- It’s finally time to talk about the one book that’s constantly been at the back of my mind since making this blog.
- No summary this time – anyone with a pulse should read this book for themselves.
I wish I could fuse with this book. No, wait, I could eat it… But then I wouldn’t be able to give it to someone else to read. Cory Doctorow’s latest nonfiction is a marvelous tour-de-force of Big Tech’s biggering, our overreliance on platforms that have become to big to fail, and what to do about them. It’s a book everyone who’s online should read. Forget that, anyone with a pulse should read it (sorry, Kissinger).
For Doctorow, interoperability is how we move on from our current predicament of mass enshittification. That is, empowering end users through legislation, giving them the right to tinker with the stuff they own, and the right to take the connections to their friends from platform to platform, among other things. Instead of walling us in, social media platforms should open us, let us communicate with each other, and move to and from platforms freely.
The story of how Big Tech got so big is often mythologised as a bunch of plucky, speccy nerds, working out of their garages, who by sheer cosmic luck each had an idea, ran with it to Silicon Valley, and lived happily ever after. The reality is much more mundane, and far less enthralling than any penny dreadful. The loosening of antitrust laws around the world, from elected officials who knew too well why they were introduced in the first place, is to blame.
Big Tech’s not a cause – it’s an effect. To understand why, Doctorow starts with a history lesson: nineteenth-century railroad owners in the US grew dominant though the merging of trusts. Essentially, all the dominant businesses in the rail industry – and oil refineries, pipelines, and oil wells – ate each other to become giant monoliths you couldn’t compete with. If you wanted to start a new railroad business, your options were join the trust, or die.
The beneficiaries of these trusts became autocrats in their own right. That wasn’t very American of them, and the government quickly moved to bring their anti-competitive practices to heel. There’s a beautiful quote from a speech given by the senator who would go on to be the author of the first antitrust bill:
If we will not endure a King as a political power we should not endure a King over the production, transportation, and sale of the necessaries of life. If we would not submit to an emperor, we should not submit to an autocrat of trade, with power to prevent competition and to fix the price of any commodity.
John Sherman
This “harmful dominance” theory of antitrust was dominant for the better half of the twentieth century. But monopolies have power. Being able to dominate an industry with little in the way of stopping you is ambrosia for the ‘Great Men’ of history. Doctorow quotes FBI informant Peter Thiel (wa-hey, there he is again) to reflect the attitudes of monopolists: “Competition is for losers.” They’d have us think their sustained premiership at the head of Big Tech companies was due to their superb leadership, but it’s not. They’d have us think they love capitalism, but they don’t.
Enter Robert Bork: Chicago school economist, member of Nixon’s cabinet, Reagan’s Supreme Court choice, and bankrolled (among others) by folks who really didn’t like antitrust law. In arguing for the weakening of antitrust law, in the 1970s and 1980s, Bork did a Vorbis1. In the trivial sense of the truth, yes, antitrust laws were necessary to protect consumers from being harmed by ‘bad’ monopolies; in the fundamental sense of the truth, antitrust law was never about breaking up all monopolies. If folks could just use their eyes, they’d find that the language used in antitrust suggested that concentration of corporate power was good, actually. Monopolies are incredibly efficient at scale, able to produce higher quality goods at lower prices, and isn’t that a great deal for consumers?
Could you imagine a world without Big Tech giants like Apple, Google, Amazon, and so on? Therein lies some food for thought – the digital revolution that has defined the past forty years has never seen an age of strong antitrust law. Weak antitrust means Big Tech companies are harder to challenge when they put locks on the stuff you own, but don’t give you the key for.
Consider smartphone apps. They’re protected by intellectual property laws that make it a crime to modify the app yourself for an improved experience. If you want to, say, modify the YouTube app to change the colour scheme, or make it more accessible for Deaf or visually impaired folks, or, shock and horror, remove ads, tough shit. If you’re an American, and the IP holder has offices in the States, you’ve just triggered Section 1201 of the Digital Millenium Copyright Act, which is punishable by a $500,000 fine and five years in prison (other such equivalent laws exist in the UK, EU and around the world):
No person shall circumvent a technological measure that effectively controls access to a work protected under this title. The prohibition contained in the preceding sentence shall take effect at the end of the 2-year period beginning on the date of the enactment of this chapter.
