed and not concentrated in any one place. You know, San Francisco was a nexus for gay liberation for lots of reasons, but one is it's like where the Navy used to discharge people who, men who had gotten into trouble for being gay, right? So they actually created, they took a trait that was widely dispersed in the population, and they created a locus for it, and it became an attractor. So if you were gay and you live somewhere else,
59:16 Speaker 00 there was a higher than median chance that you would toddle off to San Francisco. And so it created this, but now we can do that without waiting for the accident of history, right? Now we can, now, if you are trans, or if you like a weird kind of manga, or if you're trying to find people who want to buy refrigerators, you can use the internet to do that. You can use the platforms and their non-consensual surveillance dossiers to do that. The other thing that you can do with it is you can find people who might want to carry a tiki torch through the streets of Charlottesville chanting, "Jews will not replace us," right? So it does allow every kind of idea that is not yet in the mainstream to find potential co-adherence and to build up a community of mutual support that you
1:00:01 Speaker 00 can then use to try and bootstrap what's in like the demi-monde. Into the into the wider world.
1:00:07 Speaker 01 But I like I like this mental model of the pressure like the sort of the the ratio between the pressure for something that's been building up and the and the indifference to it at the same time.
1:00:18 Speaker 00 Sure.
1:00:19 Speaker 01 Like if as a founder, to be able to to have x-ray glasses where you can look around in society and see where that pressure differential is the highest, it seems like it would be a great algorithm for finding your next startup idea.
1:00:31 Speaker 00 Yeah.
1:00:32 Speaker 01 Did you have suggestions?
1:00:33 Speaker 00 To be very Canadian about this, if you're going to skate to where the puck is going to be, you know.
1:00:37 Speaker 01 Okay.
1:00:38 Speaker 00 You know, I think that one of the things we're at a tipping point for now is the understanding that concentrated industries and wealth concentration are both massive contributors to climate change. And again, there's this idea that you could either go with action or nihilism. We might say, okay, well, let's just find some feudal lords and and allow them to decide what we're going to do now that the crisis is upon us because it's too late. Or we might say, actually, we're going to pursue pluralistic solutions. Again, this is one of those things where I'm not making a prediction about which one we're going to do. I'm saying that you get to decide which side of that you're going to be on.
1:01:21 Speaker 01 Are you going to create a startup that that speculates on land that is not going to be flooded and sells it ahead of time?
1:01:28 Speaker 00 Right.
1:01:28 Speaker 01 Or can you be a part of stopping the rising of the sea from in the first place?
1:01:32 Speaker 00 And you know, this idea of I want to return to the idea of science fiction here. People tend to mistake science fiction, in which a lot of things have gone wrong at once for dystopias. It is not dystopian to imagine that complex systems will break down. Engineers who assume that complex systems won't break down are not good engineers. They're idiots, right? That's what leads you to build the Titanic and then not put any lifeboats on it, right? An engineer who understands that, you know, entropy is not just a good idea, it's the law, is someone who plans for graceful failure, right? Who looks at the way things will break down and figures out how we can get them started again. So the
1:02:13 Speaker 01 Right, dystopia is when everything goes to hell at the same time and the one thing going to hell takes the other thing down and there's a positive feedback.
1:02:20 Speaker 00 And no one can, but also where we are painted as the fallen remnants of a once noble species that could at one point figure out things like sanitation and civil, civil engineering and so on, but which has lost that praxis and could never recover it because we are, we are fallen. We have fallen from grace. That's the road, right? The road is the idea that when the lights go off, your neighbors are gonna come over and eat you.
1:02:47 Speaker 01 Right, literally.
