[00:00.830 --> 00:01.930]  Hi.
[00:02.630 --> 00:04.270]  Happy Bastille Day.
[00:05.030 --> 00:05.650]  Hi.
[00:05.950 --> 00:09.350]  What a pleasure and an honor to be back at HOPE.
[00:09.630 --> 00:13.750]  I have been reading 2600, it seems like, all my life.
[00:13.870 --> 00:29.250]  I discovered it through old text files on BBSs, and then read avidly about it with Bruce Sterling's book, The Hacker Crackdown, which later led to my now nearly 25-year association with the Electronic Frontier Foundation.
[00:29.910 --> 00:32.010]  And I'm so pleased to be here.
[00:32.430 --> 00:38.430]  I am a Once Upon a Time 2600 author, and I've spoken at HOPE more than once.
[00:38.570 --> 00:44.750]  And so this is really like a childhood dream realized and a community that I'm so proud to be a part of.
[00:44.910 --> 00:46.570]  And I thank you for supporting it.
[00:46.690 --> 00:52.890]  2600 and HOPE are a really important part of the history of the world and the future of the world, too.
[00:53.510 --> 00:58.190]  And I also thank all of you wearing your EFF swag for your support of EFF.
[00:58.190 --> 01:05.630]  I like to think that that organization that I've devoted my entire adulthood to has something important going on as well.
[01:06.130 --> 01:10.890]  So you may have heard this term that I coined enshittification.
[01:11.890 --> 01:15.650]  The American Dialect Society made it its word of the year in 2023.
[01:16.090 --> 01:26.230]  I resigned myself to having a poop emoji on my headstone, and I've quite enjoyed hearing lots of other people use it in all kinds of ways.
[01:26.230 --> 01:33.230]  One of the things that I relish about being a speaker of English is it's a language where words mean whatever they say they do.
[01:33.410 --> 01:38.810]  So if anyone ever tells you you're using enshittification wrong, tell them that I said you're allowed to.
[01:38.950 --> 01:42.110]  That there is no wrong way to enshittify.
[01:43.050 --> 01:45.770]  So how do I use enshittification?
[01:45.770 --> 01:59.670]  Well, for me, it describes how all of our platforms are getting worse, how they're all getting worse at once, and then maybe worst of all, how they keep getting worse but not dying.
[02:00.850 --> 02:07.050]  And there are three paradoxes in this enshittification hypothesis.
[02:07.610 --> 02:11.590]  The first is, like, why are there so many platforms?
[02:12.170 --> 02:19.830]  Do you remember, if you're old, do you remember that there was a time when we were excited about the Internet as a force for disintermediation?
[02:20.270 --> 02:27.030]  That it would make it possible for people to speak directly to each other without having these middlemen who got in the way?
[02:27.030 --> 02:40.730]  How was it that we declared war on all of these intermediaries, these middlemen, and then just created a bunch more middlemen, what Tom Eastman calls five giant websites filled with screenshots of text from the other four?
[02:41.250 --> 02:49.830]  And I think there's an answer to that that's actually pretty straightforward, which is that middlemen are, broadly speaking, useful, provided they know their place.
[02:50.050 --> 02:53.970]  You know, there was a time when every programmer had to write their own I.O.
[02:54.030 --> 02:56.410]  routines because OSs didn't exist.
[02:57.030 --> 02:57.270]  Right?
[02:57.490 --> 02:58.930]  That was a pain in the ass.
[02:59.070 --> 03:01.830]  You just wanted to tabulate a ballistics table.
[03:01.950 --> 03:03.090]  You didn't want to write an I.O.
[03:03.190 --> 03:03.510]  routine.
[03:03.750 --> 03:11.890]  And so, eventually, operating systems came along and we started using them, and there's nothing wrong, per se, with the existence of operating systems.
[03:11.890 --> 03:19.410]  I'm here today to speak on behalf of the operating systems and to defend them as a useful part of our technology stack.
[03:19.670 --> 03:30.930]  The problem, as we keep discovering with operating systems and other kinds of platforms, is that when they get too much power, they cease to be a way of helping you and they become a way to constrain you.
[03:32.010 --> 03:39.850]  There's a second paradox, which is, okay, we have platforms because they're useful, but why are they all going sour now?
[03:40.430 --> 03:46.790]  And then there's a third part of that paradox, which is, how is it that they keep getting worse, but we still find ourselves using them?
[03:46.970 --> 03:52.090]  And that's what the majority of this talk is going to be about, answering those other two paradoxes.
[03:52.090 --> 04:13.150]  So, to get to that answer, I want you to consider the possibility that the Internet has become pathological, that it is diseased, that it is gripped by a pandemic, that that pandemic and shittification can be broken down like any other disease by thinking about its symptoms,
[04:13.610 --> 04:21.170]  its underlying mechanisms, and its epidemiology, and that when we do that, we can figure out what to do to mitigate that disease.
[04:22.090 --> 04:23.930]  So, let's start with the symptoms.
[04:24.110 --> 04:36.170]  Let's start with the natural history of inshittification, what it looks like to observe an institution as it inshittifies from the outside, how the patient presents to the clinician.
[04:37.170 --> 04:43.470]  When a company contracts inshittification, it undergoes a very characteristic three-stage process of degeneration.
[04:44.270 --> 04:47.810]  In the first stage, it is good to its users.
[04:48.550 --> 04:56.590]  In the second stage, it abuses those users in order to acquire a surplus that it can use to make life good for its business customers.
[04:58.110 --> 05:08.070]  Finally, it abuses those business customers and claws back all the value for itself, and then it turns into a pile of shit, and that is the end stage of inshittification.
[05:08.970 --> 05:10.790]  We'll do a little case study here.
[05:11.390 --> 05:24.470]  The ultimate patient for this is obviously Facebook, a company that was founded to allow Mark Zuckerberg and his creepy friends to non-consensually rate the fuckability of their fellow Harvard undergraduates, a company that only got worse after that.
[05:26.730 --> 05:30.710]  So, in 2006, Facebook decided they were going to open the floodgates.
[05:30.850 --> 05:33.550]  Prior to that, you needed a .edu email address.
[05:33.670 --> 05:36.430]  You had to be an American college kid or know one to use Facebook.
[05:36.630 --> 05:39.310]  But after 2006, they threw open the floodgates.
[05:39.790 --> 05:48.730]  And they had a problem, which was that everyone out there who wanted social media already had an account on a rival service called MySpace.
[05:48.730 --> 05:54.050]  So they needed to pitch those users and explain to them why they should leave MySpace and come to Facebook.
[05:54.190 --> 05:57.230]  And they said, look, I know Facebook is fine for you.
[05:57.330 --> 05:58.110]  It's where your friends are.
[05:58.230 --> 06:02.310]  But has it occurred to you that Facebook... or MySpace, rather, is fine.
[06:02.690 --> 06:11.410]  But has it occurred to you that MySpace is owned by an evil, crapulent, senescent Australian billionaire, and he spies on you with every hour that God sends?
[06:12.410 --> 06:13.890]  Come to Facebook.
[06:14.870 --> 06:19.870]  We are the privacy-forward alternative to MySpace.
[06:20.870 --> 06:26.170]  We will never spy on you or monetize your data.
[06:27.490 --> 06:33.250]  All you need to do is come to Facebook and articulate your social graph.
[06:33.410 --> 06:35.150]  Tell us who matters to you in this world.
[06:35.450 --> 06:45.810]  And we will compose a feed consisting solely of the things that the people that you want to hear, hear from, have published for the consumption of the people who want to hear from them.
[06:46.330 --> 06:47.870]  And that is our pitch to you.
[06:48.130 --> 06:50.290]  And so that is stage one.
[06:50.490 --> 06:52.530]  Facebook has this surplus.
[06:52.830 --> 06:54.350]  Peter Thiel gave them a bunch of money.
[06:55.170 --> 07:01.230]  And they allocated it to these end users by giving them a feed of the things that they wanted to see and not spying on them.
[07:01.510 --> 07:04.630]  And the users piled in because that's a pretty attractive proposition.
[07:05.410 --> 07:08.330]  And the users, once they piled in, locked in as well.
[07:08.330 --> 07:11.290]  Now, there are lots of different ways that tech companies practice lock-in.
[07:11.430 --> 07:14.050]  But with social media, it's actually straightforward.
[07:14.210 --> 07:15.430]  The users lock themselves in.
[07:15.490 --> 07:20.350]  They lock themselves in through something that economists call the collective action problem.
[07:20.970 --> 07:31.770]  So the collective action problem is just the fact that, like, after this conference is over and, like, you and eight friends want to go for a drink, you're not going to be able to figure out where to go, right?
[07:31.770 --> 07:37.730]  It's the fact that you and six people in your group chat can't agree on what board game to play on Friday or what movie to see, right?
[07:37.850 --> 07:43.690]  Collective action problems are really hard for even small groups of people who like each other and have generally common interests.
[07:44.230 --> 07:47.550]  With Facebook, you know, it becomes transcendental, right?
[07:47.550 --> 07:56.550]  Once there's, like, 200 people in your social graph, and some of them are on Facebook because they have a rare disease, and that's where everyone else with the rare disease is, and it's how they support each other.
[07:56.830 --> 08:08.350]  And other people are there because that's where their customers are, or that's how their kids' little league team organizes the carpool, or it's how they stay in touch with the people in the country that they left behind when they moved abroad.
[08:08.350 --> 08:15.290]  You have to convince all of them to leave Facebook, and they have to convince all those other people to leave Facebook, and you have to do it all at once, right?
[08:15.290 --> 08:21.270]  So it's just transcendentally hard, which gives rise to another term out of economics, switching costs.
[08:22.110 --> 08:26.810]  Switching costs are everything you have to give up when you leave a platform, right?
[08:26.950 --> 08:38.670]  In this case, it would be everyone who supports you with your rare disease, the people who organize your kids' little league, carpool, the customers that you rely on, your relations with your family, and all the other people that are important to you.
[08:38.970 --> 08:56.590]  And so once users arrive at Facebook and lock themselves in, Facebook can abuse the fact that it costs a lot to leave to make life worse for them, but not just for shits and giggles, not because they're cruel, but because it allows them to bring in another group of users.
[08:56.770 --> 08:58.690]  In this case, it's business customers.
[08:59.010 --> 09:01.890]  So platforms, they have two different groups of users.
[09:01.990 --> 09:03.650]  They have end users and business customers.
[09:04.190 --> 09:09.490]  The business customers that Facebook wanted to bring in were two big groups of them at this point.
[09:09.990 --> 09:12.670]  One group were publishers and the other group was advertisers.
[09:13.210 --> 09:19.270]  So they went to the advertisers and they said, Hey, do you remember, we told these rubes that we weren't going to spy on them?
[09:19.910 --> 09:20.970]  That was a lie.
[09:21.750 --> 09:24.970]  Actually, we spy on them from asshole to appetite.
[09:25.390 --> 09:32.990]  And all we ask from you is that you give us a remarkably small amount of money and a detailed dossier on the kind of users that you want to reach.
[09:33.190 --> 09:37.330]  And we will show them ads precisely tailored to your needs.
[09:37.450 --> 09:46.070]  And moreover, we're such good-natured slobs that we've rented a whole building next door and filled it with nothing but security engineers to police ad fraud.
[09:46.290 --> 09:51.370]  So when you give us a dollar to show a user an ad, that user is going to see that ad.
[09:51.370 --> 09:53.150]  That dollar will be well spent.
[09:53.430 --> 09:54.770]  So the advertisers pilot.
[09:55.390 --> 10:07.090]  As another group of business customers, they want to reach the publishers and they turn to publishers and they say, Hey, do you remember when we told these rubes that we were only going to show them the things that they asked to see?
[10:08.230 --> 10:09.850]  That was also a lie.
[10:10.610 --> 10:28.210]  All you need to do is take an excerpt from your perfectly cromulent website where you have a great way to monetize your traffic, put it on Facebook, add a link back to that perfectly cromulent website, and we will non-consensually cram it into the eyeballs of people who never asked to see it and who
[10:28.210 --> 10:29.750]  we promised we would never do this to.
[10:30.230 --> 10:38.690]  And they will sometimes click that link and visit your website or even subscribe to you on Facebook and you will get a free traffic funnel that you can monetize as you see fit.
[10:38.990 --> 10:43.810]  So the business customers become locked in by the users and the users are locked in by each other.
[10:44.730 --> 10:47.130]  So now it's time for the third stage of initiative.
[10:47.130 --> 10:55.330]  In the third stage of initiative, the company claws back the surpluses it's allocated to these business customers as well.
[10:55.730 --> 11:03.930]  And you may have heard like if you're not paying for the product, you're the product as though paying for a product makes a company decide to treat you well.
[11:04.730 --> 11:06.610]  It's not the case, right?
[11:06.770 --> 11:12.210]  John Deere doesn't give away free ad-supported tractors and that's why they screw you if you want to fix it yourself.
[11:12.830 --> 11:16.330]  The fact is that companies will treat you as the product if they can.
[11:16.610 --> 11:18.650]  And at this point, Facebook could.
[11:19.150 --> 11:24.490]  And so for advertisers, the quality of ad targeting went down and down.
[11:24.850 --> 11:29.630]  The fraud policing went out the window and the price of ads went up and up.
[11:29.710 --> 11:30.510]  It got really bad.
[11:30.690 --> 11:35.690]  You know, Procter & Gamble was spending $100 million a year on programmatic advertising.
[11:35.690 --> 11:39.070]  It's the euphemism the industry uses for surveillance advertising.
[11:39.690 --> 11:48.010]  And when they turned that budget to a $0 per year budget for programmatic advertising, they saw a 0% change in sales.
[11:48.570 --> 11:51.510]  To a first approximation, no one saw those ads, right?
[11:51.510 --> 11:54.130]  It was just disappearing into a giant ad fraud pit.
[11:54.850 --> 11:56.730]  For publishers, things got worse too.
[11:57.210 --> 12:10.770]  Facebook started to turn the dial so that in order to get your stuff recommended to people or even shown to people who had subscribed to you, the portion of the content from your website that you had to put in that excerpt had to get longer and longer and longer.
[12:10.950 --> 12:21.330]  And of course, that makes it more and more substitutive for a click and your ability to monetize those users until eventually you had to have the whole article there perfectly substitutive.
[12:21.450 --> 12:23.250]  And then Facebook sticks the knife in.
[12:23.570 --> 12:26.530]  They say, oof, that link at the bottom of your article.
[12:26.530 --> 12:28.510]  We can't tell if it's malicious or not.
[12:28.510 --> 12:30.930]  So we're going to downrank anything that links off the platform.
[12:31.210 --> 12:32.630]  And then they turn the knife.
[12:32.770 --> 12:38.050]  And they say, also, to reach the people who subscribe to you, you're going to have to pay to boost your content.
[12:38.510 --> 12:40.370]  And so now you have to pay to boost your content.
[12:40.930 --> 12:43.090]  The only way you can monetize it is with the ad market.
[12:43.210 --> 12:44.670]  The ad market is riddled with fraud.
[12:44.670 --> 12:58.970]  And if you're a user, the quantum of material in your feed that you have asked to see dwindles to a homeopathic residue, leaving behind a void that can be filled with things that people will pay to show you, ads and boosted content.
[12:59.530 --> 13:03.490]  And yet there you are locked in because of the people who matter to you.
[13:03.810 --> 13:07.550]  Now, Facebook is trying to arrive at a very brittle equilibrium.
[13:07.890 --> 13:20.830]  They want to reach the point where they've withdrawn all the available surpluses from the platform, but left behind a sufficiency of surplus that the users stay locked to each other and the business customers stay locked to the users.
[13:20.830 --> 13:22.650]  The problem is that's a moving target.
[13:22.650 --> 13:23.770]  It's not static.
[13:24.010 --> 13:36.890]  Because the difference between, oh, I hate Facebook, but I find myself logging into it every single day and, oh, I hate Facebook, but I'm never going to log into it again, hallelujah, it's razor thin, right?
[13:36.990 --> 13:45.330]  Just one whistleblower, one Cambridge Analytica, one live stream mass shooting, and people bolt for the exits.
[13:45.330 --> 13:57.670]  Now, when that happens, Facebook gets into a lot of trouble because the markets really don't like to see platforms that are contracting or even seeing a slowdown in their growth.
[13:57.810 --> 14:05.090]  You may recall that in the first quarter of 2022, Facebook posted a slightly lower than expected growth among U.S. users.
[14:05.250 --> 14:07.830]  Not a contraction, slightly less than anticipated growth.
[14:07.830 --> 14:14.130]  And in 24 hours, they saw the largest sell-off of any company in the history of public markets.
[14:14.370 --> 14:21.490]  A quarter of a trillion dollars wiped off of Facebook's value in one day.
[14:22.050 --> 14:24.290]  Now, lots of people hold Facebook shares.
[14:24.290 --> 14:30.430]  If you're invested in an index fund with your retirement savings, you hold some Facebook shares.
[14:30.430 --> 14:37.610]  But the people whose portfolios are, like, stuffed full of Facebook shares are product managers at Facebook who get paid in stock.
[14:37.610 --> 14:45.770]  Because if you pay your product managers in stock, which you can print infinitely yourself at no price, you don't have to pay them in money, which you can exchange for goods and services.
[14:45.990 --> 14:57.690]  And so you have these people who are sitting on very Facebook-heavy stock portfolios who just watched a quarter trillion dollars get lopped off the company valuation.
[14:57.690 --> 15:03.570]  And that's when we enter the final stage of initiative-ification, the panic stage.
[15:03.790 --> 15:08.010]  Although there's a technical term for panic that the industry likes to use, which is pivot.
[15:10.150 --> 15:19.990]  So Facebook's pivot was that Mark Zuckerberg came to us atop a mountain wearing his sun-god robes and said, I have had a revelation.