Cornell Law School (https://www.law.cornell.edu/uscode/text/17/1201)
This law spits in the face of self-help. In this context, self-help means giving users the legal right to choose how they experience a service. Alongside market competition and regulation, it is one of three main disciplinary forces that should have kept Big Tech from growing too big. Weak antitrust allowed Big Tech to not only to grow big and stay big, but to also collude with itself. When was the last time you bought a phone or installed a web browser that didn’t use Google’s search engine as a default? Google practically owns the search market, and it spends fifty billion dollars a year on deals with phone vendors and web browser companies to be the default:
https://www.justice.gov/opa/press-release/file/1328941/download
The less I say about HP’s inkjet printers, their locks on printer ink, and built-in ability to remotely revoke essential services years after you’ve paid for them, the less likely it is that this post will spill over into a novella. Let their own advertising talk the talk – they’re made to be less hated:
https://geekdom.social/@bigolifacks/111545311659766447
Since the book was published, a series of articles on The Verge has chronicled the rise and fall of Beeper in its fight to bring interoperability between Apple’s iMessage and Android. In short, Beeper had been using platoons of Macs in office buildings to give Android users blue bubbles, which was horribly insecure. But a new app, Beeper Mini, employed the findings of one tinkerer’s reverse-engineering of iMessage.
I downloaded Beeper knowing full well that its continued service was a fight it had no chance of winning.2 There’s no way around the plain and simple fact that this was a hack, in the sense that Android users could now reap the benefits of Apple’s notoriously closed systems. It’s also an example of the fact that there is no technical limitations to letting one device or platform talk to another using an agreed-upon protocol. iMessage itself used to support XMPP and Google Talk.
When third parties tinker with IP, two things can happen. The first, which you can almost always bet on happening when the adversary is an individual or small group or small business, is that a press release is issued, saying that Apple is moving to protect the security and privacy of its customers, as it should. What is conveniently left out, however, is actual proof of the solution’s insecurity in the form of audit logs, CVEs, and so on. Wanna Facetime from Android? Buy an iPhone.3
Apple is hardly without sin; what Beeper did to Apple, Apple did to Microsoft nearly 20 years ago. When the MacOS version of Office kept corrupting documents in its own formats, there emerged a very serious threat of Apple customers pivoting from Mac to PC. The cost of not having a functioning productivity suite far outweighed the cost of pivoting and setting everything up again.
Microsoft Office quite famously seized its place as the top dog in productivity programs by slowly boiling the other competitors’ solutions; Office products would run flawlessly upon updating, while rival solutions had to wait for a new release that got around newly-introduced incompatibilities. Software engineers at Microsoft were being paid to sabotage their MacOS release so that when Bill Gates proposes a solution – “buy your boss a PC” – they’d see no alternative.
That wasn’t very Think Different of them. It wasn’t very engineering of them, either, come to think of it – isn’t that the engineer’s motivation, to fix things and make them better for everyone? Regardless, engineers on Apple’s side of the fence were tasked with reverse-engineering Office to figure out what made it tick. Their findings eventually culminated in Apple’s rival apps that most iOS and Mac users are familiar with today. The iWork suite could read and write to Office formats just as well, and said written files would not mysteriously corrupt themselves when sent over to a Windows machine.
Did Microsoft threaten to sue Apple? Did it smear the security of Apple’s products? No. It was all perfectly above board. Yet iWork, for all intents and purposes, was a hack, just like Beeper Mini. It subverted the assumption that only Office programs should open Office formats. Microsoft later gave up its Office formats to a standardisation body to make them open and public standards, and are now readable by any productivity suite you wish to name.