1:02:48 Speaker 00 Immediately, yeah. Right? The actual lived experience of people who've been in disasters is that the crisis is when the background refrigerator hum of petty grievance stops and you realize that you have much more in common with the people across the street than you ever had against them. And people pitch in and help each other. There's a wonderful Rebecca Solnit book called A Paradise Built in Hell. That's this closely researched historical text about how people behave in real disasters. And it opens with the 1906 quake here in San Francisco. And the elites who lived in high ground experienced, and that sociologists, they now call it elite panic. They said that the, that the poorers were coming to eat them. And so they sent
1:03:30 Speaker 00 out General Funston from the Presidio with his troops. And he was convinced of two incorrect things. One was that he knew how to use controlled demolitions to set fire breaks. And the other one was that once he set the city on fire, that he couldn't trust poor people to go back and fight the fires because they would loot. And so that's how he came to burn down a quarter of San Francisco, including the Mission. And the actual lived experience of the people who were not in the high ground, the people who built temporary villages in what's now the Golden Gate Park and so on, is, is that there was this enormous feeling of coming together, that everyone pitched in to help everyone else. There's stories of like butchers
1:04:10 Speaker 00 who cleared out everything they had and just cooked it in what's now, what's the park at 18th and just past 18th, Dolores Park, which was then, I think, I believe, a Jewish cemetery. And they set up a cook pot and they cooked all the food in the butcher shop and gave it out to people, right? Building for resilience, building for recovery, when you can see disaster on the horizon, is not just an optimistic thing that is a kindness to your species. It's also good market sense, right? That we are very soon to be in the market for graceful recovery. Not for luxury bunkers where you can wet the bed and wait for, you know, better people than
1:04:54 Speaker 00 you to rebuild civilizations so that you can emerge with your like AR-15 and your thumb drive full of Bitcoin and like live out, live out life in a, in a Frisetta painting where you get like a harem and a bunch of mercenary guards. Like actually building stuff for recovery and resilience, that's, that, there's a, a huge potential future market for that. In terms of adversarial interoperability and what's going to happen to the platforms on the way, you know, our willingness to trust that big firms have our interests at heart and that the processes that regulate them are on the side of the public interest and not the side of the firm's parochial interest, that, that is coming to an end.
1:05:34 Speaker 00 And so if you want to be where the puck is going to be, betting that when you do something that disrupts the big platforms, that public sentiment and therefore lawmakers will be on your side is a good bet. Now, many of those people will time it wrong, right? That many of those people will get there too early. Aaron Schwartz. Yeah. Aaron Schwartz, yeah. And, and others, right? I mean, you know, Power Ventures was a company that scraped Facebook along with a bunch of its competitors like LinkedIn, put it all in a single dashboard. Facebook went after them using the Computer Fraud and Abuse Act, which is the law that was used to prosecute Aaron. They successfully argued that violating terms of service was a Computer Fraud and Abuse Act violation, building on a precedent set by Blizzard when they went after people who
1:06:17 Speaker 00 made their own B-Net Battle.net server called B-Net D. And, and, and set this terrible precedent. Which shouldn't stand. I mean, it's it. First of all, I think if it if if there was a circus blood and it went to the Supreme Court, they would lose. But we should have we should have legislative clarity on this, that, you know, you cannot create a felony out of thin air by sticking a sprawling garbage novella of legalese in front of your product that says doing anything disfavored by our shareholders is illegal. Yeah. Right? I mean, that that should just not be how our law works. It's not a just system of law. It's obviously not a justice of law, not least because we're at the system of law when all of these firms were deploying it were getting started. None of them would
1:07:00 Speaker 00 have succeeded. Right? So getting the scraping time right, it's it's going to be hard. But, you know, Haikyuu was scraping LinkedIn to produce employer analytics. LinkedIn turned a blind eye to it until they launched their own analytics product. Then they tried to block them out and sue them under the Computer Fraud and Abuse Act. Not only is Haikyuu won in court and had a judge say you can't use the Computer Fraud and Abuse Act to legally punish people who scrape your site. The judge also said you can no longer use technical countermeasures against Haikyuu, which is crazy. So they're like, no rate limiting, nothing like that? No rate limiting, none of that. You can't block their IP addresses, nothing. You
1:07:45 Speaker 00 have to let them scrape you. Right? Because there this is a pro competitive, public interest based. Where was this? Activity. Ninth Circuit, here. Really? Yeah. So who would have thought it? Right? So the Ninth Circuit's really interesting because it's it's the tech circuit, right? It's also the entertainment circuit because California has this dual nature. So sometimes you get that conflict where you have crazy copyright decisions coming out. Sometimes you have ones that are like, really skeptical of copyright claims and and it's very hard to guess which judge is going to jump which way. But the other time where it gets weird is that some of the bench are like, well, I'm here to support new entrants into the market because that's the heart and soul of what makes Silicon Valley Silicon Valley. And some of them are like, I'm here to support
1:08:28 Speaker 00 the kings of Silicon Valley because they're what makes Silicon Valley Silicon Valley. What would Silicon Valley be without Google? Now, you know, Same thing Google, the same thing Silicon Valley would be without, I don't know, um, deck. Yeah, exactly. Or cray. Like HP, whatever. Silicon Graphics. Yeah. I mean, literally, I have been to the Googleplex when it was Silicon Graphics. I've been to it when it's, when it's the Googleplex. I will happily visit it again. When it's whatever. When it's the Hi-Q Plex. Yeah. And built on nothing but building better scrapers, right? Right.