[15:21.170 --> 15:29.570]  For more than a decade, I have told you that your future consisted solely of arguing with your racist uncle using a primitive text interface I designed in a Harvard dorm room.
[15:29.690 --> 15:48.870]  However, the future actually consists of me converting you and everyone you love into a legless, sexless, low polygon, heavily surveilled cartoon character inhabiting a dystopian, futuristic virtual world called the metaverse that we stole from a 25-year-old satirical science fiction novel.
[15:50.070 --> 15:54.210]  And that is the final stage of initiative-ification.
[15:55.530 --> 15:57.690]  So that's the final stage of initiative-ification.
[15:57.870 --> 16:00.750]  This is when it turns into the giant pile of shit.
[16:01.210 --> 16:08.310]  Now, all the parties involved don't like this, but all the parties involved remain stuck to the platform more or less.
[16:10.810 --> 16:15.530]  And that is a common factor among multiple platforms that have all gotten worse.
[16:15.690 --> 16:17.670]  You know, Uber used to be very convenient.
[16:18.330 --> 16:20.710]  Drivers used to boast about how much money they made.
[16:21.230 --> 16:23.350]  Customers used to boast about how little it cost.
[16:23.590 --> 16:28.710]  Now the drives cost twice as much, the drivers get paid half as much, and the company just gets creepier by the day.
[16:28.930 --> 16:31.850]  And yet lots of people still find themselves using Uber, right?
[16:31.910 --> 16:36.650]  So this is the procession that we see in platform after platform, this three-stage platform.
[16:36.650 --> 16:41.830]  First they're good to their end users, then they're good to their business customers, then they're terrible to both, and then they turn into a pile of shit.
[16:42.130 --> 16:43.910]  That is the procession of the disease.
[16:44.050 --> 16:45.110]  It is the natural history.
[16:45.230 --> 16:47.550]  It is what the patient presents with to the clinician.
[16:48.030 --> 16:50.450]  But now we need to know how it works, right?
[16:50.550 --> 16:56.270]  What is going on inside the corporate body that causes these symptoms to be displayed?
[16:56.370 --> 16:57.630]  What is its mechanism?
[16:58.430 --> 17:02.390]  Well, it starts with the inner workings of computers, right?
[17:02.390 --> 17:04.770]  Something that I hope is very near and dear to this audience.
[17:04.770 --> 17:09.030]  Computers are so marvelous and so exciting and so fascinating to us.
[17:09.170 --> 17:13.170]  The subject of amazing little quarterly magazines and cons like this.
[17:13.770 --> 17:15.270]  Because they're so flexible.
[17:15.710 --> 17:17.930]  Because they can do almost anything.
[17:18.170 --> 17:24.770]  Because the Turing Complete Universal Von Neumann machine runs every program that we can express in valid symbolic language.
[17:24.770 --> 17:36.350]  And that means that services that are built on top of digital substrates can change how they work from minute to minute and second to second at extremely low cost.
[17:36.470 --> 17:55.150]  They can twiddle the knobs on the back end of the business to change the payout, the cost, the search ranking, and every other aspect of the business logic that you might use and track in order to figure out whether you're getting a good deal, in order to be a good consumer or user of that platform
[17:55.150 --> 17:57.570]  that walks away feeling like you've been treated fairly.
[17:57.570 --> 18:02.850]  So to give you an example of how that plays out in the real world, consider how Uber drivers get paid.
[18:03.210 --> 18:09.190]  There's an anthropologist named Veena Dubel who's coined a term to describe it called algorithmic wage discrimination.
[18:09.950 --> 18:20.190]  If you're an Uber driver who has recently been selective in which rides you're willing to take, Uber will offer you more for the next ride that's available to you.
[18:20.190 --> 18:36.150]  But if you're an Uber driver who's quite indiscriminate in which rides you take, Uber will titrate your payment per mile and per minute down and down by small random intervals, sometimes toggling them up, sometimes toggling them down if you appear to be becoming more selective,
[18:36.150 --> 18:41.270]  with the net effect that over time your wage monotonically decreases.
[18:41.670 --> 18:52.010]  And whatever side hustles you had that allowed you to be so selective about Uber drivers when you started off have been jettisoned, trapping you in the platform at the lowest possible sustainable wage.
[18:52.230 --> 18:59.210]  Now, this is a pattern that many Uber drivers have experienced, and it's the kind of thing that bosses have dreamt of doing for hundreds of years.
[18:59.250 --> 19:01.830]  And it's not that they didn't do it because they lacked the imagination.
[19:02.170 --> 19:04.950]  And it wasn't because they lacked the cupidity or cruelty.
[19:05.250 --> 19:12.590]  You know, the 19th century coal bosses that Tennessee Ernie Ford wrote songs about would have happily done this to the coal miners.
[19:12.590 --> 19:25.170]  The problem was that you just couldn't build a big enough office building filled with enough guys and green eye shades to change everyone's paycheck from minute to minute and second to second to make a go of it.
[19:25.810 --> 19:33.250]  If you've ever seen the shell game being slowed down, I recommend Penn & Teller doing the ball and cups with transparent cups and very slowly.
[19:33.590 --> 19:38.030]  You know that the shell game is not a very sophisticated game, right?
[19:38.030 --> 19:42.230]  The mechanism by which the P is hidden and revealed is really straightforward.
[19:42.810 --> 19:46.470]  The thing that makes the shell game so convincing is that it's done quickly.
[19:47.230 --> 19:54.830]  And what computers allow us to do is not sophisticated tricks, but simple tricks done very fast.
[19:55.110 --> 20:04.390]  And those rapid tricks allow the platform to extract surpluses from business customers and end users that they might otherwise bulk at.
[20:04.570 --> 20:19.390]  So that same mechanism underlying algorithmic wage discrimination, that toggling up and down based on willingness to drive or not, was also present when Facebook was monotonically eroding the performance of publishers content on Facebook.
[20:19.710 --> 20:26.270]  When Facebook started, they only required a small excerpt in order to qualify for recommendation in the algorithmic feed.
[20:26.270 --> 20:30.830]  And as they started to require longer and longer excerpts, some publishers balked.
[20:30.930 --> 20:33.590]  They said, oh, no, we're not commodity backends to Facebook.
[20:33.750 --> 20:35.430]  We are a standalone competitor to Facebook.
[20:35.750 --> 20:36.830]  Facebook is a distributor.
[20:37.010 --> 20:40.290]  They're a middleman who is there to help us, not to usurp us.
[20:40.330 --> 20:42.190]  And so we're going to walk away from Facebook.
[20:42.350 --> 20:54.270]  Well, Facebook's algorithm could take notice of the fact that the cadence with which publishers were making content available on Facebook had dropped off, and it could boost the last things that it had in its cache to many users.
[20:54.270 --> 20:58.410]  And those publishers would say, oh, look at all this traffic we're getting from Facebook.
[20:58.630 --> 21:01.110]  I guess I'm the Louis Pasteur of Facebook.
[21:01.290 --> 21:04.090]  I guess I should be putting more content on Facebook.
[21:04.090 --> 21:04.970]  And they'd come back.
[21:05.090 --> 21:12.950]  And by toggling the promotion of content back and forth, they're playing those publishers like a fisherman with a line.
[21:13.150 --> 21:14.990]  And they never get tired.
[21:15.370 --> 21:21.070]  And if they never get tired and the person on the other end is a human, they will eventually wear up and then Facebook will win.
[21:23.530 --> 21:34.350]  The flexibility of digital, that proliferation of knobs and levers on the back end, that platforms can touch whenever they want, that is the underlying mechanism of machification.
[21:34.790 --> 21:45.630]  This is the system of pumps that allows for the transformation of value between different parties in the platform, the business customers, the end users, and the shareholders, that make the whole system run.
[21:45.750 --> 21:47.090]  It's how the disease works.
[21:47.090 --> 21:49.890]  Okay, so that is the mechanism of the disease.
[21:50.050 --> 21:50.770]  We know its symptoms.
[21:50.970 --> 21:51.950]  We know its mechanism.
[21:52.190 --> 21:53.510]  What about the epidemiology?
[21:54.010 --> 21:55.990]  Why is all of this going on right now?
[21:56.110 --> 21:57.770]  What led to the Enchidocene?
[21:57.890 --> 22:00.450]  Why are we experiencing a great Enchidonate?
[22:01.410 --> 22:04.990]  After all, platforms have been digital since we've had platforms.
[22:05.390 --> 22:08.430]  They could have been doing this 10, 15 years ago.
[22:08.610 --> 22:10.530]  Why are they all doing it now?
[22:10.530 --> 22:17.710]  We had the odd platform that went bad, but we never had this kind of mass Dutch elm disease of our platforms.
[22:18.330 --> 22:19.590]  So what caused it?
[22:19.690 --> 22:20.810]  What is it about this moment?
[22:21.410 --> 22:27.350]  Was it the end of the zero interest rate policy when central banks were making money available really cheaply?
[22:27.730 --> 22:30.230]  Was it a change in leadership in the tech giants?
[22:30.870 --> 22:33.210]  Is it because Mercury went into retrograde?
[22:34.170 --> 22:35.670]  I don't think it's any of these.
[22:35.870 --> 22:48.670]  After all, there was a period of extremely low cost, Fred money, and that did give these companies lots and lots of surpluses to bandy around at a very low cost, but Facebook wasn't shittifying a long time before Zerp ended.
[22:49.550 --> 22:51.850]  As did Amazon, Microsoft, and Google.
[22:52.310 --> 22:55.990]  And yeah, some of the tech giants, they did bring in new leaders, right?
[22:56.090 --> 23:05.390]  The founders quit or retired or got forced out, and they got replaced with these, like, bloodless ex-McKinsey-ites who you might expect to be the kind that would squeeze a dollar till it hollers.
[23:05.510 --> 23:08.750]  But you know, some of these platforms got worse when their founders came back.
[23:08.750 --> 23:14.850]  Like, you know, when Google had its AI panic, which, you know, they call the AI pivot, and the Google boys came back.
[23:15.130 --> 23:18.570]  Google got materially, immediately, and lastingly worse.
[23:19.230 --> 23:21.330]  I don't think it's the change in personnel.
[23:22.390 --> 23:25.130]  And I don't think that it's Mercury retrograde.
[23:25.250 --> 23:26.450]  And there's a simple reason for it.
[23:26.690 --> 23:30.710]  I'm a Cancer, and as everybody knows, Cancers do not believe in astrology.
[23:35.880 --> 23:48.560]  So, when a whole bunch of independent entities within an ecosystem all start going bad in the same way all at once, that militates for looking for an exogenous force that's causing that decay, right?
[23:48.620 --> 23:59.840]  It's much more likely that something in the environment changed to cause all those firms to behave in the same way all at once, than that something within the firms is happening in concert and that they're somehow coordinating it.
[24:00.000 --> 24:01.660]  And I think that's what happened to tech.
[24:03.140 --> 24:05.460]  Now, tech companies are like all businesses.
[24:05.640 --> 24:06.920]  They have conflicting imperatives.
[24:07.100 --> 24:10.000]  So, on the one hand, they want to make a lot of money.
[24:10.000 --> 24:13.460]  On the other hand, making money involves spending money.
[24:13.860 --> 24:27.100]  When you hire and motivate competent staff, or you make a product that's of high quality so customers want to buy it, or you charge a low price so customers buy it and feel like they've gotten a good deal, you get less money than you would if you could charge more,
[24:27.520 --> 24:31.200]  pay less, and put less energy into the product itself.
[24:32.220 --> 24:38.760]  The more value the company permits all of its stakeholders to carve off, the less value is retained for shareholders.
[24:39.500 --> 24:55.400]  Which means that when you have an equilibrium in which companies produce high quality products under ethical circumstances that are priced well, what you're living through is an equilibrium in which charging more, treating the suppliers worse, or making the products worse,
[24:55.580 --> 25:04.800]  would cost more to the company, would evince more consequences for the company than they stand to gain by cheating and shitifying, by playing dirty.
[25:05.820 --> 25:13.680]  So there are two forces that discipline every kind of company in the market that cause them to play clean instead of dirty.
[25:13.860 --> 25:15.900]  And then there's two more that are specific to tech.
[25:16.200 --> 25:18.740]  The first one, obviously, is competition.
[25:18.740 --> 25:25.260]  On balance, if a company fears losing your business, they will treat you better.
[25:25.480 --> 25:30.880]  And if they don't, then you'll probably go somewhere else and they'll do worse.
[25:31.240 --> 25:40.180]  And so, historically, when we've had competition, it has generally prevented companies from abusing their customers too badly.
[25:40.180 --> 25:43.740]  The other force that affects every company is regulation.
[25:44.440 --> 25:59.040]  If a company understands that treating its customers or its workers or its stakeholders badly, that it will pay a price that exceeds the expected return from cheating, on balance, it treats those customers better.
[25:59.840 --> 26:02.320]  Now, those affect every kind of business.
[26:02.320 --> 26:14.580]  But with tech, we do have some exceptional forces that act on tech and cause it to be disciplined in new and exciting ways that actually open up some real potential for how we can make things better.
[26:15.080 --> 26:21.340]  The first of those tech-specific forces is self-help interoperability.
[26:21.520 --> 26:30.400]  The fact that end users or their proxies can unilaterally modify the technology that they use because, after all, technology is flexible, right?
[26:30.560 --> 26:44.080]  We have these Turing-complete universal von Neumann machines that run every program that is symbolically valid, and that means that whatever effort some firm takes to enshittify their product can be disinshittified with a counter-program that undoes it.
[26:44.860 --> 26:50.300]  So, imagine that you're around a boardroom table at a product planning meeting for your company's website.
[26:50.360 --> 27:03.280]  And the person running the meeting says, All right, folks, I've done some modeling here, and I've worked out that if we make the advertising 20% more obnoxious, we will see a 2% increase in the revenue per user.
[27:03.680 --> 27:07.000]  And as we all know, our KPI is revenue per user.
[27:07.220 --> 27:12.700]  If we hit that 2% mark, we all get a big bonus, and everybody gets to take a ski holiday in Switzerland this year.
[27:13.980 --> 27:30.520]  After that little impassioned speech, someone at that table might well raise their hand, someone who does not care at all about the user's well-being, and say like, Bob, I like how you're thinking, but has it occurred to you that if we make the ads 20% more obnoxious,
[27:31.220 --> 27:35.940]  40% of our users are likely to go to the search engine and type, How do I block ads?
[27:36.520 --> 27:42.900]  And if they do that, then the revenue for those users, it doesn't stay at the status quo ante of the 100% we're getting now.
[27:42.900 --> 27:46.920]  We certainly don't rise to that 102% that you're hoping to get.
[27:47.160 --> 27:55.600]  The revenue per user drops to zero, and it stays there forever, because no user ever went back to the search engine and typed, How do I start seeing ads again?
[27:56.480 --> 28:13.060]  And so this self-help measure has always stayed the hands of digital companies, because they always had to face the possibility that if they made the ink cost $10,000 a gallon, that someone would go out and figure out how to use generic ink in their printer.
[28:13.260 --> 28:18.980]  That if they wrote a program that enshittified their product, that someone would write a program that dishitified it.
[28:19.180 --> 28:22.740]  So that was the third factor that acted to discipline companies.
[28:22.940 --> 28:31.040]  And the final factor, the second tech-specific factor, the fourth factor overall, that disciplines tech companies, is their workforce!
[28:31.640 --> 28:35.500]  So the tech workforce is a really strangely constituted one.
[28:35.500 --> 28:48.200]  It's not like any that we've really seen before, because it is a very powerful workforce that has held a lot of power over a long time scale, but it's been derived from their scarcity and not from labor rights or labor organizing.
[28:48.840 --> 28:57.280]  So there just weren't enough tech workers, which meant that tech workers could quit their jobs and get better jobs somewhere else if they didn't like their boss.
[28:57.760 --> 29:00.040]  And bosses knew it, and tech workers knew it.
[29:00.040 --> 29:10.340]  Now, this presented a problem for tech bosses, which is how do you motivate these workers to do a lot of work for you when they can quit their jobs and walk across the street and go somewhere else?
[29:10.740 --> 29:17.100]  And the strategy that they hit on, it's something that the librarian theorist Fobazi Ettar calls vocational awe.
[29:17.700 --> 29:20.460]  It's appealing to their sense of mission, right?
[29:20.540 --> 29:26.760]  To say that you are part of a great movement to digitize the human race, right?
[29:26.760 --> 29:35.760]  To bring us into a digital future in which our digital processes and our relationships are utterly transformed by technology that you make.
[29:35.840 --> 29:37.300]  You can make a dent in the universe.
[29:37.300 --> 29:38.680]  You don't have to be evil.
[29:38.800 --> 29:40.000]  You can connect all the people in the world.
[29:40.160 --> 29:52.140]  All those slogans actually worked because tech workers, while they did bargain for high wages and lots of perks around the office, they didn't bargain for mandatory overtime.
[29:53.100 --> 30:01.080]  Another word for vocational awe here is Elon Musk's phrase, being extremely hardcore, right?
[30:01.160 --> 30:09.740]  They slept under their desks, they missed their mother's funeral, they missed their kids' little league games, and they shipped to these arbitrary deadlines that their bosses demanded.
[30:09.740 --> 30:28.680]  And so long as they were on workplaces that had been transformed into whimsical campuses with free kombucha and massages on Wednesdays and a doctor who'd freeze your eggs so you could work through your fertile years, you could tell yourself that you were being pampered because you were an important
[30:28.680 --> 30:34.680]  part of a great future in the offing, and not that you were a sucker who was getting worked like a government mule.
[30:35.580 --> 30:38.480]  And so that worked well, but it failed badly.
[30:38.480 --> 30:56.680]  For a long time, when tech workers were ordered to enshittify the products that they'd worked so hard on, when their bosses said, you know, take this thing that you missed so many days with your family in order to make my deadline on and make it 30% worse so that I can have a ski holiday in
[30:56.680 --> 30:59.480]  Switzerland this year, they told their bosses to fuck off.