That is precisely the second thing that can happen when tinkering with IP. Usually, when the adversary is a Big Tech company, it’s not just a hack anymore – it’s a hack in a fancy three-piece suit. The hack becomes the norm.
Beyond the scope of this book, there’re similar tales that could be told in the 3D modelling world. Until Pixar’s open-sourcing of its work on subdivision surfaces in 2013, every 3D modelling vendor developed their own implementation of Catmull-Clark subdivision. This meant that if you modelled something with an Autodesk product, but wanted to import that model into Cinema4D, you wouldn’t get the same results. OpenSubdiv is now the de facto standard for implementing SubD in new 3D software solutions.
The Electronic Frontier Foundation, to which Doctorow was an employee, calls these acts of tinkering ‘competitive compatability’ (comcom). Many a program, application, platform, product, service, etc., that we use today, right here and now, has been improved through comcom. Yet, in the eyes of the law, we place far more trust in the companies who lock us in to protect us and moderate how we use their technology. You don’t have to look far to find out that they don’t do a very good job of this.
Honestly, each page of Internet Con is banger after banger after banger. It’s difficult to stop at just one topic, and the book details so much more than what I’ve written here. I’ve talked about what interoperability could do for platforms before, so I’ve got that covered:
https://brologue.net/2023/11/15/i-want-my-twitter-friends-back/
https://brologue.net/2023/12/14/wash-that-x-site-outta-our-hair/
Doctorow’s counter-theory to surveillance capitalism is enshittification. Tech companies do spy on end-users, profile us, and send our digital doppelgangers onto advertisers. It’s incredibly creepy – but advertisers aren’t the omniscient oracles they’re made out to be, nor are they the ones who come out on top in this arrangement. Elon Musk’s recent tantrum vs. adtech goes to show that these companies are abused by platforms as much as we are.
I’d instead like to finish by focusing on an area that Doctorow mentions in passing when referring to social media’s network effects, since it’s fair to say that social media is how most people who aren’t so concerned are experiencing enshittification right now:
By and large, Facebook users hate being Facebook users… And yet, they keep using it.
Cory Doctorow, The Internet Con: How to Seize the Means of Computation (p. 97-98, hardback)
The most common explanation for the fact that people who don’t like Facebook keep using it is that Facebook is somehow “addictive.” It’s easy to see why, because addicts are famous for saying that they hate their addictions but feel powerless to give them up.
But scientists who study addiction are very skeptical of this framing. While the underlying biology of addiction isn’t terribly well understood, online services just don’t seem capable of creating the same physiological changes as nicotine or heroin or alcohol.
He goes on to reiterate that network effects are why people keep using Facebook, and the reason why they don’t switch is because the matter of convincing all your friends to jump ship to someone else’s walled garden is a collective action problem that’s difficult to solve. Saying Facebook’s not addictive as we understand addiction is not a vote in favour of the harms it has brought.
You can’t say it any plainer than he has:
Facebook is a company that was founded to nonconsensually rate the fuckability of Harvard undergrads, and it only got worse after that.
Cory Doctorow, An Audacious Plan to Halt the Internet’s Enshittification and Throw It Into Reverse
There’s no denying that social media overuse can bring about real harm with regards to mental health. We know the harms it can bring, and so does Facebook (the second stage of enshittification is bringing in business customers after cheating your end users). Just as enshittification addresses the same problems as surveillance capitalism from a different perspective, so too should we try approaching social media addiction assuming that these platforms will eventually cease to be.
Much research has been done to try and quantify what social media addiction is, and you can find no end to the qualitative anecdotes online. Here’s another one: Besides hosting a landing page for his business on Facebook, my Dad loves his reels. Minute for minute, reels are how he spends a good chunk of his downtime when he needs to recover from painting and decorating. Watching reels, one after the other, is a pretty low-energy decision to make, like letting YouTube videos roll on autoplay.
