The Luddite

Everything is cluttered and bloated. I mean this at all scales, in its most expansive and generalized sense.

Perhaps the most obvious and pressing example is pollution. It seems that we cannot help but clutter our world with trash. We're filling the air with greenhouses gases and other pollutants. We're filling the ocean and the land with plastics. We've even recently gotten really into PFAS, or the so-called "forever chemicals" because they take forever to break down, which can now be found in most Americans' blood. We are cluttering our own blood.

There are also more subtle examples. At the beginning of the COVID pandemic, when governments were shutting down air travel, I heard a public health expert on the radio complaining that travel bans are counterproductive. They "can hinder the sharing of information, make it harder to track cases and their contacts, and disrupt the medical supply chain, potentially fueling shortages of drugs and medical supplies in the areas hit hardest by the outbreak [emphasis added]." At first, this made sense to me, but then I realized what it means: In an emergency, we are incapable of moving around a relatively tiny pool of supplies without operating the entirety of the consumer airline apparatus. What a bloated mess.

This is a widely, openly acknowledged problem at the personal level, too. Marie Kondo has made a living telling people that we have too many things and we should get rid of them, because having too much stuff is a common problem for a regular household. Then, even if you manage to keep your house physically de-cluttered, the second you turn on your favorite streaming service, you let clutter back in, because Netflix et al are increasingly filled with low-quality garbage.

We've even gotten so good at cluttering the Earth that we've moved on to cluttering space, which is a real problem. In fact, the United States just issued its first ever fine for space debris.

The bloat grows on every possible medium, which then itself becomes new growth media, upon which there will be new bloat, and so on forever.

This is a technology blog, so to drill down on this, we'll use the example of the internet. Let's begin with the user experience, or the top layer, so to speak. This website does a wonderful, humorous, and succinct job illustrating how bloated the internet user's experience has become. We have already explored one facet of this top layer bloat, in the post about Google ads. Recall that, in that post, I compared the recipe sites of 10 years ago to the blogspam, ad-filled ones of today, and then compared it again to the cleaned-up versions of the modern ones once they've gone through a recipe extraction tool. Here, still in this top layer, we see the fractal nature of our phenomenon. Recipes sites, not long ago, were completely new. Once they were created, they immediately began to clutter up the also new medium of the web, only to then start to bloat themselves. This becomes the medium upon which new things grow, like recipe extraction tools, which then also start running ads, asking for your email, offering paid services, etc.

Going slightly deeper, there's the proliferation of native apps. Whereas in the distant past, maybe circa 2013, a website would've sufficed, companies now often make a website, an iOS app, an Android app, a macOS app, a Windows app, and however many native TV and video game console systems there are. With relatively few exceptions, these apps could just be websites, and each of these apps is simply duplicated functionality, but for a different platform. They only exist because native apps are better for tracking/advertising, and companies are in a war with interoperability. Every day, in the self-proclaimed most efficient economic system, companies spend millions upon millions of dollars on thousands upon thousands of software engineers to recreate the same functionality, but for each of these native platforms.

Even the actual websites themselves, from a technical perspective, are a bloated mess.¹ The technical details, like code bloat, aren't important, but for our purposes, we'll use a wonderful, naturally occurring case study by examining nitter. Nitter is an alternative front-end for Twitter. If you take a Twitter URL, like https://twitter.com/erowidrecruiter, you can replace "twitter.com" with "nitter.net" and see the same page on the Nitter front-end, like this: https://nitter.net/erowidrecruiter.²

Because Nitter is purposefully built to be a simple, non-commercial, privacy-respecting alternative, functionally similar but without loading tons of worthless frameworks, ads, and so on, we can get a rough sense of how much of Twitter is bloat. On my computer, the Nitter page linked above loads 435 KB worth of data; the equivalent Twitter page loads 8.6 MB. Twitter loads 200x as much data as Nitter, which means it is, by this metric, 95% bloat.

This might seem outrageous, but it's actually fairly normal. The home page of the LA Times loads a similar amount as Twitter, about 8.2 MB, in the initial 5 to 10 seconds it takes for the site to be usable, but continues loading data for another 4.8 minutes, loading a total of 25.4 MB. This is before I've even interacted with the page. As soon as I so much as scroll an inch, it immediately starts climbing, in part because the page still hasn't even loaded most of its images. For reference, the entirety of The Luddite's repository — this is the entirety of the site, not just any one page — is about 11 MB, and 99% of that is the images.