[30:59.960 --> 31:04.640]  I said, the other guy across the street will give me a better job if I leave.
[31:05.500 --> 31:08.220]  I did not miss my mother's funeral for this.
[31:08.460 --> 31:11.820]  I am experiencing a sense of profound moral injury.
[31:12.040 --> 31:23.720]  That is the problem with motivating your workers by appealing to their sense of mission, is that they'll feel a sense of mission, and they'll tell you that they won't defy it.
[31:24.640 --> 31:30.500]  So that was the fourth force that acted on technology to discipline it.
[31:30.620 --> 31:34.180]  Those tech workers who represented the final bulwark against enshittification.
[31:35.040 --> 31:40.000]  Which means that the pre-enshittification era wasn't a time of better leadership.
[31:40.180 --> 31:42.080]  It wasn't a time when the executives were better.
[31:42.360 --> 31:52.100]  It was a time when the executives were more constrained, when their worst impulses were checked by competition, by regulation, by interoperability, and by the tech workforce itself.
[31:52.580 --> 31:59.800]  And what happened to bring us to this moment is that one by one, each of those constraints eroded until it dissolved.
[31:59.800 --> 32:05.500]  It left the enshittificatory impulse unchecked, and it ushered in the enshitticine.
[32:06.800 --> 32:13.860]  The collapse of those constraints began with the collapse of competition law.
[32:14.460 --> 32:19.880]  From the Gilded Age until the Reagan years, the purpose of competition law was right what it said on the tin.
[32:20.260 --> 32:22.160]  It was there to promote competition.
[32:23.300 --> 32:28.800]  U.S. antitrust law historically treated corporate power as dangerous, and said that it should be kept at bay.
[32:28.800 --> 32:42.900]  But starting in the neoliberal era, a fringe theory in economics, called the consumer welfare theory, took over the orthodoxy of economic practice, first in the U.S., then in Europe and the U.K., and finally all over the world.
[32:43.440 --> 32:59.460]  And under consumer welfare, this theory, this radical theory of antitrust law, when you find a monopoly in the wild, you find a store that everyone shops at and a product that they all buy there, you have not found someone who is cheating and needs to be disciplined by the state,
[32:59.680 --> 33:01.540]  you've found someone who is succeeding.
[33:02.180 --> 33:13.480]  If we all shop at the same store and buy the same thing, that's the best store and it's the best thing, and it would be crazy for our elected representatives to shut that down and punish those companies for succeeding.
[33:14.020 --> 33:18.560]  And so the world's governments, they just stopped enforcing competition law.
[33:18.560 --> 33:22.940]  Enforcers ignored antitrust law as companies flouted it.
[33:23.100 --> 33:29.300]  They allowed the companies to merge with their major competitors to absorb smaller companies before they could grow to become big threats.
[33:29.460 --> 33:40.960]  They oversaw an orgy of consolidation that produced the most inbred industries imaginable, whole sectors so incestuous that they developed the corporate equivalent of Habsburg Jaws.
[33:40.960 --> 33:55.920]  From eyeglasses to sea freight, from glass bottles to payment processing, from vitamin C to beer, from professional wrestling to cheerleading, most of our global economy is dominated by five or fewer large global firms.
[33:56.800 --> 34:13.640]  Moreover, when smaller companies that had successful businesses declined to sell to those cartels that had taken over the sectors, the members of the cartel were free to abuse antitrust law, violate the law without fear of enforcement through things like predatory pricing.
[34:14.320 --> 34:27.740]  So back before Amazon was the everything store, when it was just a bookstore with pretensions, it went around and it found successful e-commerce retailers that had conquered one vertical or another and it bought them.
[34:28.140 --> 34:29.860]  And it folded them into the company.
[34:30.000 --> 34:30.760]  That's how Amazon grew.
[34:30.880 --> 34:39.520]  It's not a company that mastered the art of selling things to everyone, it's a company that bought the companies that mastered the art of selling things to everyone.
[34:39.940 --> 34:41.200]  But one company wouldn't sell.
[34:41.360 --> 34:43.020]  There was a company called Diapers.com.
[34:43.920 --> 34:45.840]  No points for guessing what they sold.
[34:46.460 --> 34:48.420]  And Diapers.com, they had a good business.
[34:48.860 --> 34:53.040]  And so when Amazon came knocking and said, we want to buy you out, But Diapers.com said, no thanks.
[34:53.060 --> 34:54.840]  We'll just keep selling diapers.
[34:55.240 --> 35:05.620]  And so Amazon tapped the capital markets, raised $100 million, and set it on fire, selling diapers significantly below cost until Diapers.com collapsed.
[35:06.000 --> 35:21.560]  And then they bought Diapers.com for pennies on the dollar and shut them down and taught every single investor who might invest in a company that might turn Amazon down that that's not a smart business strategy, that they can remain irrational longer than any investor can remain solving.
[35:24.340 --> 35:27.000]  And so competition became a distant memory.
[35:27.200 --> 35:31.240]  So we got Tom Eastman's five giant websites filled with screenshots of text from the other four.
[35:31.420 --> 35:34.980]  So these companies stopped fearing the loss of our business.
[35:35.300 --> 35:44.360]  You may remember Saturday Night Live, Willie Tomlin used to come out and do this character, Ernestine, the bell operator, as I guess a good one for 2600 magazine crowd.
[35:44.760 --> 35:49.900]  And Ernestine would do these pretend commercials for the AT&T bell system.
[35:49.900 --> 35:53.620]  And each one was funny in its own way, but they all ended the same way.
[35:53.740 --> 35:56.760]  She would turn to the camera and she'd say, we don't care.
[35:57.320 --> 35:58.360]  We don't have to.
[35:58.640 --> 35:59.880]  We're the phone company.
[36:01.980 --> 36:04.960]  Today's tech giants, they're not constrained by competition.
[36:05.300 --> 36:06.280]  They don't care.
[36:06.920 --> 36:07.940]  They don't have to.
[36:08.500 --> 36:11.980]  They're Google and they have 90% of the search market.
[36:13.100 --> 36:14.720]  That's the first constraint gone.
[36:14.940 --> 36:15.380]  Competition.
[36:15.720 --> 36:19.280]  And as competition slips away, regulation is not far behind.
[36:19.880 --> 36:25.240]  When an industry consists of hundreds of small and medium-sized enterprises, it's not a cartel.
[36:25.240 --> 36:25.840]  It's a mob.
[36:25.980 --> 36:26.780]  It's a rabble.
[36:27.180 --> 36:31.660]  Hundreds of companies cannot agree on what to tell Parliament or Congress or the European Commission.
[36:31.660 --> 36:35.740]  They can't even agree on where to hold their annual industry meeting or how to cater it.
[36:35.880 --> 36:39.360]  And as a result, they have a hard time capturing their regulators.
[36:39.900 --> 36:44.540]  But when a sector dwindles to a handful of dominant firms, it's not a rabble anymore.
[36:44.540 --> 36:45.500]  It's a cartel.
[36:45.500 --> 36:55.500]  Five companies or four or three or two or just one company that dominates its sector, they find it very easy to converge on just one thing to tell the regulator.
[36:55.900 --> 37:05.440]  And because they're not having their profits eroded by wasteful competition, they've got an awful lot of money to mobilize that consensus and turn it into law.
[37:06.340 --> 37:12.660]  Tech's regulatory capture allows it to flout the rules that constrain less concentrated sectors.
[37:13.840 --> 37:22.900]  Tech uses its regulatory capture to pretend that when it violates our labor, consumer, and privacy laws that everything is fine because they did it with an app.
[37:23.680 --> 37:25.240]  That's why competition matters.
[37:25.540 --> 37:30.680]  It's not just because competition makes companies work harder and share values with their customers and their workers.
[37:30.960 --> 37:38.940]  It's because it keeps companies from being too big to fail, which means that they'll never become too big to jail, which means that they can't become too big to care.
[37:39.840 --> 37:43.640]  There are plenty of things that we don't want to use competition to fix.
[37:43.840 --> 37:48.940]  Plenty of things that we don't want improved, that we want extinguished, like the commercial surveillance sector.
[37:49.140 --> 38:01.060]  It would be perverse to say that the answer to our commercial surveillance epidemic is to pit those commercial surveillance companies against one another to see which one can violate our human rights most efficiently.
[38:01.540 --> 38:04.280]  We don't want them to get better at violating our human rights.
[38:04.360 --> 38:07.160]  We want them to stop violating our human rights.
[38:07.380 --> 38:11.480]  When companies harm us, we don't want them to get better at the things that they're doing that are bad.
[38:11.780 --> 38:15.280]  We want to extinguish the things that they're doing that are harming people.
[38:15.900 --> 38:20.900]  But the more concentrated a sector is, the harder it is for regulators to hold it to account.
[38:21.440 --> 38:28.700]  Instead, the companies that dominate the sector also dominate their regulators who get to ignore the harms that they inflict.
[38:28.700 --> 38:35.440]  So you have Google and Facebook acting as though they're immune to privacy law because they violate our privacy with an app.
[38:35.640 --> 38:41.700]  Or Uber pretending that it can violate our labor laws because it violates our labor laws with an app.
[38:41.840 --> 38:47.880]  Or Amazon, which violates consumer protection laws but says it doesn't matter because they do it with an app.
[38:48.500 --> 38:53.720]  You will often hear when people talk about the ad tech sector that it's not really a duopoly of Google and Facebook.
[38:53.720 --> 38:59.600]  They say, no, no, no, it's a triopoly because Amazon is a $38 billion advertising business.
[38:59.820 --> 39:06.080]  But Amazon's $38 billion advertising business is not like the advertising businesses that Meta and Google run.
[39:07.220 --> 39:10.360]  Amazon's $38 billion advertising business is payola.
[39:10.920 --> 39:13.880]  It's pay for search placement and search results.
[39:14.100 --> 39:19.400]  The top results in an Amazon search are not the results that match your query best.
[39:19.600 --> 39:21.600]  They're the results that paid the most to be there.
[39:21.600 --> 39:31.680]  If you search for common products on Amazon, the first search result on the top of the page, on average, costs 29% more than the best match for your query.
[39:31.960 --> 39:37.380]  That top row costs 25% more, on average, than the best result for your query.
[39:37.540 --> 39:43.140]  The best result for your query on Amazon is below the fold, 17 places down.
[39:43.140 --> 39:55.380]  Now, if you walked into Best Buy and said, give me some Duracell batteries, and instead they gave you a house brand that cost more and worked worse, the New York Attorney General would go in and smack them around for you.
[39:55.480 --> 39:59.540]  But when Amazon does it, they just call it an ad, and they get away with it.
[39:59.940 --> 40:03.060]  It doesn't count because they did it with an app.
[40:03.780 --> 40:07.900]  Now, this is where that third constraint, self-help, would sure come in handy.
[40:07.900 --> 40:13.200]  After all, if you don't want your privacy violated, you don't have to wait for a privacy regulator to step in.
[40:13.360 --> 40:15.120]  You can just install an ad blocker.
[40:15.480 --> 40:18.900]  More than half of all web users have installed ad blockers.
[40:18.940 --> 40:32.700]  The next time someone tells you users will never install third-party software, tell them that the largest consumer boycott in world history is underway, with the majority of global web users installing aftermarket ad blockers.
[40:33.280 --> 40:35.000]  But the web is an open platform.
[40:35.100 --> 40:38.660]  It's a more noble weapon from a more civilized age.
[40:38.820 --> 40:49.960]  It dates back to a time when standards were developed by large groups of small and medium-sized enterprises that couldn't agree on a single regulatory policy, and as a result held each other in check.
[40:50.500 --> 40:53.280]  And the web is being devoured by apps.
[40:53.680 --> 40:57.700]  Apps are the product of a duopoly, and they are ripe for enshittification.
[40:57.700 --> 41:00.340]  They were born enshittification ready.
[41:00.580 --> 41:02.340]  It was in the requirements document.
[41:02.940 --> 41:08.260]  Because regulatory capture, it's not just the ability to flout regulation.
[41:08.750 --> 41:13.140]  It's also the ability to mobilize regulation against your competitors.
[41:14.000 --> 41:18.120]  Today's tech giants, they got big by exploiting self-help measures.
[41:18.380 --> 41:32.700]  You know, when Facebook told MySpace users that they needed to escape Rupert Murdoch's evil Australian empire, they didn't end the pitch with, we have a better privacy policy, come hang out on Facebook and re-read it until your friends wise up and leave MySpace.
[41:33.380 --> 41:34.540]  I gave them a bot.
[41:35.020 --> 41:48.620]  And if you gave that bot your login and your password, it would go to MySpace several times a day, authenticate as you, scrape all the waiting messages in your inbox, put them into your Facebook inbox, let you reply to them and autopilot them back out again.
[41:48.620 --> 41:55.220]  So you didn't have to choose between Facebook and its better privacy and MySpace where your friends were.
[41:57.060 --> 41:58.980]  But every pirate wants to be an admiral.
[41:59.180 --> 42:06.300]  When Facebook, Apple, and Google did this adversarial interoperability, this reverse engineering, this hacking, they call that progress.
[42:06.620 --> 42:09.640]  When we do it to them, they call it piracy.
[42:10.500 --> 42:15.900]  If you try to make an alternative client for Facebook, they'll say you violated the Digital Millennium Copyright Act.
[42:15.900 --> 42:24.360]  Try to make an Android program that runs iPhone apps and play back the data from the media stores that Apple operates in iOS, Apple will nuke you till you glow.
[42:24.800 --> 42:30.280]  And God help you if you scrape Google the way that Google scraped all of us.
[42:31.000 --> 42:34.580]  The regulatory capture of tech cannot be overstated.
[42:35.080 --> 42:44.400]  A couple of times now I've mentioned the Digital Millennium Copyright Act, 1998, a sort of giant gnarly hairball of a law that Bill Clinton signed.
[42:44.400 --> 42:52.460]  And 2600 readers will know that 2600 plays an important role in the history of the DMCA, a case called Corley in 2001.
[42:53.940 --> 43:04.060]  The Section 1201 of the DMCA is a blanket prohibition on reverse engineering and modifying any access control for a copyrighted work.
[43:04.360 --> 43:13.620]  Now, when that law came out, it was about stopping you from modifying your Sega so it could play homebrew games or altering your DVD player so it would play discs from out of region.
[43:13.860 --> 43:16.200]  But the penalties were very stiff.
[43:16.500 --> 43:20.620]  A five-year prison sentence and a $500,000 fine for a first offense.
[43:21.320 --> 43:41.340]  In the years since, the combination of the proliferation of computerization into every corner of our lives combined with the monotonic expansion of that law by tech companies themselves has meant that we are now living in a world in which the DMCA can be used to shut down almost anything you want to do.
[43:42.480 --> 43:44.900]  So, let's take audiobooks as an example.
[43:45.140 --> 43:51.040]  This is an example near to my heart, maybe not to yours, although lots of people listen to audiobooks, but I'm an author.
[43:51.780 --> 43:55.080]  And in the publishing world, audiobooks are really important now.
[43:55.180 --> 43:58.000]  They sell as many units as hardcovers or paperbacks.
[43:58.140 --> 44:00.180]  So, like, they're the third pillar of the publishing industry.
[44:00.420 --> 44:10.040]  There's a big difference, which is that 90% of the audiobooks sold go through one store, which is Audible, a division of Amazon acquired through an anti-competitive acquisition in 2008.
[44:10.680 --> 44:23.880]  Now, Audible has an ironclad requirement, no exceptions made, that any title that they carry in their store has to be locked with Amazon's digital rights management, has to be encrypted so that it only plays back on Amazon's apps.
[44:24.340 --> 44:34.580]  Now, that means that if I decide I want to leave Amazon, take my customers with me, and go find another store that will pay me a bigger share, or maybe one that's a little more forthright.
[44:34.760 --> 44:47.820]  After all, Amazon was implicated in the 2022 scandal, Audiblegate, in which they were discovered to have stolen at least $100 million from independent audiobook authors with a sleazy accounting trick.
[44:48.020 --> 44:49.660]  So maybe I want to leave Amazon.
[44:50.060 --> 45:06.840]  And in order to take you with me as my customers, I give you a tool that lets you unlock the audiobook that Amazon sold you, that I made, maybe an audiobook that not only did I write, but went into the studio and recorded, and as I sometimes do, even paid the producer and the director and the voice
[45:06.840 --> 45:12.020]  actor myself rather than doing it through a publisher because I've been lucky enough to retain my audiobook rights.
[45:12.060 --> 45:25.280]  If I give you the tool to take that book that I wrote, that I read, that I paid for, so that you can follow me to a rival platform, I commit a felony punishable by a five-year prison sentence and a $500,000 fine.
[45:25.280 --> 45:29.980]  That is a harsher penalty than you would pay if you torrented the audiobook.
[45:30.460 --> 45:35.880]  It's also a harsher penalty than you would likely pay if you shoplifted the audiobook on CD from a truck stop.
[45:36.460 --> 45:43.020]  It may be a lighter penalty, a harsher penalty than you would pay if you stuck up the truck that delivered the CDs.
[45:44.640 --> 45:46.800]  This is regulatory capture.
[45:48.120 --> 45:54.780]  So think again about that product meeting where we're planning our website and what the advertising strategy is going to be like.
[45:54.780 --> 45:59.580]  So Bob has heard us all and he's not going to make the advertising on the website 20% more obnoxious.
[45:59.720 --> 46:02.300]  He's moved on to the next item which is the advertising in the app.
[46:02.780 --> 46:07.480]  And he says, all right folks, I've had the same back of the napkin calculations here.
[46:07.600 --> 46:18.260]  If we make the ads in the app 20% more obnoxious, we will also get a 2% KPI increase and we can all take that vacation in Switzerland.
