The Trouble With Tech Companies  ## Metadata - Author: [[HBR IdeaCast]] - Full Title: The Trouble With Tech Companies - Category: #podcasts - URL: https://share.snipd.com/episode/a177dcd2-32e0-46ad-a915-7029babcb1f4 ## Document Notes ## Highlights - **Three-Stage Platform Decline** - Platforms initially prioritize end users to build scale and lock-in. - Once users are locked, platforms shift value to business customers and then shareholders. Transcript: Cory Doctorow Platforms are first good to their end users. They lure those end users in. They find a locus of lock-ins so the users can't readily leave. Once those users are locked in, they make things worse for those end users in order to tempt in business customers. One way to think about this that's very straightforward is Facebook went from promising users they would never spy on them to spying on them a lot. Having locked in those business customers, they then abuse those business customers as well. And the end goal is to draw out all ([Time 0:03:38](https://share.snipd.com/snip/c48f1bfd-2f0b-4e18-9263-3c97ad975aea)) - **How Lock-In Works** - Lock-in arises from many mechanisms: legal, technical, and social. - These switching costs prevent users and businesses from leaving even when platforms worsen. Transcript: Alison Beard Well, that's where the lock-in comes in. Cory Doctorow And maybe I'll work backwards here and talk about lock-in of business customers. So we have got a lot of discursive and theoretical attention on monopoly, right? Powerful sellers. But we don't pay a lot of attention to monopsony, which is powerful buyers. If you operate a firm, you understand that if a customer accounts for say 20 or 30% of your turnover and they just stop doing business with you overnight, that is catastrophic. Anyone who's ever done business with Walmart knows what this looks like, but also everyone who's still putting videos on YouTube as they crank down the compensation to creators knows What this looks like. It's very hard to escape a powerful buyer. But in terms of consumer lock-in, there's many different ways of ([Time 0:04:28](https://share.snipd.com/snip/22018ed6-9149-4aa4-bf29-bad704e2b3d2)) - **Printer Ink Example** - Doctorow uses printers to show software locks making cartridges expensive and switching hard. - Firmware access controls create legal penalties that enforce vendor lock-in. Transcript: Cory Doctorow This is, I think, the primary source of innovation in Silicon Valley is finding ways to lock-in end users. You can do it in concert with the state. Since 1998, America has had a law called the Digital Millennium Copyright Act. And Section 1201 of the DMCA establishes a felony punishable by a five-year prison sentence and a $500,000 fine for defeating an access control. And so if there's software in a device, and the only way to modify it is to bypass something that says, like, don't modify the software. The act of modifying the software, even if you don't infringe anyone's copyright, even if you don't violate anyone's law, is a felony that you can go to jail for five years for. And so this is how you get locked in with say printer ink, right? So the companies have merged to a cartel, four giant companies control all the printers. They have fitted the printer cartridges with security chips that are a software lock and access control. And it checks to see whether you're using an OEM cartridge with the original ink in it, or whether you've refilled or bought a third-party ink cartridge. ([Time 0:05:13](https://share.snipd.com/snip/e46385d1-741b-4f96-bc03-f911523f2b0b)) - **Uber's Market Capture** - Uber practiced predatory pricing, losing money to drive out rivals over years. - That strategy left many regions with no effective transport alternatives when prices rose. Transcript: Cory Doctorow You have Uber, which was able to violate the black letter of antitrust law by doing predatory pricing, selling goods below cost in order to remove competitors from the market. They lost 41 cents on every dollar they made for 13 years, and they crushed all potential competitors except for an also-ran, which is Lyft. And so in many places now, the bus isn't coming, there isn't a regular taxi company, and it's just Uber. And so Uber raises the price. ([Time 0:06:49](https://share.snipd.com/snip/e490b771-70c4-41f1-b1b2-ef0e8e7092ab)) - **Facebook's Three-Phase Shift** - Facebook's path: promise no spying, amass users, then monetize by selling surveillance to advertisers. - Later it withdraws value from users and advertisers to enrich shareholders. Transcript: Cory Doctorow Great stuff to crappy stuff. So when Facebook started off as a public service, they had a problem, which was that everyone who could have been using Facebook and who wanted social media already had a social media Account on another service called MySpace. And so they made a pitch to them. They said, come to Facebook. We are like MySpace, except that we will never spy on you and we'll never show you anything you didn't ask to see. So that was stage one. So as those end users piled into Facebook, they did Facebook's work for them. They found a way to lock themselves in. And they did so through something economists call the collective action problem. When there's 200 of you and you're all stuck to Facebook. And some of you are