If we inspect where this traffic is going, we find more clutter and bloat. Browsers allow websites to load data from other domains. This has a combination of reasonable uses, some of which are essential for a functioning, modern internet, and bad ones, like tracking and advertising. The LA Times loads stuff from some 20 domains,⁴ most of which are advertising servers and trackers. By comparison, The Luddite loads all its assets from one single domain, theluddite.org, in large part because we don't track you or sell ads. Even if you just load the page and don't touch it, as far as I can tell, it keeps uploading and downloading data forever, mostly to and from a bunch of tracking and/or advertising companies.

All this is just the easily observable bloat and clutter in the initial page load. We have yet to even interact with the page. There are near infinite examples of this bloat-first approach to making web applications. In my professional capacity, I have written entire reports, longer than this blog post, about the design and security ramifications of single pages of web applications, but those are boring as hell (and no one ever actually reads them anyway), so let's leave it there.

While we're done documenting observable bloat, I do want to continue and make some inferences about that which we can't even see. Each of these websites and apps is constantly uploading data to all these aforementioned trackers, who are then storing this massive bloat of pointless, wasted cyberspace in giant data warehouses, so that companies can spend unspeakable amounts of computational resources figuring out how to better target the advertisements the next time, thus continuing the perpetual cycle of clutter.

I don't think readers of this blog need convincing that the internet matters, but even if, for the sake of argument, we concede that it doesn't, the internet has direct impacts on the physical world. It uses somewhere between 3.6% and 6.2% of all the electricity worldwide. Further consider that many of these data centers are in drought-stricken communities. A single data center can use several millions of gallons of water per day to cool the servers, which they do because it's cheaper to cool servers in the desert with water than to use traditional air conditioning elsewhere. We have bloated data centers cluttering the desert with data at the expense of some people's drinking water and everyone's privacy.

We've built a surveillance-to-drought pipeline. Each step in the pipeline represents, under the logic of GDP, productive economic activity. When I write a post on The Luddite and you read it, that doesn't contribute to GDP.⁴ It's when I put ads on the page that the GDP goes up, until I've finally added so many ads that someone makes a blog post content extractor, and this goes on until there are companies that you can pay to delete your personal data online — that's how you get a strong economy.

Fractal bloat maximizes GDP. It's a job creator. That company linked above has, according to crunchbase, between 51-100 employees, and it has competitors. That's hundreds of people, at least, working full time in salaried positions on a single task many layers deep into the bloat.

Here too we see the appeal of generative AI. Nothing can bloat and clutter our online world as quickly as an LLM. We instantly see its value because we have internalized the bloat's desire to grow itself. Salespeople are apparently celebrating tools for automatically generating proposals for clients. If a salesperson's job were to communicate with and identify the needs of a client, then this wouldn't make sense, but a salesperson's job is to sell as much of a thing as possible. It is to bloat and clutter our world. Of course the people whose job it is to create bloat are excited about being able to create bloat faster.

What else are they supposed to do? How else are we to understand the world? The bloat's structure is never-ending and self-similar, no matter the scale. Look down a level, at the internal working at any company, and we find many jobs that exist purely to please bosses, produce meaningless compliance reports, or fit elsewhere within an entire taxonomy of pointless work.

Look up a layer, and we find entire sectors consisting exclusively of bloat. People need housing. Houses are real, but people take out a loan to buy a house. The bank then bundles that loan with other loans and sells them as some new bundled security. Other banks, hedge funds, and "financial services" firms then buy, sell, and rebundle that into yet more financial instruments, each layer with an increasingly tenuous connection to actually doing anything, but the connection is there when the whole thing implodes under the weight of its own bloated, cluttered absurdity, and people lose their actual, real houses.

So lost are we in this eternal bloat that we don't see it. It is the water in which we swim. We do notice its symptoms, again at every scale. We notice when people lose their houses en masse, or when the air in our cities is unbreathable. We notice that, in our daily lives, no one seems to have any time, but there's still so much left to do, because the bloat is eternal, filling our days, weeks, months, years, and lives with commuting, emails, meetings, shopping for car insurance, automated customer service phone trees, and quarterly self-evaluation performance reviews.

There is no compromising with the bloat. Attempts by governments to regulate it are inevitably consumed by and become it. In trying to slow down the emissions cluttering our air, for example, we've created entire new industries, which themselves became bloat. Our half-hearted attempts to constrain the bloat have turned our legal system, the mechanism by which such attempts are enforced, into a bloat of its own. Settling disputes among its citizens is a core, legitimizing function of government, yet our legal system is so bloated that regular people cannot access this most basic function of the rule of law.

The bloat is unsustainable. It will eventually end, though we don't yet know how. We can end it now, together, thoughtfully, or we can let it accumulate until it ends on its own. The former will be unpleasant and painful, but I suspect the latter will be catastrophic.