[46:18.600 --> 46:23.240]  And he looked pointedly at the person who raised their hand last time and said, what about the ad blockers?
[46:23.660 --> 46:28.380]  Everyone looks at that person and he says, Bob, why are you stopping at 20%?
[46:29.000 --> 46:40.140]  Why don't we make those ads 100% more obnoxious and get a 10% increase because it doesn't matter if the user goes to the web, to the search engine and types, how do I block ads in the app?
[46:40.340 --> 46:44.880]  Because the answer is you can't because reverse engineering the app is a felony.
[46:45.840 --> 46:57.540]  And so apps are really just like web pages skinned in enough IP to make it a felony to mod them so that they respect you instead of the company that made them so that they serve your priorities instead of someone else's.
[46:58.500 --> 47:06.780]  And as a result, companies all over are trying to force you to install their apps instead of using their websites.
[47:07.780 --> 47:24.860]  Now, by the same token, there's no reason that gig workers who are being algorithmically twiddled by Uber and its algorithmic wage discrimination couldn't reply to that with their own counter app that allowed drivers to coordinate geographically within a region to reject all rides below a certain
[47:24.860 --> 47:39.680]  threshold and force that pricing algorithm to keep the wages high for all of them except that the toolsmith that built that would be charged with violating the Digital Millennium Copyright Act, the Computer Fraud of an Abuse Act, they'd be accused of tortious interference with contract,
[47:40.000 --> 47:43.900]  maybe a little copyright trademark and patent thrown in at the same time.
[47:44.060 --> 47:46.860]  In other words, they'd be said to have violated IP law.
[47:47.020 --> 47:56.160]  In this case, IP law is just any rule that lets a company reach outside its walls and exert control over the conduct of its competitors, its critics, and its customers.
[47:57.040 --> 47:58.020]  They don't care.
[47:58.500 --> 47:59.580]  They don't have to.
[47:59.880 --> 48:01.480]  They're the mobile phone duopoly.
[48:02.860 --> 48:07.440]  So, self-help is no longer a constraint that we can rely on because they criminalized it.
[48:07.620 --> 48:08.700]  But what about that fourth constraint?
[48:08.820 --> 48:10.040]  What happened to workers?
[48:10.560 --> 48:16.980]  For decades, tech workers' high degree of bargaining power and vocational awe put a ceiling on and shitification.
[48:17.440 --> 48:32.000]  Even after the tech sector shrank to a handful of giants, even after they captured their regulators so they could violate our consumer privacy and labor rights and say it didn't matter because they did it with an app, even after they created this felony contempt of business model business and
[48:32.000 --> 48:42.740]  extinguished self-help for tech users and the toolsmiths who served them, those companies were still constrained by their own workers and their sense of moral injury in the face of the insidificatory imperative.
[48:43.240 --> 48:45.540]  Those workers were the last line of defense.
[48:45.940 --> 48:47.540]  But that dream is over.
[48:48.200 --> 48:53.940]  All that's left these days for tech workers is work for a tech giant until they fire your ass.
[48:54.220 --> 49:05.340]  Like those 12,000 Googlers they fired last year, just months after an $80 billion stock buyback that would have paid their wages for the next 27 years.
[49:05.920 --> 49:08.480]  They fired 260,000 tech workers in the U.S. last year.
[49:09.020 --> 49:11.200]  We're up to 100,000 this year and we're only halfway through.
[49:11.780 --> 49:15.500]  Workers are no longer a check on their boss's worst impulses.
[49:15.860 --> 49:22.640]  The response today to no boss, I refuse to make that product worse, is, yeah, don't forget to turn in your badge on the way out.
[49:22.760 --> 49:25.280]  It's one of those cool smart badges and they're expensive to replace.
[49:26.420 --> 49:41.440]  In other words, the same executives with the same impulses, tried the same tactics to extract value from their platforms, making things worse for their business customers and their end users, but because they no longer face the constraint of competition,
[49:41.440 --> 49:51.680]  of regulation, of self-help, and of their labor, they can, when they apply the same amount of force to the enshitification lever behind their desk every day, it moves a lot further.
[49:53.300 --> 49:55.660]  Okay, so, this is a little depressing.
[49:56.300 --> 49:57.560]  Maybe it's really depressing.
[49:57.560 --> 49:58.860]  But hear me out.
[49:59.020 --> 50:00.940]  We have identified the disease.
[50:01.020 --> 50:03.080]  We have traced its natural history.
[50:03.240 --> 50:10.420]  We've learned how to spot its symptoms, the locking in of users and then of business customers and then the transfer of value from both to the platform.
[50:10.620 --> 50:12.560]  We understand its underlying mechanism.
[50:12.560 --> 50:20.300]  We understand that digital is different, that these fast platforms allow crude tricks to happen so quickly that they fool you.
[50:21.380 --> 50:38.320]  We've unraveled its epidemiology, the collapse of those constraints that once prevented enshitification from taking root in so many firms and that killed off the firms that it did take over rather than letting them shamble on, tormenting their users and their business customers long after they should
[50:38.320 --> 50:39.300]  have been put in the ground.
[50:39.920 --> 50:42.840]  And now that we understand the disease, we can work on a cure.
[50:43.600 --> 50:46.720]  Remember, four constraints prevent enshitification.
[50:47.180 --> 50:50.680]  Competition, regulation, interoperability, and labor.
[50:51.420 --> 50:58.080]  To reverse enshitification and to guard against its re-emergence, we have to restore and strengthen each of these.
[50:58.560 --> 51:02.080]  Now, on the first one, competition, it's actually weirdly good.
[51:02.460 --> 51:16.900]  I mean, notwithstanding the latest Supreme Court stuff, which is not the last word on things because this is a global phenomenon, we have seen more action on competition and antitrust enforcement in the last four years than we have in the last 40.
[51:17.240 --> 51:19.040]  And it's not just in the U.S. in the EU, in the UK, in Australia, and in Japan and China, they're doing more in competition than they have done in two generations.
[51:26.860 --> 51:28.140]  They're blocking mergers.
[51:28.300 --> 51:29.800]  They're unwinding existing mergers.
[51:29.900 --> 51:34.360]  They're banning predatory pricing and really smacking around the companies that do it.
[51:34.560 --> 51:36.140]  They already have the laws to do it.
[51:36.200 --> 51:37.500]  They just hadn't been enforcing them.
[51:37.620 --> 51:40.640]  They don't need to wait for their legislatures to pass new laws.
[51:41.080 --> 51:44.180]  Now, the enshitifiers are not taking this lying down.
[51:44.340 --> 51:49.120]  The business press can't stop talking about how stupid and old-fashioned all this stuff is.
[51:49.280 --> 51:52.260]  They call people like me hipster antitrust.
[51:52.440 --> 51:59.340]  But they reserve their worst opprobium for any regulator that actually shows up for work and does their job.
[51:59.840 --> 52:12.480]  Chief in their monster manual is Lena Kahn, the incredible chair of the Federal Trade Commission, youngest chair in its history, who has done more in her term than all of her predecessors since I was in short pants.
[52:12.780 --> 52:22.420]  The Wall Street Journal has run more than a hundred op-eds saying that Lena Kahn can't get anything done and that she's wasting our tax dollars.
[52:23.000 --> 52:30.380]  Sure, Rupert Murdoch pays the editorial board of the Wall Street Journal to write a hundred editorials about someone who's not getting anything done.
[52:36.440 --> 52:41.860]  Competition has a long way to go, but all over the world, competition law is seeing a massive revitalization.
[52:42.220 --> 52:47.860]  Reagan and Thatcher might have put it in a coma in the 1980s, but it's awake, it's back, and it's really pissed.
[52:48.920 --> 52:50.100]  So how about regulation?
[52:50.720 --> 52:56.900]  How will we get tech companies to stop doing the one weird trick of adding with an app to their crimes and escaping enforcement?
[52:57.500 --> 53:00.180]  In the European Union, they're starting to figure it out.
[53:00.420 --> 53:06.720]  There are these two new laws that came into effect this year, the European Digital Markets Act and Digital Services Act.
[53:06.960 --> 53:08.320]  They're kind of mixed bags.
[53:08.500 --> 53:14.560]  There's parts of them that I think are great and parts that I think are really misguided, but what's very interesting about them is their enforcement mechanism.
[53:15.240 --> 53:22.280]  So historically, the European Commission, when they wanted to regulate tech, they made regulations, which then had to be implemented in each EU country as a law.
[53:22.660 --> 53:27.920]  And that meant that if a company broke the law, you would have to sue them in the country where they were headquartered.
[53:28.180 --> 53:32.640]  Well, all the tech companies are headquartered in Ireland because Ireland is a crime haven.
[53:33.020 --> 53:35.560]  It didn't start out as a crime haven, it started off as a tax haven.
[53:35.880 --> 53:48.280]  But if you're going to offer preferential tax treatment to companies that are so nimble and footloose that they can pretend to be Irish, then you have to possibility that next week they'll pretend to be Cypriot or Maltese or Luxembourgian or Dutch on a bad day,
[53:48.560 --> 53:54.640]  which means that you've got to really pamper them, lest they take their ass out the door and give their money to someone else.
[53:54.920 --> 54:00.120]  And so as a result, Ireland is a country where the privacy commissioner basically never gets out of bed.
[54:00.280 --> 54:12.020]  And when he does, he doesn't put his pants on, he just sits around eating cereal and watching breakfast cartoons, which is why Ireland, where you have to go to enforce Europe's privacy laws if the big tech companies, which are all pretending to be Irish, has violated you.
[54:13.500 --> 54:20.320]  That's why they hear about 17 major cases a year, while in Germany it's 500 major cases a year.
[54:20.720 --> 54:24.380]  So with the Digital Markets Act and the Digital Services Act, they just federalized it.
[54:24.680 --> 54:38.060]  They said these are not regulations, they're acts, they're pan-European laws, they don't have to be implemented in the European member states, they're implemented Europe-wide by the commission, by the parliament, and all of the enforcement starts in the federal court.
[54:38.460 --> 54:40.860]  And so this is going to be some pretty amazing stuff.
[54:40.860 --> 54:52.420]  Now, one of the provisions of the Digital Markets Act is mandatory interoperability, setting standards for how different services talk to each other so you can port from one service to another and maintain interop between them.
[54:53.760 --> 54:59.600]  Now, that seems like a reasonably good idea to me, although I can think of a million ways that it can go wrong.
[55:00.280 --> 55:07.880]  I could see a complementary approach in the United States that we're actually starting to see little bits and pieces of around the edges with right-to-repair laws.
[55:08.040 --> 55:14.900]  So in Oregon, they just passed a right-to-repair law where they just said, okay, we can't repeal the DMCA because that's a federal law.
[55:14.980 --> 55:17.560]  We're just going to make it illegal to use DRM on your parts.
[55:17.760 --> 55:20.640]  You can't sell a thing in Oregon if there's DRM on your parts.
[55:20.920 --> 55:34.920]  And so if a technician swaps a part in and the part tries to do a handshake with the central unit of the device and the technician can't complete that handshake because there's DRM, then it's illegal to sell that device in Oregon.
[55:35.140 --> 55:36.900]  So this is a very American approach.
[55:37.020 --> 55:38.280]  It's a kind of YOLO approach.
[55:38.520 --> 55:48.080]  Instead of saying, okay, we're going to tell you how you have to design your products, instead of what they say is, we're just going to withdraw the state's protection from your products.
[55:49.560 --> 55:51.440]  America is doing much better on privacy.
[55:51.860 --> 55:54.500]  Europe, we may see privacy enforcement with the DMA.
[55:54.920 --> 55:58.120]  America, we haven't had any real privacy enforcement for quite some time.
[55:58.320 --> 56:08.840]  To give you an idea of how bad privacy is here in the United States, the last United States federal consumer privacy law to be enacted was something called the Video Privacy Protection Act.
[56:09.100 --> 56:17.840]  It was enacted in 1988, and it prohibits video store clerks from telling newspapers what VHS cassettes you took home.
[56:18.620 --> 56:21.500]  This is like a trifecta of dead media.
[56:22.260 --> 56:26.160]  It's been a minute since America got a new federal consumer privacy law.
[56:26.280 --> 56:34.640]  And the thing is, there are a lot of people in America who are really pissed off about things that have some connection to America's failure to fix its privacy laws.
[56:34.840 --> 56:38.120]  Are you worried that Facebook turned rampy into a QAnon?
[56:38.300 --> 56:40.660]  Or that Insta made your teen anorexic?
[56:40.780 --> 56:45.500]  Or that TikTok brainwashed a millennial in your life into quoting Osama Bin Laden?
[56:45.800 --> 56:50.980]  Are you worried that the cops rolled up the identities of every protester in a Black Lives Matter protest?
[56:50.980 --> 56:53.680]  Or all the people on January 6th?
[56:53.760 --> 56:59.160]  Or that red state attorneys general follow teenagers across state lines and into abortion clinics?
[56:59.520 --> 57:06.660]  Or that black people get discriminated against when they apply for loans or mortgages or jobs because of the surveillance data that's been acquired on them?
[57:06.820 --> 57:09.820]  Or, you know, someone's making a deepfake AI porn of you?
[57:10.780 --> 57:26.280]  Having a federal privacy law with a private right of actions and that would allow consumers, individuals, to sue even if the state attorney general didn't want to take their case would go a long way to addressing every single one of those problems, which means that the coalition for a privacy law in
[57:26.280 --> 57:29.440]  America is larger and more powerful than it's ever been.
[57:29.580 --> 57:36.480]  And we are seeing more movement on real privacy laws than we've seen in my lifetime as an activist.
[57:37.520 --> 57:38.780]  But what about self-help?
[57:38.920 --> 57:43.240]  Well, like I say, the European Union is going to mandate some interop.
[57:43.540 --> 57:47.440]  America is sneaking in a little bit of adversarial interop through the back door.
[57:48.280 --> 57:50.480]  but it's pretty weak sauce, I have to say.
[57:50.720 --> 58:02.160]  And that's a pity because that's the fight that I started in on and that really 2600 has the biggest nexus with thanks to its publication of DCSS and its lawsuit over the DMCA.
[58:02.680 --> 58:03.680]  It's lagging.
[58:04.680 --> 58:11.800]  But there's an idea out of finance, something called Stein's Law, that holds that anything that can't go on forever eventually stops.
[58:12.420 --> 58:20.220]  And the idea that manufacturers get to tell you how you have to configure and use the products that you rely on is grossly offensive.
[58:20.380 --> 58:40.000]  And it's a moral hazard that invites companies to dream up the most evil, ridiculous bullshit that makes people frankly outraged, whether that's Mercedes renting you the accelerator pedal in your car by the month, or HP staffing an entire building full of engineers who do nothing day and night but
[58:40.000 --> 58:46.840]  figure out new ways to make you spend $10,000 a gallon on ink to print out your boarding cards to-do lists, and kids coloring book masters.
[58:48.080 --> 58:49.440]  It is offensive.
[58:50.100 --> 59:06.880]  And the only people who agree are the people who are running those monopolies in all those other industries, like medtech monopolists who are locking in insulin pump users to their own glucose monitors and eventually to their own brand of insulin, turning people with diabetes into walking inkjet
[59:06.880 --> 59:07.300]  printers.
[59:08.680 --> 59:09.980]  Now finally, there's labor.
[59:09.980 --> 59:14.000]  We have seen a massive surge in tech union activity.
[59:14.240 --> 59:28.360]  Tech workers are realizing that they are not founders in waiting, that the days of free massages and facial piercings and getting to wear those cool black t-shirts that say things your boss don't understand, that they weren't the result of your boss liking you.
[59:28.520 --> 59:33.220]  They were the result of your boss being afraid that they couldn't fire you and replace you.
[59:33.420 --> 59:36.240]  And now that they can, those days are ending.
[59:37.120 --> 59:46.800]  In Seattle, tech workers at Amazon walked out with the warehouse workers at Amazon because they understood that they are all workers, not temporarily embarrassed founders.
[59:51.820 --> 01:00:03.240]  The only reason that the tech workers at Amazon aren't being monitored by that AI that rats them out to their boss if they go to the toilet during work hours is that rapidly dwindling market power.
[01:00:03.380 --> 01:00:10.260]  And the way things are going, it's only a matter of time until Amazon's programmers also have to keep a bottle to piss in next to their computers.
[01:00:10.700 --> 01:00:16.320]  You know, for a guy who built a penis-shaped rocket, Jeff Bezos really hates our kidneys.
[01:00:18.120 --> 01:00:24.440]  So we're seeing some bold, muscular action on competition and regulation and labor, though self-help could use a bit of a goose.
[01:00:24.540 --> 01:00:29.960]  And it's not a moment too soon because the bad news is, and shittification is coming to every industry.
[01:00:30.220 --> 01:00:36.360]  If it's got a networked computer in it, the people who make it can run the first lesson of the Darth Vader MBA.
[01:00:36.780 --> 01:00:38.360]  I have altered the bargain.
[01:00:38.720 --> 01:00:40.280]  Pray I don't alter it further.
[01:00:41.160 --> 01:00:43.560]  They can change the rules from moment to moment.
[01:00:43.680 --> 01:00:47.060]  They can violate your rights and say, it's okay, we did it with an app.
[01:00:48.120 --> 01:00:50.700]  We were wrong to think the software would eat the world.
[01:00:50.880 --> 01:00:52.880]  What software did was it had shittified it.
[01:00:53.000 --> 01:00:55.700]  We were concerned with the wrong end of the elementary canal.
[01:00:57.180 --> 01:01:04.860]  Just as with privacy law in the United States, the potential coalition against shittification is massive and unstoppable if we can gel it.
[01:01:04.860 --> 01:01:08.780]  Now, the cynics among you may be skeptical that any of this will make a difference.
[01:01:08.980 --> 01:01:13.380]  After all, isn't enshittification just another funny word for capitalism?