there because that's where the people have the same rare diseases that meet. And some of you are there because that's how they plan the carpool for the little league their kids are in. And some are there because they emigrated and that's the only way to stay in touch with the people they left back home. And some are there because that's how they find their audience or their customers. Then the fact that you are dissatisfied with Facebook has to overwhelm the fact that you are very much dependent on your friends. And so, so long as you love your friends more than you hate Mark Zuckerberg, Mark Zuckerberg can make things worse for you in order to make things better for another group of users. So that's where stage two starts. We draw down some of the value that has been given to end users in order to allocate it to a different group of users, the business customers. In Facebook's case, that was publishers and advertisers. So Mark Zuckerberg violates his promises and he goes to the advertisers and he says, do you remember when these rubes signed up for Facebook, we told them we weren't going to spy on them. The actual fact is we spy on them with every hour that God sends. And for a remarkable small amount of money, we will target ads to them with incredible fidelity. We will devote enormous resources to make sure that when you give us a dollar to show an ad to a specific kind of user that is the kind of user and they definitely get shown the ad before you Get built so that business customers they pile in too and they become dependent on those end users because very quickly facebook becomes a critical source of business to their firms And they can't replace them overnight they can't afford to go away. So that's stage two. In stage three, Facebook exploits the fact that business customers and end users are all locked together. And so it starts to withdraw value from both of them in order to allocate it to shareholders and investors. So if you're an advertiser, you find that the price of advertising has gone way up, but the fidelity of ad targeting has gone way down. So advertising costs go up. Ad fraud goes through the roof. In 2017, Procter& Gamble zeroed out their programmatic advertising spend. They had been spending $200 million a year on surveillance ads, programmatic ads. And when they went to zero, they saw a 0% drop in sales. Because to a first approximation, P&G's ads were either not being seen by anyone, right? It was just being consumed by ad fraud or they were being seen by the wrong people. They were just being misdirected by kind of indifferently tended ad targeting systems. Publishers also faced rising costs. So they had to put longer and longer excerpts just to be shown to their own subscribers. And to get recommended, the excerpts had to be longer still, and nothing would show up if they put a link back to their own website and Facebook said, oh, we can't tell if those are malicious Links. With the effect that they become commodity backend suppliers to Facebook, right? The stuff they post is perfectly substitutive for the material on their own website. Meanwhile, end users have found that the share of content that they've asked to see in their feed is almost non-existent to create a void into which Facebook can stuff things people Will pay to show them to see. That's the end of it. ([Time 0:07:30](https://share.snipd.com/snip/ba0108e2-ec8d-450f-beb5-c5f0a04690c1)) - **Gateways To Knowledge Can Fail** - Large platforms like Google can entrench dominance via deals that prevent competition. - Market gateways that perform poorly have outsized social and economic effects. Transcript: Cory Doctorow But it's working fine for me. Why should our listeners care? Well, I think increasingly it's not working for people. So in a way that I think most people can relate to Google with a 90% search market who were just allowed to go on by a federal judge, given permission to go on paying Apple $20 billion a year, Not to enter the search market with a competitor that might impose discipline on them, even having lost an antitrust case. ([Time 0:11:50](https://share.snipd.com/snip/935417c5-44ad-4569-b90e-3e164f7fd970)) - **Surveillance Used To Suppress Wages** - Gig platforms use data to price wages based on a worker's economic vulnerability. - This transfers wealth from labor to capital and harms service quality for users. Transcript: Cory Doctorow And it used to be that contract nurses were supplied by local firms, several local firms in every market. And now they've been replaced by, depending on how you count, three or four giant apps, each of which calls itself Uber for nurses. And because the monopolies capture their regulators, those giant apps are able to violate the privacy of those nurses in ways that shifts giant amounts of money from labor to capital. So before a shift is given to a nurse, the app buys their credit history from a data broker. Data brokers are unregulated in America. Under Biden, we got a rule regulating them. Under Trump, that rule was rescinded. And the last time Congress gave us a law about this, a consumer privacy law was in 1988 when the Video Privacy Protection Act banned video store clerks from disclosing your VHS rental History. We haven't had a single new consumer privacy law since then. And so they're able