1:08:58 Speaker 01 Just circling back to Y C founders then, uh, you think that they can now proceed because of this Ninth Circuit decision with, uh, with less fear and build some companies based on massive scraping?
1:09:09 Speaker 00 So I am not a lawyer and I'm not their lawyer, but I will say that, uh, if I were talking to an investor about how I was gonna nibble Facebook to death, I would be showing them the Hi-Q decision and not every investor is going to be convinced, uh, not least because a lot of those investors are now invested in companies whose bottom line depends on there not being scraping that disrupts their monopolies. Uh, but money talks, bullshit walks. There is a market opportunity there. It is clearly a giant market.
1:09:42 Speaker 01 And there are a lot of VCs.
1:09:43 Speaker 00 And there are a lot of VCs and there, uh, is a, and that if you do it, you will be building a constituency for more legislative and legal clarity, uh, and also for a better public understanding that scraping is legitimate. You know, I, I got on to this adversarial interoperability kick when I went and spoke at the 50th anniversary of the University of Waterloo's Computer Science Program. I'm a proud U Waterloo Computer Science dropout and, uh, uh, they're whip smart. They call themselves, you know, the MIT of the north, uh, and, um,
1:10:22 Speaker 01 Just like McGill is the Harvard of the north.
1:10:24 Speaker 00 That's right. Yeah. Yeah. Yeah. It's hilarious. As Canadian as possible under the circumstances. Uh, the, uh, The Q and A afterwards, a super smart grad student said like, "Well, I hear what you're saying about Facebook and its monopoly, but how would I ever convince all my friends to leave it at the same time as me? And if I didn't, how would I talk to them?" I'm like, "Are you kidding me? You're a grad student in an elite computer science program. You know, write a very small shell script, done in one, right?" Like, that part has been solved for a long time, and it is a recurring motif in the history of our industry and of every other technological industry. You know, where did cable TV come from? Cable
1:11:07 Speaker 00 TV came from slurping down broadcast signals on a big antenna in rural Pennsylvania that was too far for anyone with a TV to reach either Philly or Pittsburgh, and then charging money to retransmit them over wires to people who own televisions in the small town. It was created by a TV salesman, and it was called Community Access TV or Community Antenna TV, CATV, before it was called cable. It was adversarial interoperability. You know, if broadcasters had had the copyright or patent or computer fraud and abuse defenses or whatever, we would have never had cable TV. Every industry has an incumbent disrupted by a new entrant, not by making a thing that is totally disjoint from the new entrant, not by asking people to leap the chasm from one to the other, to let go of the vine they're
1:11:51 Speaker 00 holding on before they've got a secure handle on the next one. Everyone figures out how to allow people to have a foot on both fields while they figure it out, you know?
1:12:01 Speaker 01 Well, I think that's consistent with the hacker spirit of the YC community, and I'd love to see more of that happening. Thanks for the inspiration.
1:12:07 Speaker 00 Well, thank you very much.
1:12:08 Speaker 02 Thanks so much, Corey. If someone wants to read your work or find out more about you, where do they go?
1:12:11 Speaker 00 I'm often the first Corey in Google, so it's pretty easy to find me. Otherwise, my personal site is called craphound.com and-
1:12:20 Speaker 02 As well as your podcast.
1:12:21 Speaker 00 As well as my podcast, and I'm one of the owners of a website called Boing Boing, boingboing.net, and I'm Dr. O on Twitter.
1:12:28 Speaker 02 Perfect. Thanks so much.
1:12:29 Speaker 01 Thank you. Thanks.
1:12:32 Speaker 02 All right. Thanks for listening. So as always, you can find the transcript and the video at blog.ycombinator.com. And if you have a second, it would be awesome to give us a rating and review wherever you find your podcasts. See you next time.
Previous #151 - Startup School Week 10 Recap - Ali Rowghani on How to Lead and Kevin Hale and Adora Cheung on Startup School 2019 by the Numbers Next #154 - Matt Cutts Page updated October 26, 2026 · transcript and summary are machine-generated