[01:01:14.240 --> 01:01:15.040]  It's not.
[01:01:15.540 --> 01:01:17.720]  Look, I'm not going to cape for capitalism here.
[01:01:17.800 --> 01:01:24.200]  I am not a true believer in markets as the most efficient allocators of resources or arbiters of policy.
[01:01:24.440 --> 01:01:30.780]  I think if there is ever any doubt of that, capitalism's total failure to grapple with a climate emergency has erased it.
[01:01:31.280 --> 01:01:42.820]  But 20 years ago, capitalism made space for a wild and woolly Internet, a space where people with disfavored views could find one another, offer mutual aid, organize, and gather conferences like this.
[01:01:43.260 --> 01:01:44.900]  And the capitalism of today?
[01:01:45.300 --> 01:01:55.120]  It produced a global digital ghost mall filled with bot shit, crap gadgets from companies whose names are all consonants, and cryptocurrency scams.
[01:01:56.220 --> 01:01:59.800]  The Internet is not more important than the climate emergency.
[01:02:00.160 --> 01:02:06.040]  It's not more important than gender justice, or racial justice, or inequality, and it's certainly not more important than genocide.
[01:02:06.820 --> 01:02:10.040]  But the Internet is the terrain on which we will fight those fights.
[01:02:10.180 --> 01:02:14.060]  Without the free, fair, and open Internet, we lose those fights before we join them.
[01:02:14.200 --> 01:02:17.400]  And we can reverse the enshittification of the Internet.
[01:02:17.600 --> 01:02:21.220]  We can halt the creeping enshittification of every digital device.
[01:02:21.220 --> 01:02:32.580]  We can build a better enshittification-resistant digital nervous system that is fit to coordinate the mass movements we will need to fight fascism and genocide, save our planet, and our species.
[01:02:33.280 --> 01:02:41.860]  There is nothing about this revolutionary struggle that will be improved by making the lives of marginalized people worse because their technology isn't shittified.
[01:02:41.960 --> 01:02:47.100]  We have everything to gain by making the lives of the people that we fight alongside of better.
[01:02:47.940 --> 01:02:57.080]  Martin Luther King said, it's true that the law can't make a man love me, but it can stop him from lynching me, and I think that's pretty important.
[01:02:57.080 --> 01:03:15.400]  And it's true, the law can't force corporate sociopaths to conceive of you as a human being entitled to dignity and fair treatment, and not just as an ambulatory wallet, a source of nutrition, and inconvenient gut flora for the transhuman colony organism we call the Limited Liability Corporation.
[01:03:16.400 --> 01:03:26.320]  But it can make that company act as though they believe that you are entitled to dignity and fair treatment, even though they don't think you deserve it.
[01:03:27.040 --> 01:03:28.720]  And I think that's pretty important.
[01:03:29.420 --> 01:03:30.100]  Thank you.
[01:03:34.220 --> 01:03:34.820]  Thank you.
[01:03:39.920 --> 01:03:40.560]  Thank you.
[01:03:42.120 --> 01:03:43.240]  It's very kind of you.
[01:03:43.360 --> 01:03:43.900]  Thank you very much.
[01:03:44.120 --> 01:03:46.400]  So, we have some time for questions.
[01:03:48.140 --> 01:03:51.580]  Yeah, we have, we have like 30, 40 minutes for questions.
[01:03:53.020 --> 01:03:56.640]  And there is a mic down here so that you can be on the live stream.
[01:03:57.340 --> 01:04:04.580]  While you folks are gathering at the mic, let me say that a good question has one part and not two.
[01:04:05.220 --> 01:04:07.940]  It's rarely more of a comment than a question.
[01:04:08.100 --> 01:04:09.660]  It's usually an actual question.
[01:04:10.120 --> 01:04:10.620]  All right.
[01:04:10.780 --> 01:04:11.320]  Go ahead.
[01:04:12.400 --> 01:04:13.760]  Thank you for your talk.
[01:04:13.880 --> 01:04:14.680]  I was curious.
[01:04:14.780 --> 01:04:20.360]  We have examples of other attempts for legislation to require interoperability.
[01:04:20.560 --> 01:04:25.900]  For example, electronic health records, and that seems to have been a after shit.
[01:04:27.260 --> 01:04:32.260]  So, do you have any ways forward with that or broad principles?
[01:04:32.500 --> 01:04:32.680]  Yeah.
[01:04:33.000 --> 01:04:46.820]  So, the question is about electronic health records, EHRs, which are these gadgets that are these software programs that have your health records in them that different doctors, medical practices, hospital groups use to keep records on you.
[01:04:46.920 --> 01:04:50.640]  And they silo your records that can make it really hard for you to move from one doctor to another.
[01:04:50.780 --> 01:04:56.160]  And they also give an enormous amount of market power to these companies, which can then screw the medical practices.
[01:04:56.400 --> 01:05:00.440]  And we mandated some interop among them, and it didn't work very well.
[01:05:00.560 --> 01:05:03.140]  It's been really gross and bad.
[01:05:03.340 --> 01:05:08.000]  So, there's a couple of different approaches to making good standards.
[01:05:08.280 --> 01:05:09.740]  So, really, that's what this is about.
[01:05:09.880 --> 01:05:10.980]  It's how do you make a good standard.
[01:05:11.180 --> 01:05:18.720]  And I'm sure some of you have done some standards work, and all of you have worked with standards and cursed the people who do standards work.
[01:05:19.120 --> 01:05:22.160]  I have done my turn in the standards salt minds as well.
[01:05:23.100 --> 01:05:33.520]  There's a kind of contradiction at the heart of standardization, which is that standards bodies tend to be organized as non-profits, and the people who do the work in them tend to be volunteers.
[01:05:34.820 --> 01:05:44.800]  And that sounds like a kind of gnarly free software project where people who are really into the community get together and try to find the best way to do it.
[01:05:44.900 --> 01:05:51.560]  In practice, what it means is that large companies pay their employees to volunteer to fill out these committees.
[01:05:52.480 --> 01:06:07.460]  And when there are a lot of large companies that dominate a standards body, right, when there's like five large companies that the standards body is dependent on in order to pay its bills, then an open standardization process becomes those five companies deciding how to make a meal out of the rest
[01:06:07.460 --> 01:06:07.840]  of us.
[01:06:07.980 --> 01:06:24.100]  Even if an academic or people from non-profits or from small startups are sitting there on that standards committee, they're just like one person doing it in their spare time, whereas the secretary, the president, the chair, the co-chair, and everyone else who's like on it full-time is working for
[01:06:24.100 --> 01:06:24.820]  one of these large firms.
[01:06:24.920 --> 01:06:29.840]  And you tend to get these extremely parochial standards out of them that reflect the priorities of vendors.
[01:06:30.100 --> 01:06:34.140]  So one of the first projects I ever did at EFF was after Bush v. Gore.
[01:06:34.720 --> 01:06:35.540]  You remember Bush v. Gore?
[01:06:38.260 --> 01:06:41.420]  The stolen election in 2000, the actual stolen election?
[01:06:43.800 --> 01:06:45.960]  You'll recall that those voting machines sucked.
[01:06:46.380 --> 01:06:51.780]  And so Congress ordered voting machine companies to find a standard and adhere to it.
[01:06:52.140 --> 01:06:56.740]  And the IEEE very unwisely said, oh, we'll oversee that.
[01:06:56.880 --> 01:06:58.880]  We'll be the standards body where you can convene.
[01:06:59.000 --> 01:07:06.640]  And so the committee to make the standards for the next generation of voting machines was just composed of the same people who made the shitty machines that caused Bush v. Gore.
[01:07:07.460 --> 01:07:09.720]  And they said, you know, we don't have a lot of time here.
[01:07:09.920 --> 01:07:10.920]  There's an election coming up.
[01:07:10.920 --> 01:07:12.840]  We want to make sure that the standard is in place.
[01:07:12.940 --> 01:07:20.480]  So we're going to make a descriptive standard, which is to say we are going to describe the machines that we already use and call them the standard.
[01:07:21.040 --> 01:07:21.300]  Right?
[01:07:21.540 --> 01:07:24.380]  So there are many ways the standardization can go very wrong.
[01:07:24.480 --> 01:07:29.980]  Now, in that case, we rallied IEEE members who convinced the IEEE to de-charter the committee.
[01:07:30.320 --> 01:07:32.460]  And that standard never emerged.
[01:07:32.640 --> 01:07:36.360]  It's sort of the unglamorous, weird work of wonky tech stuff.
[01:07:37.060 --> 01:07:47.960]  In general, standards work best when they're made by a bunch of small companies of equal size and not when they're dominated by a few large companies, which is to say, if you wanted to get there, I wouldn't start from here, right?
[01:07:48.080 --> 01:07:56.300]  We are now at the point where it's really hard to run a fair process because we've allowed a small number of firms to metastasize to the point where they can corrupt things.
[01:07:56.300 --> 01:08:00.220]  So we need to look for other mechanisms to check the quality of standards.
[01:08:00.520 --> 01:08:07.360]  And I really favor a two-part system for standardization, one of which is we have a standard.
[01:08:07.660 --> 01:08:14.480]  And maybe it's arrived at by industry, it's overseen by NIST or by the European Commission or by a blue-ribbon panel of experts, whatever.
[01:08:14.680 --> 01:08:17.840]  So we make a standard, and this is the standard, and it is described.
[01:08:18.220 --> 01:08:24.480]  And then we have a safe harbor for people who reverse engineer to meet the purposes of the standard.
[01:08:24.740 --> 01:08:39.140]  So we say, you can violate the DMCA, copyright, patent, trademark, all the neighboring rights, and so on, provided that you do so to further the interests of the standard, and then you don't violate labor, privacy, or consumer rights while you do so.
[01:08:39.400 --> 01:08:42.100]  And so now you have two different ways of interoperating with something.
[01:08:42.220 --> 01:08:45.600]  So you can either follow the standard for the EHR and just do the EHR, right?
[01:08:45.660 --> 01:08:52.560]  Or you can hack the EHR and figure out how to export its data or make a middleware layer or do whatever else it is you want to do.
[01:08:52.560 --> 01:08:58.920]  Now, no one who's got a good standard is going to bother with all the reversing and sort of tedious stuff, right?
[01:08:59.820 --> 01:09:04.180]  Preferentially, we will start with the path that has been laid out for us.
[01:09:04.340 --> 01:09:15.260]  So on the one hand, this tells us, right, if everyone who re-implements the standard or makes the interoperable product is reversing the dominant product, then you know that the standard sucks, right?
[01:09:15.260 --> 01:09:17.960]  It's like it's a good reason to go back to the drawing board.
[01:09:18.120 --> 01:09:23.980]  And on the other hand, it means that that you will still have a solution, even though the standard sucks.
[01:09:24.820 --> 01:09:35.580]  I think that this is like a two-part epoxy, where you have something that's quite rigid, right, and very strong, which is a defined standard.
[01:09:35.780 --> 01:09:48.340]  And you have something that's quite gooey, which is reverse-engineered solutions, and scraping and bots and, you know, intermediary layers and whatever, that every time they change their back end, you've got to hastily rewrite your client so that it continues to work.
[01:09:48.900 --> 01:09:53.640]  And that when you put the two of them together, you get something that is strong and flexible.
[01:09:53.980 --> 01:10:04.480]  And I think that's like a, it's a, I'm sorry, this is a very nerdy policy wonk answer, but this is how you get something that I think regulators could actually administer and figure out whether it was working.
[01:10:04.480 --> 01:10:18.460]  And that would overcome the otherwise insurmountable problem of starting with an industry that is already dominated by a handful of firms that would otherwise just sabotage any standardization effort.
[01:10:18.660 --> 01:10:19.280]  Yeah.
[01:10:20.140 --> 01:10:20.860]  Yeah.
[01:10:21.080 --> 01:10:43.600]  So, I would love to get your thoughts on some of the new, quote, unquote, privacy and national security acts that have been proposed in regards to TikTok and DJI, where the government is claiming these are for security and, and, um, for privacy concerns while going directly after other companies for
[01:10:43.600 --> 01:10:46.760]  protectionist reasons under the guise of privacy.
[01:10:47.040 --> 01:10:51.840]  So, I think you've put a logical, uh, exclusive order where maybe you want an and.
[01:10:52.320 --> 01:11:10.120]  Like, I, I think the coalition for banning TikTok, which I, let me say, I don't think we should ban TikTok, but I think the, the political economy that produced the coalition to ban TikTok includes people who in good faith believe that they are doing something for privacy and people who are just
[01:11:10.120 --> 01:11:12.600]  protectionist and want to just make things easier for Facebook.
[01:11:12.920 --> 01:11:15.540]  And that's what, that was the, that's the key ingredient, right?
[01:11:15.660 --> 01:11:21.820]  Is they, they created a coalition, not merely of people who want to do trade war, right?
[01:11:21.940 --> 01:11:23.500]  Or, you know, Sinophobes or whatever.
[01:11:23.860 --> 01:11:27.460]  They, they use together two different groups, right?
[01:11:27.560 --> 01:11:29.980]  This is how things get done is, is through coalition.
[01:11:30.340 --> 01:11:45.880]  And so we also know where the fissure line is in that coalition, which is, um, uh, the privacy people are only getting this tiny pocket of privacy, uh, in a way that is not particularly important because of course, if what you're worried about is, uh,
[01:11:46.100 --> 01:11:52.980]  Chinese state actors acquiring information on Americans, banning TikTok without regulating the data brokerage sector does nothing.
[01:11:53.100 --> 01:12:00.660]  It just changes who they had a credit card to, to get, you know, the asshole to appetite, non-consensual dossier on you and everyone you love, right?
[01:12:00.900 --> 01:12:15.900]  And so, um, it, what we need to do if we're going to break that coalition apart is we need to focus on the fact that the people who are acting in good faith and who care about privacy have, uh, not solved their problems, right?
[01:12:16.020 --> 01:12:18.800]  And we need to show them that there's a path to actually solving the problem.
[01:12:18.900 --> 01:12:23.040]  And that's through an actual privacy law that reigns in all these other firms as well.
[01:12:23.240 --> 01:12:28.940]  I mean, the idea that these firms have a national identity is on the one hand, it's very silly, right?
[01:12:29.060 --> 01:12:38.160]  Like I'm sure Mark Zuckerberg has like 11 passports at this point, you know, so that he's, he's got lots of luxury bunkers he can hide in when they start building, building guillotines on his lawn.
[01:12:38.480 --> 01:12:50.880]  Uh, but you know, at the same time, there's something quite pervasive in the minds of lawmakers and their constituencies where they do really identify these firms with their nations, right?
[01:12:50.920 --> 01:12:58.600]  One of the reasons the European Union is making such muscular rules about the United States is that these are not, or about tech companies rather, is that they're not Nokia.
[01:12:58.880 --> 01:13:00.180]  They're not Olivetti.
[01:13:00.380 --> 01:13:01.520]  They're not Deutsche Telekom, right?
[01:13:01.600 --> 01:13:03.000]  Those companies are already in the ashes.
[01:13:03.320 --> 01:13:09.000]  And so they have like, they feel like they can just go ahead and, and do stuff about them.
[01:13:09.100 --> 01:13:14.800]  Um, partly because like the voters won't get really upset that they're destroying a beloved national treasure.
[01:13:15.040 --> 01:13:29.520]  And I think that lawmakers are, uh, that the, the, the dilemma whose horns they're on is the fear that if they go after Facebook or these other companies, uh, that they will be perceived as doing something hostile to our national interest.
[01:13:29.860 --> 01:13:44.080]  Uh, and so what we need to do is really, again, to break that coalition, we need to heighten the fact that these are transnational entities that are happy to elect dictators in Cambodia and, uh, you know, do all kinds of other shady stuff in other countries and really are, uh, you know, completely, uh, unmoored from any kind of U.S.
[01:13:49.800 --> 01:13:56.380]  domicile, except where it benefits them because the State Department will act on their behalf or because legislators will hold their tongue.
[01:13:56.520 --> 01:13:59.140]  And we need to make sure that Americans know these aren't American companies.
[01:13:59.720 --> 01:14:04.600]  And then American lawmakers won't be as nervous about smacking them around a little.
[01:14:06.140 --> 01:14:06.680]  All right.
[01:14:06.780 --> 01:14:07.080]  Thank you.
[01:14:07.280 --> 01:14:07.620]  Thank you.
[01:14:08.220 --> 01:14:08.780]  Yeah.
[01:14:09.580 --> 01:14:15.040]  Uh, Mike, Mike, I don't exactly have a question more of something I'd like you to elaborate on.
[01:14:15.200 --> 01:14:15.420]  Sure.
[01:14:15.660 --> 01:14:17.620]  But when...
[01:14:21.020 --> 01:14:21.580]  Sorry.
[01:14:21.800 --> 01:14:22.360]  Wait a time.
[01:14:23.520 --> 01:14:37.020]  When you have the surveillance that sells you ads, it also sells you the actual contents of websites like Twitter and Facebook that do, like, turn your grandma into a Q-on 5G tower-burning person.
[01:14:37.020 --> 01:14:43.720]  Uh, how do you... how do you think this... this affects, like, how hard it is to leave the platform?
[01:14:43.940 --> 01:14:45.040]  Not just network effect.
[01:14:45.120 --> 01:14:45.240]  Right.
[01:14:45.360 --> 01:14:47.160]  But changing your neurochemistry.
[01:14:48.400 --> 01:14:48.600]  Right.
[01:14:48.760 --> 01:14:50.560]  Are they... are they, like, hacking our dopamine?
[01:14:50.900 --> 01:14:51.180]  Yeah.
[01:14:51.380 --> 01:14:51.540]  Yeah.
[01:14:51.940 --> 01:14:54.480]  I'm really skeptical of the dopamine-hacking claims.
[01:14:54.580 --> 01:14:56.040]  Well, I... not dopamine.