to buy the credit history of nurses and they are looking for how much credit card debt the nurses are carrying and how delinquent it is. And the more economically desperate you are, the lower the wage you're offered. ([Time 0:12:29](https://share.snipd.com/snip/62db8205-7169-427b-a7ba-3d4fafd3a9ea)) - **Erosion Of Tech-Specific Discipline** - Multiple tech-specific disciplines eroded: competition, worker scarcity, and interoperability. - Policy choices and IP law growth removed these restraints, enabling platform decline. Transcript: Cory Doctorow It's that there were other forces of discipline that worked on tech platforms and that those forces of discipline that were tech-specific also went away. And those two forces, one is that the tech workforce was historically very powerful because they were scarce. And a lot of them, I call them Tron-pilled, they fought for the user. Tech worker scarcity has gone away. We saw half a million layoffs in the sector in the last three years. And then the other one is a very interesting one called interoperability, which is making a technology that makes another technology work better. That used to be the order of the day. If a company made its technology worse, someone else would make a technology that made it better and steal your customers from you. But the growth of IP law has neutralized that as well. And so the erosion of discipline has led to the worsening of things. It's not the iron laws of economics or the great forces of history. ([Time 0:15:32](https://share.snipd.com/snip/06afe5a5-efb0-4175-8026-3f38799d2584)) - **AI's Economic Motive** - AI investment centers on replacing labor and extracting wage savings for capital. - AI excels at personalized surveillance pricing, shifting value from workers and consumers. Transcript: Cory Doctorow Have to understand that the investment in AI, which is extraordinary, has been driven by exactly one thing. And it's not making cool pictures. It's not AI search. It's the prospect of displacing labor. When Gartner says there's $13 trillion in AI, what they mean is that there's $13 trillion in wages that can be shifted from labor to capital. That's foundationally what they're talking about. And while there's lots of other stuff going on, I think that we have to keep our eyes on the prize. And I think that it's important to note that whatever merit AI has for search or anything else, it's actually not very good at doing people's jobs. AI salesmen are much better at convincing a boss to fire a worker and replace them with an AI that can't do their job than AI is at doing that worker's job, partly because they're pushing On an open door. Bosses would love to reduce their wage bills. That is like the first characteristic of being a boss is to find ways to reduce your wage bills because it's an input that is something that feels like you should be able to reduce it through Automation. As to whether AI is going to make this better or worse, I think mostly it's going to make it worse. One of the use cases where AI really is very good is surveillance pricing. It's really good at figuring out how to tighter an offer on a per-user basis, whether that's a wage offer or a pricing offer, to find the minimum you're willing to accept as a worker and The maximum you're willing to pay as a customer. ([Time 0:17:00](https://share.snipd.com/snip/6d8aff2a-348c-417b-8fa4-525cc534d9cf)) - **Businesses Must Change Policy Not Just Products** - Businesses should organize and push policy change, not just build better products. - Join industry associations that exclude monopolists and pursue antitrust actions and private rights. Transcript: Alison Beard How about we start with businesses who sort of appreciate your argument and say, yes, we've lost power. We've lost agency. We're not happy. I mean, your P&G example, we're spending all this money and getting nothing in return. What can business do to try to change all of this? Cory Doctorow Well, I have to say, I don't think there's much businesses or individuals can do. This is a policy question. I mean, what can businesses do? They can shift the policy landscape. One of the things we have to recognize is that Chamber of Commerce is extraordinarily bad at representing the interests of SMEs and often lobbies for things that SMEs owners are extremely Hostile to in order to please its largest members. And those largest members are predatory upon SMEs. But there are some large enterprises that have taken big swings at changing the policy landscape. And the best example I can think of here is Epic, makers of the video game Fortnite. Epic has sued both in the EU and the US, Google and Apple, with mixed success, but some stunning successes, including a federal case against Google, rather, for antitrust violations. It's going to benefit every SME that uses those platforms, having 30 cents taken out of every dollar they make in processing fees. Often the local chambers of commerce are better and they need to find industry associations that don't incorporate the monopolists that prey on them and that are seeking to collect Rents from them because they have these choke points in the economy. Because you can't clear the choke point just by having a better product. ([Time 0:19:00](https://share.snipd.com/snip/6900e000-f287-43ea-8f69-e38314122a25))