Alarming algorithms, digital decay

Hey there, time traveller!
This article was published 01/05/2024 (691 days ago), so information in it may no longer be current.

Have you noticed that every platform we use online, from Google to Amazon to Facebook, is just sort of … shitty now?

Cory Doctorow has. Last year, the Toronto-born, Los Angeles-based tech blogger, journalist, science-fiction author and activist coined the term “enshittification” to describe the decline of these major digital platforms on which we’ve come to depend, and it struck a nerve. The American Dialect Society even made it its Word of the Year for 2023.

“I’ve been writing about the problems of platforms and problem of technological self-determination — just getting to decide as individuals or as a community how you relate to the technology — for almost a quarter of a century,” Doctorow says over the phone from L.A.

“It’s been a theme in both my activism and my artistic work, and I’ve made up all kinds of words to try and describe different aspects of it and none of them have really caught fire the way this one did.”

In addition to a knack for capturing the zeitgeist, Doctorow, 52, is also extremely prolific, writing nine books during the COVID-19 lockdowns. His latest novel, The Bezzle, came out in February.

Doctorow will be delivering a talk in Winnipeg Wednesday called Escaping the Enshittocene: Why Everything Is Terrible and What To Do About It, and gave the Free Press a little preview. This conversation has been edited for length and clarity.

Free Press: Can you give our readers a definition of what “enshittification” is?

Cory Doctorow: Enshittifcation is a way of talking about platform decay. Platforms are this endemic form to the internet, these two-sided markets; they’ve got business customers and end users. So, you’ve got eBay, who has sellers and buyers. Uber’s got drivers and riders. Google’s got publishers and advertisers on one side, and web searchers on the other web, and so on.

It’s how these platforms decay. It describes that decay the way you describe a pandemic. So, it describes what outward symptoms are, or like what it feels like to be on a platform that is enshittifying. It describes the mechanism, like what happens inside the platform that makes it enshittifying. And it describes the epidemiology, like why are all these platforms enshittifying now.

I think that pathology would be very recognizable to anyone. You have a platform that is good to its end users, the end users pile in, they get locked in somehow, then the platform is worse to those end users.

It’s good to its business customers, they pile in too, to serve those end users. And then things are withdrawn from those business customers too and the end state is a system in which all the value that can be extracted has been and just the sort of homeopathic residue of value remains on the platform and it’s just enough to keep all of those users locked in and all of those business customers locked in, and everything else has been returned to the shareholders. And that’s the point where the platform becomes a pile of shit.

FP: And yet, we’re all still on or using these platforms. Facebook is unusable now and billions of people are on it.

CD: Our use of technology is … not an entirely individual decision. People think about Facebook as being addictive, when really what it is is sticky. And the reason it’s so sticky is because it has a dynamic that you could say is a mutual hostage-taking. You joined Facebook because the people who were there were people you wanted to talk to and then other people joined because they wanted to talk to you.

This is called the network effect and lots of digital businesses have it. The more people who are potential dating partners on a dating service, the more reason there is to join that dating service and see if any of them are right for you. We’re all familiar with that.

But the problem with Facebook is you also face something called the collective-action problem, which is an economist name for the fact that, like, you and your five friends in your group chat can’t agree on where to have dinner this Friday. It’s just hard to make decisions with other people.

Let’s say you have 200 friends and you’re all on Facebook. Some of you are there because you have a rare disease and that’s where your support group is. And some of you are there because that’s where your kid’s hockey team’s carpool gets organized. And some of you are there because that’s where your customers are. And some of you are there because you moved from another country and its how you stay in touch with the people you love back home. It’s just really hard to leave. But it’s really hard to leave because you have to choose between giving up all those people, or going somewhere better. And the problem of convincing them all to go, even if they all agree that Facebook sucks, and where to go next is basically impossible.

FP: It’s also hard, I imagine, to get people to care about this problem when many of these platforms — specifically social media platforms — have just become one big passive, numbing scroll.

CD: I actually think the trajectory of social media was that it started in a very active medium, where the way that you used it was to tell the social media platform who mattered to you by following them, and then to get a feed that you interacted with so people you have both social and parasocial relationships with.

The goal of the platforms has been to transform that into an algorithmic feed. One of the reasons for that is that algorithmic feeds are easier to enshittify.

Periodically, you get these platforms that have what feels like genuine breakthroughs in predicting what you might like to see that’s different from the things you’ve already asked to see. So that would be like TikTok, where a follow on TikTok is more like a suggestion to TikTok, and really what’s mostly in the feed are things that you never asked to see but that, by and large, heavy TikTok users really seem to say are good predictors of what they want to see.

But the problem for those users is that once you have a passive feed, TikTok can twiddle the knob, and show you more things that you don’t want to see but someone will pay to show you. Some of that’s ads, and some that’s boosted content, and some of that is stuff that for TikTok’s own reasons they want to show you.

They have this thing called the heating tool that allows them to increase the reach of posts to people who aren’t algorithmically likely to enjoy them, and that’s a way of tricking performers into relocating to the platform.

FP: I always laugh when people say they are on Instagram but not Facebook when they are both Meta. What role do monopolies play in enshitification?

CD: Lina Khan, who is the chair of the Federal Trade Commission, says the problem with monopolies is not that they’re too big to fail, they’re too big to care.

Google has a 90 per cent search market share. And the way they got that 90 per cent search market share was not merely by making a really good search engine — which, let’s be clear, 25 years ago, Google made a search engine that was so good that it felt like magic.

But the way they maintained that monopoly was not by maintaining the search engine, but rather by using the extreme branch of interactives from vertical integration, where they integrate an ad network and a mobile network, videos tool and so on. And that vertical integration was accomplished entirely by acquisitions that were themselves all anti-competitive. So, Google has had one good in-house product that really succeeded and almost, without exception, everything else that they do is with a company they bought and historically, they would have been prohibited from buying.

That resulting vertical monopoly was so powerful and made so much money, that they were able to spend between US$25 and US$40 billion a year making sure that they were the default search option on every single platform. Which means that no one ever organically tries a search engine that isn’t Google, which means no venture capitalist will ever fund a search engine that isn’t Google.

They don’t have to be good. They just have to be the only thing you end up using.

FP: Taken all together, this all feels very grim.

CD: Let me tell you, I don’t think it’s grim. I think it’s actually good.

FP: Wow, OK, say more about that.

CD: We know what policy choices produced this outcome, and we know how to reverse those policy choices. We know what to do to make it better.

If we want to fix this, we just go through all of those ways in which companies lost. We unionize tech workers. We carve out exceptions to IP law that allow for people to remediate the worst ideas at tech companies by modifying how their technology works. We end regulatory capture, and we enforce the normal laws for privacy, consumer protection and labour against tech platforms. The way we do all of that is we start with competition law that stops these companies from being too big to care.

And we are in a golden age of competition law enforcement. We’re seeing more in the last three years than we saw in the 40 years before, in the U.S., the U.K., the EU, Australia, and even China. We are at a moment where things are really changing.

And you’re right to be frustrated. You should be furious. You should be scared that these platforms that you’re using can collapse before you figure out how to leave them and take with them all the value that they delivered to you.

But you should be hopeful because this is the moment in which everything is changing and we have more potential for change now than we’ve had in any time in my life.

jen.zoratti@winnipegfreepress.com