[01:14:56.300 --> 01:14:58.660]  Right, but, like, the addict... the addictiveness quality.
[01:14:58.660 --> 01:15:11.480]  When I'm... when I'm browsing Mastodon or co-host, this has a lot... it has a lot more content that's actually worth reading, but I can't focus on it long enough because I'm... my brain is so broken by Twitter.
[01:15:12.360 --> 01:15:12.780]  Hmm.
[01:15:12.940 --> 01:15:14.080]  That's an interesting theory.
[01:15:14.180 --> 01:15:28.860]  I think there's too many confounders to decide whether, like, the reason that you find it hard to concentrate is that, uh, Twitter broke your brain, or that the circumstances that are coterminal with the rise of Twitter, which is, like, um, a bunch of stuff happening all at once.
[01:15:29.000 --> 01:15:39.580]  Like, we... you know, there... people complain rightly that, you know, so the 24-hour news cycle created this thing where there's just a bunch of stuff happening, which is true, but at the same time, there is a bunch of stuff happening.
[01:15:39.760 --> 01:15:53.220]  There is an actual bunch of stuff happening, and I find it hard to concentrate, too, and I think it's because there's a lot on my There's a lot of stuff happening that is legitimately worrisome.
[01:15:54.120 --> 01:15:57.140]  And I've experienced both sides of this.
[01:15:57.900 --> 01:16:06.800]  As a reader and a writer, when I fall into a book, and I do manage to lose myself in a book, it's bliss.
[01:16:07.140 --> 01:16:12.840]  Because for a while, the world goes away, and that kind of horrible chattering dies.
[01:16:13.020 --> 01:16:14.980]  And it does help when I don't use a device.
[01:16:15.640 --> 01:16:24.260]  You know, I'm keenly aware in the back of my mind that as I'm reading a book on my phone, that I'm like one tap away from a guy sticking a lemon up his nose on YouTube.
[01:16:24.300 --> 01:16:27.080]  And it's very hard not to do that.
[01:16:27.140 --> 01:16:29.320]  But when I manage it, it's amazing.
[01:16:29.480 --> 01:16:30.820]  And the same is true for writing books.
[01:16:30.900 --> 01:16:32.480]  So I write when I'm anxious.
[01:16:32.520 --> 01:16:33.960]  I write when I'm worried.
[01:16:34.400 --> 01:16:36.680]  And during lockdown, I wrote nine books, right?
[01:16:37.020 --> 01:16:39.680]  And I just forgot that the world was going on.
[01:16:39.680 --> 01:16:50.240]  My wife can't understand how I listened to the same nine Talking Heads albums on a loop for 12 months in the hammock in the backyard, nonstop, without going nuts.
[01:16:50.340 --> 01:16:52.600]  And I'm like, because everything went away, right?
[01:16:52.740 --> 01:16:55.840]  So I think it's very hard to say.
[01:16:55.840 --> 01:17:04.240]  And I think there's a very self-serving story that tech giants tell about how they broke our brains, right?
[01:17:04.240 --> 01:17:06.960]  So this is a thing leave and sell at Virginia Tech.
[01:17:07.180 --> 01:17:08.360]  He's a STS scholar.
[01:17:09.080 --> 01:17:12.440]  He calls it critahype, like criticism and hype put together.
[01:17:13.000 --> 01:17:19.400]  And he says that like sometimes these tech companies make claims about what they're doing that are themselves just marketing claims.
[01:17:19.580 --> 01:17:21.820]  And critics hear them and they take them seriously.
[01:17:22.320 --> 01:17:25.880]  And they repeat them, but with a negative valence, but it still helps the company.
[01:17:25.880 --> 01:17:30.340]  So like Facebook goes around and they say like, oh, we've hacked your dopamine.
[01:17:30.620 --> 01:17:31.480]  We've made it addictive.
[01:17:31.640 --> 01:17:34.340]  We've figured out how to bypass your critical faculties.
[01:17:34.480 --> 01:17:35.760]  We're doing all these things with big data.
[01:17:35.840 --> 01:17:36.760]  We have persuasive tech.
[01:17:36.940 --> 01:17:47.660]  We did this election experiment where we turned out 240,000 more people to vote than we were expected when we put them in the experimental arm of the trial, right?
[01:17:47.660 --> 01:17:56.060]  And then critics go around and they go, Facebook is hacking your dopamine and bypassing your critical faculties and they're addicting you.
[01:17:56.360 --> 01:18:03.460]  And then Facebook goes back on its sales calls and they're like, so you're wondering why you should pay me a 40% premium to advertise on Facebook.
[01:18:03.600 --> 01:18:11.100]  Well, I don't know if you've heard my critics, but they'll tell you that I'm an evil dopamine hacking wizard and that I can bypass your critical faculties and make you addicted and so on.
[01:18:11.100 --> 01:18:23.100]  And it's clearly the case that I use computers more than I would if I had like perfect control over my will and my desires.
[01:18:23.100 --> 01:18:27.000]  And I think that's true of all of us, but like when I interrogate the reasons for that, there are a lot of them.
[01:18:27.640 --> 01:18:34.260]  One is that it is actually just distracting me from what's legitimately a very worrying and difficult time to be alive.
[01:18:35.140 --> 01:18:41.220]  And another is that when it comes to social platforms, it's where all the people I love are.
[01:18:41.400 --> 01:18:55.040]  And they're stuck there because you can't just like number port from Facebook to Mastodon, not because it's difficult to make that work as a matter of reverse engineering, but because they'd like destroy you if you did with the courts.
[01:18:55.440 --> 01:19:05.160]  And so I think we're going to have a hard time figuring out to what extent, which parts, which factors are in play, but we know which ones we can change right away, right?
[01:19:05.180 --> 01:19:07.360]  We know which levers are like at our hands, right?
[01:19:07.360 --> 01:19:23.640]  So like if it turns out that making the world better and also making it easier to leave Facebook so companies run by like foundationally unserious people like Mark Zuckerberg are no longer important forces in our geopolitics, if that turns out not to solve this problem, we'll still be better off.
[01:19:25.460 --> 01:19:27.420]  So I'm like okay with starting there.
[01:19:27.720 --> 01:19:33.880]  You know, I have this friend, Mimi Ito, she's a sociologist who studies how young people use networks.
[01:19:33.880 --> 01:19:34.900]  She's at UC Irvine.
[01:19:35.340 --> 01:19:40.620]  And I once brought her in for a talk and someone in the audience said, how are iPads affecting our kids' brains?
[01:19:40.640 --> 01:19:41.600]  This was like 2013.
[01:19:42.320 --> 01:19:46.180]  And she said, well, look, you can't think without your brain being affected.
[01:19:46.360 --> 01:19:48.620]  So clearly it's affecting our kids' brains in some way.
[01:19:49.060 --> 01:19:56.640]  We will know in 40 years when we go and we look at like cohort studies to figure out how kids' brains were affected by iPads.
[01:19:56.720 --> 01:19:59.840]  But I'll tell you what, none of them will be kids anymore and we won't have iPads anymore.
[01:19:59.840 --> 01:20:16.080]  So, like, while those questions are interesting and important and I would like to know more about them, I am skeptical of accounts that say that the companies have lucked into durable ways to change the way we relate to their products.
[01:20:16.360 --> 01:20:18.400]  Maybe there's low-hanging fruit, right?
[01:20:18.520 --> 01:20:29.360]  Like, you know, wander around the meatpacking district and you'll see like faded signs on cool old buildings that say like, use pond soap, five cents, it will make you clean.
[01:20:29.700 --> 01:20:31.700]  Because apparently that once sold soap, right?
[01:20:31.780 --> 01:20:34.480]  And then we all got inured and we were like, well, maybe they're lying.
[01:20:34.480 --> 01:20:37.900]  And so they had to switch to like, axe, body, spray, we'll make you love God.
[01:20:38.580 --> 01:20:44.740]  Because, you know, we build up a callous over the soft spots in our attention and our cognition and our critical faculties.
[01:20:45.320 --> 01:20:58.660]  So, the idea that there's like a durable way to alter our behavior as opposed to like a short-term way that briefly works the way an upworthy headline does until you suddenly realize, you know, seven amazing things about socks you won't believe number three is a false promise.
[01:20:59.420 --> 01:21:00.880]  And then you just move on.
[01:21:01.020 --> 01:21:03.300]  I don't think... I'm skeptical that that's the case.
[01:21:03.400 --> 01:21:05.880]  If it turns out not to be true, it will be important.
[01:21:06.160 --> 01:21:07.760]  But I don't know what we do about that.
[01:21:07.860 --> 01:21:12.820]  And I know what we do about the other part of it, which I'm pretty sure is a really significant piece of the story.
[01:21:12.820 --> 01:21:14.600]  I hope that answers your question a little.
[01:21:15.900 --> 01:21:27.180]  Also, sorry for overstaying my welcome, but I think it should be analyzed as like a continuation of the 24-7 news stream.
[01:21:27.540 --> 01:21:32.520]  And how, like, the phrase, Americans have the political memory of a flea.
[01:21:32.780 --> 01:21:33.060]  Right.
[01:21:33.220 --> 01:21:38.560]  That like, oh, years ago you thought Donald Trump was an egotistical businessman from New York.
[01:21:38.560 --> 01:21:47.220]  And it dissuades us from actually analyzing the material conditions and how that brings about all of these things that control us.
[01:21:47.360 --> 01:21:51.000]  I don't think that's distinct... that's incompatible with my thesis here, right?
[01:21:51.080 --> 01:22:01.760]  The idea that, like, there isn't... that we're American... the American political character is a toddler before the advent of object permanence that forgets anything that's not in its immediate vision, I think, is very plausible.
[01:22:02.920 --> 01:22:19.020]  I also think that it's quite possible that, you know, that the reason that we are currently feeling battered by the news cycle is not merely that we have these 24-hour conduits that need to fill the news, but also that a lot of stuff is objectively going on.
[01:22:19.120 --> 01:22:20.300]  Like, both of those things can be true.
[01:22:21.400 --> 01:22:21.800]  Yeah.
[01:22:22.880 --> 01:22:24.740]  That's a question from the Internet.
[01:22:25.080 --> 01:22:36.560]  Kirk asks, with all the potential for standards and legislation, do you believe we will see real positive change or continued incentivification where the power has a vested interest in continuity?
[01:22:37.080 --> 01:22:41.740]  So the question is whether we're going to see real change or whether things are just going to continue to get worse.
[01:22:41.880 --> 01:22:44.640]  I guess this is the pessimism-optimism question, right?
[01:22:44.980 --> 01:22:47.560]  And I don't like optimism or pessimism.
[01:22:47.720 --> 01:22:50.120]  I think that they're both flavors of fatalism.
[01:22:50.120 --> 01:22:57.260]  You know, when you unpack optimism and pessimism, what you get is things will get worse no matter what we do or things will get better no matter what we do.
[01:22:58.180 --> 01:22:59.220]  So I have this colleague.
[01:22:59.480 --> 01:23:01.540]  If you're science fiction readers, you should look her up.
[01:23:01.620 --> 01:23:02.480]  Her name is Ada Palmer.
[01:23:02.680 --> 01:23:03.800]  She's a great science fiction writer.
[01:23:03.960 --> 01:23:07.220]  She's also a tenured professor of Renaissance history at the University of Chicago.
[01:23:07.480 --> 01:23:14.560]  And she's infamous there for this annual thing she does where she has her undergraduates LARP the election of the Medici's Pope for four weeks.
[01:23:14.760 --> 01:23:21.640]  They all get the name of a real cardinal or head of a great family and they spend four weeks like trading favors and stabbing each other in the back.
[01:23:21.820 --> 01:23:23.840]  There's this fake Gothic cathedral that you see.
[01:23:24.160 --> 01:23:28.020]  She has all these costumes that she's bought from theater companies that were going out of business.
[01:23:28.020 --> 01:23:31.960]  She dresses them all up in Renaissance drag and they elect a pope every year.
[01:23:32.720 --> 01:23:36.760]  And out of the final four, two are always the same, right?
[01:23:36.900 --> 01:23:43.760]  Two are always the heads of the two largest, most powerful factions in Florence at the time of the Medici's.
[01:23:44.180 --> 01:23:47.120]  The other two have never once been the same, right?
[01:23:47.180 --> 01:23:52.280]  Because human action has an envelope in which it operates that changes the facts.
[01:23:52.660 --> 01:23:54.720]  History is always contingent, right?
[01:23:54.720 --> 01:23:56.400]  It always depends on what we do.
[01:23:56.780 --> 01:24:03.120]  Now, look, if you've ever done big, gnarly programming problems, you have encountered a data set that is too big to enumerate, right?
[01:24:03.200 --> 01:24:08.040]  And when you have a data set that's too big to enumerate, you don't sort it, you climb it, right?
[01:24:08.040 --> 01:24:08.920]  You have a heuristic.
[01:24:09.180 --> 01:24:14.440]  You do hill climbing through it where you say, okay, I am here in the data set.
[01:24:14.900 --> 01:24:21.020]  Which of the points that are adjacent to where I am goes furthest up the gradient towards the peak that I'm seeking?
[01:24:21.260 --> 01:24:23.960]  And you take a step in that direction, right?
[01:24:24.080 --> 01:24:31.160]  And when you reach that point, you ask yourself again, which of my adjacencies brings me closer to the goal that I want to get to?
[01:24:31.160 --> 01:24:43.260]  And in that stepwise way, and not a particularly efficient way, perhaps, and certainly a way that might strand you on a local maximum, you can get to high ground from low ground without ever seeing the whole terrain.
[01:24:43.820 --> 01:24:46.520]  Now, the name for that is hope, right?
[01:24:46.760 --> 01:24:49.980]  HOPE is the idea, you know, very apt given this conference.
[01:24:50.460 --> 01:25:08.640]  HOPE is the idea that even if you can't figure out how to get from A to Z, that if you can materially improve your position in any way, you may ascend to some piece of terrain from which you can see previously invisible ways to get even further up towards your goal,
[01:25:08.820 --> 01:25:09.340]  right?
[01:25:09.560 --> 01:25:11.800]  HOPE doesn't always solve our problems, right?
[01:25:11.860 --> 01:25:15.300]  You can tread water for as long as you want after the Titanic sinks.
[01:25:15.620 --> 01:25:18.220]  If no boat comes, you won't be rescued.
[01:25:18.340 --> 01:25:21.800]  But everyone who doesn't tread water sinks, right?
[01:25:21.800 --> 01:25:23.080]  And so I'm a believer in hope.
[01:25:23.240 --> 01:25:24.600]  I'm a believer in human agency.
[01:25:25.540 --> 01:25:32.700]  Whether we have a better future or a worse one, or in which ways it's better and worse, I think is up to us, right?
[01:25:32.760 --> 01:25:34.980]  It's not a streetcar.
[01:25:35.140 --> 01:25:36.260]  It's got a steering wheel.
[01:25:36.440 --> 01:25:37.740]  We get to choose where it goes.
[01:25:39.220 --> 01:25:39.320]  Yeah.
[01:25:40.300 --> 01:25:40.940]  Thanks for the talk.
[01:25:41.140 --> 01:25:41.920]  Thanks for wearing a mask.
[01:25:46.530 --> 01:25:53.670]  What thoughts do you have on structures within companies that might be able to prevent future and certifications?
[01:25:53.930 --> 01:26:13.130]  So assuming that there's huge success with labor organizing to empower workers to resist it, assuming that regulation is also hugely successful, for people in organizations like platform co-ops, worker co-ops, what sort of structures, safeguards might be possible to have within the tech company?
[01:26:13.930 --> 01:26:15.390]  Sorry, I'm missing the question.
[01:26:15.550 --> 01:26:16.950]  What can tech companies... what?
[01:26:17.110 --> 01:26:17.490]  I'm sorry.
[01:26:17.690 --> 01:26:19.270]  I think the monitors are pointing that way.
[01:26:19.430 --> 01:26:21.970]  So the questions from the mic are going that way and bouncing off that wall.
[01:26:21.970 --> 01:26:24.990]  What sort of structures or safeguards...
[01:26:24.990 --> 01:26:26.410]  What... destructive?
[01:26:27.230 --> 01:26:27.830]  Structures.
[01:26:28.190 --> 01:26:28.970]  What are the structures...
[01:26:29.670 --> 01:26:32.130]  Or safeguards could you have within a company...
[01:26:32.130 --> 01:26:32.910]  Oh, to prevent...
[01:26:32.910 --> 01:26:35.650]  ...over like a platform co-op or something like that?
[01:26:35.850 --> 01:26:36.050]  Right.
[01:26:36.050 --> 01:26:49.310]  So I think that the macro phenomena that produced enshitification were held at bay very successfully for a long time by muscular competition enforcement.
[01:26:49.310 --> 01:26:49.830]  Right.
[01:26:49.950 --> 01:26:59.550]  Just by the idea that, broadly speaking, if a company gets very powerful, irrespective of whether it got powerful by doing things that we like or things that we don't, it presents a risk.
[01:26:59.890 --> 01:27:00.350]  Right.
[01:27:00.450 --> 01:27:05.390]  And that that risk can only be mitigated by finding different ways to blunt its power.
[01:27:05.390 --> 01:27:07.390]  So maybe that is breaking up firms.
[01:27:07.570 --> 01:27:12.530]  Maybe that is regulating them in certain ways that cause their power to be limited.
[01:27:12.790 --> 01:27:24.110]  I've just read a very good book on the history of the Bell system that contrasts federal regulation in Canada with municipal regulation in the United States and how they had their own different political economy at different times in the history.
[01:27:24.110 --> 01:27:25.310]  So regulators have to be nimble.
[01:27:25.530 --> 01:27:29.610]  It's, it's basically, it's like a, it's a project of management and administration.
[01:27:29.650 --> 01:27:31.710]  It's not like a thing that you do.