The internet, and the platforms hosted on it, are tools that we use in so, so many ways: They are places we can watch videos on autoplay, see radical opinions from people across the country (or globe – where such radical opinions might just be the norm and not so radical), get community updates on local traffic jams, break the news, find out the names of things, find resources on how to do something…
But if Big Tech is an effect of lax antitrust laws, and the assumption that platforms coming out of Silicon Valley will be around forever, then so too might we suggest that if the platforms were broken up, so would our attachment to them. It’s several orders of magnitude more difficult to imagine the end of Facebook and YouTube than it is to imagine the end of the world. Nonetheless, we have to consider what the future of social media would look like if we just let the big silos burn.
I’d better address the elephant in the room. Mastodon’s not the panacea to our predicament – we’ve been trying to solve technology’s problems with more technology for years – but the two things federated platforms have that make them objectively better than silos is that 1) They fail really, really well, and 2) instances being so easy to pivot from, without ditching your friends, incentivises their owners to make a better place to be.
Users have more of a say over how they experience the Fediverse. You can say “orange” to the instance owner and move on if you feel like something’s missing; you can say “orange” to your Mastodon app, and move to another one (or make your own app if you’ve got the skills). You can say “orange” to massive federated instances, like Threads, adding an extra layer of content moderation. Similarly, if your instance is centred around certain political views, you can still reach the people who want to hear them, but your posts are not entitled to reach the feeds of federated communities who don’t.
You can’t say “orange” to autocrats of trade like Elon Musk or Caesar Marcus Augustus Zuckerberg. Musk’s safe word is vox populi, but only ever hears vox Dei. You can’t say “orange” to how you experience the X app, either, unless you want to eat a Double Felony Contempt-of-Business with a Section 1201. While end users have the right to develop and install extensions that can change X’s layout, you can’t bring Twitter back. It’s dead, Jim.
Mastodon, for what it’s worth, just multiplies the number of intermediaries we can choose from. We could go further – have no intermediaries, and do everything peer-to-peer, like Napster or BitTorrent:
Which begs the question, again: if these massive platforms broke up, or disappeared overnight, how would our approach to tackling social media’s problems change if, say, federated or P2P platforms became the norm? How would the problems themselves change?
No-one, especially Doctorow (who posits Internet Con as a ‘shovel-ready’ solution), believes solving the Big Tech problem will be a cakewalk. It’s certainly not going to be an overnight job. One thing that gives me hope, though, is how nice Big Tech companies appear to their customers. You have to hope your adversary is a good entity. Hope like hell.
Chekhov-Pratchett’s Post-It: If in Act I a monopoly has you at its mercy, hope like hell that monopoly promises to not be evil. Not being evil means you have to write a lot about how nice you are. Not being evil means you have to get a lot of people in on not being evil – on post-its, DMs, in guest talks, in the canteen kombucha…
Hope like hell the monopoly doesn’t do evil – by Act III, someone’s going to dig up the post-its and find that nice ain’t good.
The Internet Con can be bought DRM-free, directly from Doctorow’s website: https://craphound.com/category/internetcon/
It’s also available at Verso Books: https://www.versobooks.com/en-gb/products/3035-the-internet-con
It’s ALSO available as a DRM-free audiobook at libro.fm: https://libro.fm/audiobooks/5184389278340-the-internet-con
- See my review of all the Discworlds I’ve read this year: https://brologue.net/2023/12/06/that-old-back-catalogue-part-i-discworld/ ↩︎
- Since it has come to an end, I’d like to point out that Beeper has open-sourced a solution using Matrix bridges:
https://github.com/beeper/imessage
You might be able to tell that this solution, while it technically works, it’s effectively untenable.
↩︎ - It’s at this point that it’d also be fair of you to argue, “Why not install WhatsApp, Signal, Telegram, etc.?” What Beeper did was it made a point of saying that making iMessage interoperable – which it proved was possible – would incentivise Apple to make a better, more secure default service. Likewise, these third-parties would be incentivised to do better in response.
↩︎