[01:27:31.990 --> 01:27:36.790]  It's like you examine the facts on the ground and you go, okay, which parts of this are working, which parts aren't.
[01:27:36.970 --> 01:27:38.950]  And then you adjust the parts that aren't.
[01:27:38.950 --> 01:27:49.130]  And the prerequisite for doing that is making sure that firms can be bid by democratically accountable institutions, right?
[01:27:49.150 --> 01:27:55.690]  If firms don't, can thumb their nose at regulators or at laws, then everything else goes out the window.
[01:27:56.290 --> 01:28:06.310]  So long as firms can be bid by laws and democratic institutions, then you have, you know, a separate question of like, how do you gather data on how the firms are performing?
[01:28:06.310 --> 01:28:09.890]  How do you adjust that and check that your rules are going on and so on?
[01:28:09.930 --> 01:28:17.050]  And there are a lot of, there's a lot of good gnarly discussion to be had about like, what makes a regulation administratable and what doesn't?
[01:28:17.110 --> 01:28:29.270]  Like, I really like, for example, we could have a rule that says, companies have to provide you with a machine readable standards defined way to move your social graph from one social platform to another.
[01:28:29.430 --> 01:28:38.890]  So basically the file that Mastodon spits out when you leave a Mastodon server, just a little XML file that, or maybe it's Jason, I forget, that's got all of, all the people who follow you and all the people you follow.
[01:28:39.090 --> 01:28:48.530]  And there's like a, an XML directive, like the RSS directive that just says to the next, the next time someone who follows you goes to your server and says, Oh, I'd like your updates.
[01:28:48.530 --> 01:28:50.150]  It says, no, no, look for them over here.
[01:28:51.170 --> 01:28:54.330]  Um, we could force companies to provide that on demand to users.
[01:28:54.330 --> 01:29:03.670]  And the reason I think that's a, uh, uh, interesting regulation is on the one hand, I think it'd be really powerful to make it really easy for users to go from one platform to another and then change their mind and go back.
[01:29:03.790 --> 01:29:07.970]  It would discipline platforms that, and cause them to treat their users better on balance.
[01:29:07.970 --> 01:29:11.230]  And if they didn't, it would make it possible for those users to leave all of that good stuff.
[01:29:11.410 --> 01:29:15.010]  But also it's really easy to figure out if a company's following it, right?
[01:29:15.110 --> 01:29:21.630]  If, if like you write to the attorney general and say, you know, I left Corey's server, uh, but he didn't give me the file.
[01:29:22.130 --> 01:29:27.290]  The attorney general doesn't even have to, like, open a hearing and figure out whether I gave you the file or not.
[01:29:27.370 --> 01:29:29.070]  They can just say, what's, what's your name?
[01:29:29.210 --> 01:29:30.330]  Or what name can I call you?
[01:29:31.250 --> 01:29:31.470]  Me.
[01:29:31.710 --> 01:29:31.870]  Yeah.
[01:29:32.110 --> 01:29:32.450]  Kalis.
[01:29:33.090 --> 01:29:33.430]  Kalis?
[01:29:35.170 --> 01:29:35.510]  Kalis.
[01:29:35.730 --> 01:29:40.610]  They could just say, um, Corey, Kalix says you didn't give them your file.
[01:29:41.110 --> 01:29:42.450]  Uh, just give them the file.
[01:29:42.810 --> 01:29:45.170]  Like, just, just like, give it to me and I'll give it to them.
[01:29:45.330 --> 01:29:45.670]  Right?
[01:29:45.810 --> 01:29:46.950]  We don't have to have a hearing.
[01:29:47.210 --> 01:29:48.450]  We don't have to have a process.
[01:29:48.450 --> 01:29:53.390]  We just have like a way that these files can be like verified and moved on.
[01:29:53.530 --> 01:30:00.510]  So it's not like these things where we have ideas like, oh, we'll make the platforms police their users for harassment conduct, which is like, great.
[01:30:00.650 --> 01:30:02.090]  Okay, so you say you're harassed.
[01:30:02.250 --> 01:30:04.130]  Now we have to agree on what harassment is.
[01:30:04.230 --> 01:30:05.930]  Then we have to agree on whether that was harassment.
[01:30:06.090 --> 01:30:10.350]  Then we have to decide whether or not Facebook did what they could to prevent the harassment.
[01:30:10.530 --> 01:30:17.730]  And since like, to a first approximation, everybody who understands Facebook's infrastructure works for Facebook, that means like deposing their engineers for a decade.
[01:30:17.730 --> 01:30:20.870]  So 10 years later, we're like, all right, you got harassed.
[01:30:20.970 --> 01:30:22.250]  Here's your reward, whatever.
[01:30:22.470 --> 01:30:23.770]  We take down the harassing content.
[01:30:24.090 --> 01:30:26.410]  In the meantime, 40 million other people have shown up.
[01:30:26.870 --> 01:30:29.450]  Meanwhile, a rule that just says, you can leave.
[01:30:29.930 --> 01:30:31.490]  You can just leave.
[01:30:31.730 --> 01:30:33.710]  And we can make it easy for you to leave.
[01:30:33.810 --> 01:30:37.490]  And after you leave, you can talk to the people who you're talking to before you left.
[01:30:37.910 --> 01:30:41.030]  And all that changes is the company loses your business.
[01:30:41.870 --> 01:30:45.030]  That's a thing that we can like administer really readily.
[01:30:45.990 --> 01:30:52.150]  And that like reasonable parties can all look at the enforcement regime and tell you whether it's going well.
[01:30:52.410 --> 01:30:56.490]  So there are a lot of that kind of like highly administratable rules that we can have.
[01:30:56.610 --> 01:31:08.190]  But the prerequisite for all of them is an overarching rule that you might call anti-monopoly that keeps corporate power smaller than the power of the democratic institutions that oversee it.
[01:31:08.190 --> 01:31:10.930]  And I've talked about coalitions a few times here.
[01:31:11.110 --> 01:31:16.070]  The coalition for keeping companies small, I think, is bigger than we imagine.
[01:31:16.070 --> 01:31:23.050]  So those of us in tech circles have a lot of friends who consider themselves one kind or another libertarian.
[01:31:23.910 --> 01:31:30.650]  And lots of libertarians think, well, the primary duty or maybe the only duty of the state should be to enforce contracts.
[01:31:30.870 --> 01:31:31.630]  Okay, fine.
[01:31:31.750 --> 01:31:39.830]  If that's what you think companies should be doing or states should be doing, it follows that the state has to be more powerful than the entity whose contracts it's enforcing.
[01:31:39.830 --> 01:31:40.490]  Right?
[01:31:40.590 --> 01:31:43.550]  So if you want a small state, you need smaller companies.
[01:31:44.130 --> 01:31:44.530]  Right?
[01:31:44.610 --> 01:31:53.670]  The smallest your state can be is big enough to push around the biggest company in the state, if all it's going to do is enforce contracts, even if that's all it's going to do.
[01:31:53.770 --> 01:32:00.510]  So if you're a small government weirdo who'd vote for a dead gopher if it'd knock a nickel off your taxes, you still want antitrust.
[01:32:01.770 --> 01:32:07.810]  So anyway, that's the baseline that makes everything else possible.
[01:32:08.270 --> 01:32:09.990]  I hope that answers your question.
[01:32:10.530 --> 01:32:12.010]  It didn't, but it was super interesting.
[01:32:14.870 --> 01:32:15.850]  Hi, thank you.
[01:32:16.230 --> 01:32:26.070]  I just wanted to ask a quick question because you're talking about sheetify or un-sheetify shit and you go through four points and it's very long.
[01:32:26.070 --> 01:32:32.930]  It would take me a hundred pages to propose to the commission of European what to do.
[01:32:33.150 --> 01:32:37.310]  So if you were to do it in three paragraphs, where would we start?
[01:32:38.070 --> 01:32:41.370]  And you would last into a one to three page document.
[01:32:41.730 --> 01:32:49.790]  What issue of un-sheetifying can we start with as a baby step and would you write it in a three page paper?
[01:32:49.790 --> 01:32:54.310]  I've done tons of these and I've done, I've written these as editorials, short editorials and so on.
[01:32:54.390 --> 01:32:58.050]  They can absolutely be boiled down, but you lose a lot of the nuance.
[01:32:58.570 --> 01:33:08.190]  And you know, you say, look, companies, you may have noticed companies start off good, but then they're bad to their users and then they're bad to their business customers and then it seems like they're bad to everyone, huh?
[01:33:08.390 --> 01:33:09.810]  And yet they're still sticking around.
[01:33:10.010 --> 01:33:13.210]  Well, it's not because the companies are being run by worse people.
[01:33:13.410 --> 01:33:14.470]  It's the same people running them.
[01:33:14.470 --> 01:33:17.790]  What's changed is they don't face consequences for doing it.
[01:33:18.030 --> 01:33:24.070]  And the consequences they used to face was they would lose your business or a regulator would smack them around or someone make a compatible product.
[01:33:24.230 --> 01:33:25.530]  But we made all that impossible.
[01:33:25.830 --> 01:33:28.990]  So if we're going to fix it, what we need to do is restore those things.
[01:33:29.350 --> 01:33:31.790]  So anyway, and the book is more than a hundred pages.
[01:33:32.010 --> 01:33:33.470]  I'm nearly, I'm halfway through it.
[01:33:33.570 --> 01:33:38.130]  It's going to be 70,000 words and it'll be a 2026 title.
[01:33:39.950 --> 01:33:40.510]  Yeah.
[01:33:40.510 --> 01:33:42.350]  Hi, thanks for the wonderful talk.
[01:33:42.590 --> 01:33:51.630]  So the fourth modulating factor you mentioned for institutionification is tech worker power, which I'm sure makes us all feel important here.
[01:33:52.830 --> 01:34:06.330]  But in the, you mentioned, and I think a lot of us have felt in the last few years, this feeling that tech workers have lost a lot of their power that they were given by or felt by the companies.
[01:34:06.330 --> 01:34:11.710]  And, you know, some of us are trying to fight back now with unions and try to reclaim that.
[01:34:11.850 --> 01:34:19.270]  But I was wondering if you had thoughts as to what exactly led to that loss of power, because there have been tech layoffs before.
[01:34:19.550 --> 01:34:19.970]  Right.
[01:34:19.990 --> 01:34:22.110]  And that hasn't really stopped it.
[01:34:22.250 --> 01:34:27.370]  And there hasn't been like a huge influx in supply or a huge drop in demand for tech workers.
[01:34:27.570 --> 01:34:30.050]  So what kind of made that happen?
[01:34:30.350 --> 01:34:33.230]  I think it is actually a drop in demand for tech workers.
[01:34:33.230 --> 01:34:41.290]  I think, you know, one of the things that Elon Musk proved is you could fire a whole ton of people if you didn't really care about product quality.
[01:34:41.610 --> 01:34:53.490]  You know, you could like run a hundred million, two hundred million user service service with like, you know, immiserated H1B shackled to your sinking ship by their visa status.
[01:34:54.030 --> 01:35:07.770]  You know, and I think that there are lots of firms that are like, yeah, we can just if we don't care about making the product better, if we don't care about improving things, we can just we can just make it worse and keep going.
[01:35:07.770 --> 01:35:12.410]  And the other thing that's happened is a lot of decomposition of former tech roles into quasi tech roles.
[01:35:12.570 --> 01:35:21.210]  So think about the games industry and, you know, the carving off of Tester and a lot of other jobs that are really tech jobs.
[01:35:21.430 --> 01:35:21.830]  Right.
[01:35:21.930 --> 01:35:26.670]  But that aren't treated like tech jobs and that weaken the power of tech workers.
[01:35:26.670 --> 01:35:38.270]  I think it's the best way to understand the, you know, incredible horniness for co-pilot is it's like it's it's just to it's not to really improve your efficiency.
[01:35:38.370 --> 01:35:53.110]  It's to create a kind of programmer who is easier to train to the level where they can make code that someone else can review by prompting, you know, a spicy autocomplete.
[01:35:54.130 --> 01:36:10.410]  And and then and then you can use the fact that that there that so much of your job is now being done, that bosses can use the fact that so much of your job is now being done by this person who's, you know, a code monkey to treat you worse.
[01:36:10.610 --> 01:36:10.690]  Right.
[01:36:10.750 --> 01:36:20.370]  To make you easier to replace, you know, I mean, the source, the loss of tech workers power comes from the loss of the scarcity of a technically skilled workforce.
[01:36:20.370 --> 01:36:21.890]  I think that's just that's it.
[01:36:22.170 --> 01:36:22.610]  Right.
[01:36:22.750 --> 01:36:28.850]  You know, it was the source of the power is the loss of the power like the two ways that workers have power, solidarity and scarcity.
[01:36:29.390 --> 01:36:36.370]  If you rely on scarcity, then as soon as market forces drive more people to enter the sector, the power goes away.
[01:36:36.890 --> 01:36:40.910]  And that's just like the the history of every kind of worker.
[01:36:41.450 --> 01:36:45.090]  You know, there are ways in which tech is exceptional, but that's not one of them.
[01:36:45.770 --> 01:36:46.450]  Thank you.
[01:36:47.170 --> 01:36:47.330]  How?
[01:36:50.620 --> 01:36:51.200]  All right.
[01:36:51.420 --> 01:36:53.520]  We have like seven more minutes.
[01:36:53.740 --> 01:36:57.540]  So I tell you what, ask your questions and I'll answer as many of them as I can.
[01:36:58.400 --> 01:36:59.620]  But we'll go back and forth.
[01:37:00.340 --> 01:37:00.560]  OK.
[01:37:00.560 --> 01:37:15.760]  I would say, as somebody that's actually never had a Facebook account, you know, and I'm a Zucker vegan as well, that not to deny the significance of, you know, confronting big tech on this.
[01:37:15.760 --> 01:37:20.440]  But I'm interested in micro enshitifications.
[01:37:21.600 --> 01:37:30.980]  I'm thinking of things that sometimes aren't even necessarily a product or like I'm wondering whether that's even something that you would consider within enshitification.
[01:37:30.980 --> 01:37:31.700]  Why?
[01:37:31.900 --> 01:37:32.420]  Why?
[01:37:32.540 --> 01:37:36.180]  Why my landlord can tell me, oh, you can only unlock the fucking building with an app.
[01:37:36.680 --> 01:37:36.940]  Why?
[01:37:37.260 --> 01:37:39.140]  I can't like my favorite, my personal favorite.
[01:37:39.380 --> 01:37:45.480]  I can't find fucking therapy in a non recorded video environment.
[01:37:45.860 --> 01:37:46.160]  Yeah.
[01:37:46.380 --> 01:37:46.520]  Right.
[01:37:47.420 --> 01:37:50.760]  I can't park downtown without a fucking smartphone.
[01:37:51.300 --> 01:37:51.400]  Yeah.
[01:37:51.940 --> 01:37:52.240]  Good.
[01:37:52.360 --> 01:37:52.820]  Good question.
[01:37:52.980 --> 01:37:54.360]  Go, sir.
[01:37:54.520 --> 01:37:55.600]  And that green hat here.
[01:37:56.320 --> 01:37:56.840]  You.
[01:37:57.080 --> 01:37:58.500]  Oh, thank you.
[01:37:58.500 --> 01:38:04.580]  I was just thinking, um, what if we, we've, let's see, we sold and shitification.
[01:38:04.940 --> 01:38:09.780]  There's, uh, we restored labor rights, we restored competition, we restored regulation.
[01:38:10.300 --> 01:38:19.240]  But, uh, isn't there also need to address the harm that and shitified, especially tech has caused to society?
[01:38:19.240 --> 01:38:26.360]  But, uh, well, how do we de-radicalize the racist grandmas and the uncles and the misogynist nephews?
[01:38:26.360 --> 01:38:28.900]  And there's hundreds of millions of them out there.
[01:38:29.060 --> 01:38:40.600]  And is there any need also to hold these, especially tech companies accountable for the damage they have done to individuals, families, communities and society?
[01:38:40.920 --> 01:38:41.340]  Got it.
[01:38:41.440 --> 01:38:41.680]  Thank you.
[01:38:41.680 --> 01:38:42.560]  Yeah.
[01:38:42.800 --> 01:38:51.020]  So trying to convert people to better platforms, like say Signal, I've encountered what I call swapping fatigue, where there's like, oh, no more messaging platforms.
[01:38:51.240 --> 01:38:53.100]  But then they turn around and get things like threads.
[01:38:53.440 --> 01:38:59.380]  How do big companies convert people to new platforms when we can't to our own open ones?
[01:38:59.920 --> 01:39:00.440]  Great.
[01:39:00.940 --> 01:39:01.060]  Great.
[01:39:01.060 --> 01:39:01.080]  Go.
[01:39:01.800 --> 01:39:13.260]  I'm relaying this question from the Matrix chat where a user asked, while you are correct that we are seeing a massive buildup of power in the tech world, something similar has happened within the U.S. government.
[01:39:13.720 --> 01:39:16.820]  For example, has anything really changed after Snowden?
[01:39:16.980 --> 01:39:23.220]  So why would regulation help when the enforcers of such regulation are also subject to enshittification?
[01:39:24.600 --> 01:39:25.560]  All right.
[01:39:27.580 --> 01:39:28.060]  Good.
[01:39:28.320 --> 01:39:28.540]  Okay.
[01:39:28.760 --> 01:39:29.160]  And last.
[01:39:29.400 --> 01:39:29.600]  Hi.
[01:39:30.800 --> 01:39:36.200]  Earlier in the talk, you said that enshittification was held off even like 10 to 15 years before today.
[01:39:36.500 --> 01:39:40.260]  But the DMCA for user self-help was passed in 1998.
[01:39:40.780 --> 01:39:44.120]  Competition antitrust laws were already weakened in the 1980s.
[01:39:44.280 --> 01:39:47.580]  The tech companies were starting to merge even in the early 2000s.
[01:39:47.580 --> 01:39:52.920]  So this implies that the only factor remaining was vocational awe or labor.
[01:39:53.940 --> 01:39:54.380]  Okay.
[01:39:54.620 --> 01:39:54.860]  Great.
[01:39:55.140 --> 01:39:55.340]  Okay.
[01:39:55.500 --> 01:39:59.620]  So the question of micro-enshittification and why everyone is abusing you freely now.
[01:40:00.280 --> 01:40:06.420]  So I would say that in many cases, they're riding the coattails of monopoly regulatory capture.
[01:40:06.700 --> 01:40:16.640]  So in other words, the fact that there is no effective privacy or consumer protection law is why you have this proliferation of parking apps and landlord apps and bossware and surveillance tech.
[01:40:16.640 --> 01:40:16.640]  Right.
[01:40:17.300 --> 01:40:30.120]  And it's not like those companies went to bat for, you know, bullshit landlord apps in order because they really wanted to help those those bullshit landlords do bullshit landlord stuff.
[01:40:30.220 --> 01:40:32.600]  It's just like it's a spillover effect.
[01:40:32.740 --> 01:40:32.940]  Right.
[01:40:33.040 --> 01:40:35.080]  They just they were like, no privacy law.
[01:40:35.280 --> 01:40:35.560]  Right.
[01:40:35.880 --> 01:40:38.060]  Privacy stops with the VHS cassette.
[01:40:38.460 --> 01:40:38.780]  Right.
[01:40:38.780 --> 01:40:40.360]  And and then everything else.
[01:40:40.480 --> 01:40:45.220]  And so what this leaves is a vacuum that can be filled by companies that that can abuse you.
[01:40:45.340 --> 01:40:48.200]  And I think the lesson here, like, I really hate this phrase.
[01:40:48.340 --> 01:40:50.320]  If you're not paying for the product, you're the product.
[01:40:50.520 --> 01:40:55.860]  I think it's much safer to say if the company can get away with treating you as the product, they will.
[01:40:56.460 --> 01:40:56.680]  Right.
[01:40:56.680 --> 01:41:01.440]  Companies are like they're they're going to treat you as badly as they can.
[01:41:02.320 --> 01:41:09.340]  And, you know, the the starving away of the administrative state, the weakening of a regulatory capacity and so on.
[01:41:09.340 --> 01:41:17.500]  It's meant that, like, on the one hand, we don't have the broad regulations like privacy that would stop someone from, you know, making these parking apps and whatever.
[01:41:17.980 --> 01:41:21.180]  Because the reason these parking apps exist is to get your data.
[01:41:21.360 --> 01:41:21.460]  Right.
[01:41:21.540 --> 01:41:25.500]  Like it's the parking part is like secondary or data is like not the new oil.
[01:41:25.640 --> 01:41:30.660]  It's the new oily rag and everyone wants a garage full of them and they want someone else to pay to put out the fires.
[01:41:31.300 --> 01:41:43.500]  And and, you know, that that it's also then created a system where like all because they they're more broadly invested in just making democratic institutions weak.
[01:41:44.460 --> 01:41:50.460]  That all of the other remedies that might step in like municipal government or whatever, they're just not there, you know, and you have.
[01:41:51.900 --> 01:41:52.400]  Yeah.
[01:41:52.540 --> 01:41:58.220]  And we have the but also, you know, we have these like national and multinational firms which arose out of this deregulatory environment.
[01:41:58.220 --> 01:42:05.920]  So the other thing is that, you know, the yokels who run your local parking meters are like easy pickings.
[01:42:06.060 --> 01:42:07.860]  You know, this is Lyle Langley in the monorail.
[01:42:08.180 --> 01:42:14.700]  You know, they're they're showing up and saying and saying like, well, Shelbyville put in the parking lot, the parking meter app.
[01:42:14.820 --> 01:42:16.760]  Why don't you and look at how well they're doing?
[01:42:17.000 --> 01:42:23.560]  And so everyone's just like a rube next to these guys because they're they're playing an iterated game across multiple jurisdictions.
[01:42:23.560 --> 01:42:28.660]  And, you know, these are, you know, local counselors who are who are just getting rooked.
[01:42:29.040 --> 01:42:30.800]  OK, de-radicalization.
[01:42:31.560 --> 01:42:40.220]  I don't know why people have become conspiratorialists and gotten rotten and whether or not we can make the tech companies pay reparations to fix it.
[01:42:40.940 --> 01:42:49.060]  I'm inclined to think that like the reason, you know, your grandpa became super racist in 2016 is not because Facebook hacked his dopamine loops.
[01:42:49.240 --> 01:42:56.840]  I think your grandpa was probably always a bit of a racist, but like Facebook connected him with other people who are racist and kind of emboldened him to be more racist.
[01:42:56.840 --> 01:43:01.820]  And like also like just upregulated the salience of racism to his life.
[01:43:01.980 --> 01:43:05.920]  Like people have all kinds of odious beliefs that they don't focus their lives on.
[01:43:06.080 --> 01:43:06.180]  Right.
[01:43:06.520 --> 01:43:17.220]  And and what I think happened in 2016 and what's happened since is a bunch of odious beliefs that people might have held as part of a bouquet of things that they were part of their identity have become their whole thing.
[01:43:17.660 --> 01:43:17.980]  Right.
[01:43:18.120 --> 01:43:18.120]  Right.
[01:43:18.140 --> 01:43:22.860]  Like just like just like junior high kids who are like, well, now my whole thing is I have a fedora.
[01:43:23.360 --> 01:43:29.600]  There's just like there's there's old people who are just waking up one day and going like my whole thing is great replacement.
[01:43:30.080 --> 01:43:30.400]  Right.
[01:43:30.520 --> 01:43:35.940]  And and, you know, it might be a thing that they kind of believed in the back of their minds, but it wasn't the thing, the only thing they thought about.
[01:43:36.080 --> 01:43:37.560]  And I think that's kind of what's changed.
[01:43:37.840 --> 01:43:39.500]  I think they kind of egged each other on.
[01:43:39.500 --> 01:43:45.260]  If you know anyone who's got gang stalking delusion, which is like this really sad form of paranoid schizophrenia.
[01:43:45.580 --> 01:43:55.740]  And I think everyone who's sort of tech adjacent hears from people who have this because it's people who think that they're that they're coded messages on the Internet being targeted at them.
[01:43:55.860 --> 01:43:57.360]  It's a very awful way to live.
[01:43:57.640 --> 01:44:05.580]  And I think that like the worst thing that happens with gang stalking delusion is that people find forums for other people with gang stalking delusion.
[01:44:05.580 --> 01:44:15.920]  And whenever someone says, yeah, you know, my shrink said that this tweet was not about me, but I... and someone else says, oh, no, no, it's totally about you.
[01:44:16.040 --> 01:44:16.640]  And you can tell.
[01:44:16.740 --> 01:44:17.360]  And here's why.
[01:44:17.520 --> 01:44:23.880]  And they construct this kind of awful, like, mirror world LARP in which everything is part of the gang stalking delusion.
[01:44:24.080 --> 01:44:25.620]  And I mean, it wrecks their lives.
[01:44:25.780 --> 01:44:32.500]  And I think that's what's happening to, you know, weird boomers who suddenly become, like, racism hobbyists.
[01:44:33.760 --> 01:44:36.280]  You know, I don't think that they were, like, made racist.
[01:44:36.440 --> 01:44:39.380]  I think their racism was upregulated as the center of their identity.
[01:44:40.320 --> 01:44:43.160]  And then elaborated through kind of collective storytelling.
[01:44:43.660 --> 01:44:44.760]  Swapping fatigue and threads.
[01:44:44.920 --> 01:44:46.340]  So why do people all get on threads?
[01:44:46.480 --> 01:44:47.840]  They got on threads because it was a one clip.
[01:44:48.420 --> 01:44:48.740]  Right?
[01:44:48.920 --> 01:44:58.100]  We could, like, threads actually proves that if we could just mandate portability for social media, you could go from zero users to 100 million users overnight.
[01:44:58.380 --> 01:44:58.520]  Right?
[01:44:58.520 --> 01:45:02.360]  The reason threads worked is you took your Insta followers with you when you went there.
[01:45:03.200 --> 01:45:05.100]  You know, that's, like, not a...
[01:45:05.920 --> 01:45:10.440]  It's not an impossible technical feat to make Facebook offer that through an API.
[01:45:10.680 --> 01:45:17.560]  And it's also not impossible to do with scraping and, like, you know, just enumerating your followers or, like, reversing the app.
[01:45:17.560 --> 01:45:20.420]  It's just that it's illegal to do one and Facebook won't do the other.
[01:45:20.420 --> 01:45:30.840]  So it's actually, I think, a great proof that we can overcome swapping fatigue just by, like, making it really easy to click one button and bring your followers with you somewhere else.
[01:45:32.740 --> 01:45:36.520]  Snowden and isn't all regulation enshitification.
[01:45:36.760 --> 01:45:38.800]  No, bad regulation isn't shitification.
[01:45:39.240 --> 01:45:43.560]  The fact that some regulation is bad doesn't mean that all regulation is bad.
[01:45:43.560 --> 01:45:49.320]  And the fact that, like, we have made a bunch of shitty regulations doesn't mean we're incapable of making good ones.
[01:45:49.420 --> 01:45:52.900]  Like, we are standing here underneath these reinforced steel joists.
[01:45:53.320 --> 01:45:53.720]  Right?
[01:45:53.860 --> 01:45:55.480]  And I'm not a material scientist.
[01:45:55.480 --> 01:45:57.160]  And most of you probably aren't either.
[01:45:57.440 --> 01:46:05.760]  Even if some of you are, most of us are not qualified to figure out whether or not we should be in this auditorium or whether we're all going to be crushed under the ceiling.
[01:46:06.180 --> 01:46:06.540]  Right?
[01:46:06.620 --> 01:46:13.140]  It's because we have a good regulation where, like, people sat down and actually just did the math and had the argument.
[01:46:13.160 --> 01:46:16.760]  And, you know, they all had their own parochial reasons for wanting to do X or Y.
[01:46:16.860 --> 01:46:18.100]  Some of them had patents on things.
[01:46:18.240 --> 01:46:20.720]  Some of them were good at installing one kind of thing and not the other.
[01:46:20.840 --> 01:46:22.000]  And they figured it out.
[01:46:22.160 --> 01:46:29.260]  And, like, the ability to figure out how to have potable water and ceilings that don't fall in on you, it's not a lost art of a fallen civilization.
[01:46:29.400 --> 01:46:31.040]  It's something like embalding pharaohs.
[01:46:31.340 --> 01:46:31.580]  Right?
[01:46:31.660 --> 01:46:33.340]  Like, we figured out how to do this in living memory.
[01:46:33.340 --> 01:46:34.660]  We can still do it.
[01:46:34.660 --> 01:46:46.540]  I think that the side that wants to be able to just YOLO everything and have the roof fall in on you and have that be your problem and not theirs likes bad regulation.
[01:46:46.560 --> 01:46:51.640]  Because the more bad regulation there is, the more reason there is to not even try regulating.
[01:46:52.140 --> 01:46:52.320]  Okay.
[01:46:52.720 --> 01:46:57.000]  And final question about DMCA and the timeline.
[01:46:58.040 --> 01:47:01.200]  So my argument is not that each factor is dispositive.
[01:47:01.200 --> 01:47:06.700]  That ending competition law immediately produced the anti-competitive and shittification forces.
[01:47:07.120 --> 01:47:14.120]  Rather, there are a series of threshold effects that build up through these four forces, competition, regulation, and so on.
[01:47:14.220 --> 01:47:16.080]  And that they're mutually reinforcing.
[01:47:16.460 --> 01:47:20.380]  So, like, the less competition there is, the more regulatory capture is.
[01:47:20.380 --> 01:47:27.120]  The more regulatory capture there is, the easier it is to concentrate your sector, right, through doing things like vertical acquisitions and integrations.
[01:47:27.120 --> 01:47:29.080]  But that they don't all happen at once.
[01:47:29.260 --> 01:47:38.680]  And I think what happened is, all of these forces that constrain corporate conduct, including one more that I didn't really talk about, which is corporate memory.
[01:47:38.680 --> 01:47:53.400]  So, like, if you're IBM and you went through an antitrust proceeding under the DOJ from 1970 to 1982, in which you had to outspend the entire Department of Justice on outside counsel for 12 consecutive years, in what they call antitrust Vietnam.
[01:47:54.080 --> 01:48:08.060]  That when 1982 comes along, and you've escaped the antitrust trap, right, like, you know, Reagan has been elected, he's dropped the charge against you, you're still, you've internalized a flinch response, right?
[01:48:08.140 --> 01:48:12.840]  So, when someone says, oh, you know, we should make our own OS for this PC, they're like, no, God, no, are you kidding?
[01:48:12.900 --> 01:48:15.380]  Don't you remember what the DOJ did when we tied software to hardware?
[01:48:15.380 --> 01:48:18.600]  So, you go and you find Microsoft, and you get them to build it.
[01:48:18.880 --> 01:48:20.200]  Same with Microsoft, right?
[01:48:20.240 --> 01:48:25.840]  They go through the seven-year process of antitrust, same thing, they wait until George Bush gets elected, they get off the hook.
[01:48:26.340 --> 01:48:28.500]  And meanwhile, like, Bill Gates is deposed.
[01:48:28.680 --> 01:48:33.960]  And if you've never seen Bill Gates' deposition video from the Microsoft deposition, it was the first viral video, it's amazing.
[01:48:34.200 --> 01:48:37.040]  It is Bill Gates in full panic mode.
[01:48:37.040 --> 01:48:40.860]  You have never seen him more freaked out, and just awful.
[01:48:41.100 --> 01:48:42.800]  And it ruined him.
[01:48:42.940 --> 01:48:44.360]  Like, he's still angry about it.
[01:48:44.360 --> 01:48:50.760]  And so, for years, when people would say, like, at Microsoft, hey, let's do something terrible, someone else would go, are you kidding?
[01:48:51.040 --> 01:48:53.380]  They'll put Bill back on the stand, right?
[01:48:53.600 --> 01:48:55.820]  And so, you had these after effects.
[01:48:55.960 --> 01:49:01.760]  Meanwhile, you have companies that are like two Stanford undergrads and their frat buddy who's just, you know, finished law school there.
[01:49:02.120 --> 01:49:05.820]  And, like, they turn to him and they go, like, hey, drop the papers.
[01:49:05.820 --> 01:49:08.900]  We're going to buy this company because they're a competitor of ours and people like them better.
[01:49:08.900 --> 01:49:14.140]  And unlike a lawyer who'd been in one of those companies during the antitrust, who'd be like, Jesus Christ, no!
[01:49:14.460 --> 01:49:16.400]  He's just like, yeah, I'll just draw up the papers.
[01:49:16.600 --> 01:49:22.920]  Which is why you get Zuck tripping over his own dick in the most epic dick-tripping episode ever.
[01:49:23.240 --> 01:49:32.880]  He sends an email at 12.30 in the morning to his CFO going like, hey, Bill, you want to know why we're buying Instagram for so much money, even though there's, like, 16 people working there and they don't have many users?
[01:49:32.880 --> 01:49:36.240]  It's because people like it better than Facebook and our users are leaving.
[01:49:36.440 --> 01:49:50.500]  And then two hours later, he sends another email, presumably after getting a panicked phone call from the General Counsel, that has the weird cadence of someone doing, like, denouncing the American military regime in a Korean POW camp with a gun to the back of his head,
[01:49:50.560 --> 01:49:55.780]  saying, like, hey, Bill, as you know, I am a patriotic American who values competition.
[01:49:55.780 --> 01:50:04.480]  And when I said that we should buy Instagram because our users preferred it more, I meant that there were great synergies between our two competitive American firms.
[01:50:04.680 --> 01:50:10.380]  You know, so, like, the norms changed, right?
[01:50:10.500 --> 01:50:13.560]  The things that companies tried changed over time, right?
[01:50:13.600 --> 01:50:15.460]  They pushed on the open door and it swung.
[01:50:15.600 --> 01:50:18.340]  And then they pushed on it harder and it swung further.
[01:50:18.480 --> 01:50:19.920]  It took time for them to do it.
[01:50:20.180 --> 01:50:21.880]  Sometimes different stuff blocked it.
[01:50:22.520 --> 01:50:23.880]  And they were able to...
[01:50:23.880 --> 01:50:27.200]  And they got burned and they were like, no, the door hit them in the nose.
[01:50:27.320 --> 01:50:29.040]  And they said, okay, we're not going to do that now.
[01:50:29.220 --> 01:50:42.820]  But over time, they pushed out the boundary until we got to this point where there were just no more constraints at all, where they just get up every day and they go to that enshittification lever and they pull on it and it goes all the way to 100 every time.
[01:50:42.820 --> 01:50:45.200]  And they're like, oh, this is what I've done since the day one.
[01:50:45.320 --> 01:50:46.660]  It's just the lever moves a lot more freely.
[01:50:46.680 --> 01:50:48.020]  This is like how I run a business.
[01:50:48.180 --> 01:50:49.100]  It's how I've always run a business.
[01:50:49.580 --> 01:50:50.580]  Nothing's changed about me.
[01:50:50.580 --> 01:50:53.000]  I just pull the lever every day, right?
[01:50:53.180 --> 01:50:54.260]  And this time it's moving.
[01:50:54.260 --> 01:50:56.960]  So we are living in a favorable business environment.
[01:50:57.720 --> 01:50:58.260]  All right.
[01:50:58.560 --> 01:50:59.600]  Well, thank you all very much.
[01:50:59.700 --> 01:51:00.600]  Thank you for supporting 2600.
[01:51:01.380 --> 01:51:02.840]  Thank you for supporting EFF.
[01:51:03.220 --> 01:51:03.820]  Thank you.