How to Save the Internet

This article was originally posted to my substack in 2023. It is a topic I am very interested in, hope to learn more about, and wish to write much more about. It is the first dive into various ideas which have informed the creation of this website and the beginning of this project. I have decided to rewrite and expand sections of the essay. I think I did a bad job the first time, and additional comments and clarifications will make this post better.

Introduction
Posted to an online forum in 2021, the Dead Internet Theory laments the modern internet. The author, IlluminatiPirate, describes how the internet today feels sanitised, safe, and unoriginal. They argue this is because the US government uses AI bots to generate content, and keep us docile.

The Dead Internet Theory has a flawed explanation for what is wrong with the internet. But it draws on a sincere concern. Facebook is the website that radicalises your grandmother. Instagram is a one-stop-shop for multi-level marketing schemes. TikTok's algorithm is crack cocaine. Amazon is a minefield of premium subscription buttons. LinkedIn is a psychopath support group. Google is always listening. Paywalls surround everything. Nothing is true. Twitter is X.

Many netizens will share the feeling that there is something wrong with the internet. For instance, a recent essay by Maria Farrell and Robin Berjon covers similar topics, advocating for a 'rewilding' of the internet to save the online world from its coming decline. Another essay, by Maggie Appleton, argues that our current online ecosystem is transforming the internet into a 'dark forest'; where online interactions and content must be sealed off by niche communities else they be poisoned by the coming onslaught of AI generated content.

Why is the Internet Bad, Now?
Why does the internet feel like it's getting worse? The dominant economic model of the internet is a good place to start. The internet is highly centralised around a few websites, notably Google and Facebook, whose business model involving capturing user attention to feed people advertisers. This, in turn, incentives these sites to prioritise visual and algorithmic designs which keep people scrolling, and naturally, transforms how people use these sites--particularly influencers who themselves are looking to profit. This, on its own, can contribute to a hollowness, or swallowness, surrounding the late-stage internet.

But, more importantly, eyeballs only become dollars when advertisers can be attracted to these centralising platforms. Hence, the perogative for most of these internet titans is making a web which is advertiser friendly. A quite famous example of this is YouTube. In 2017, YouTube was accused of showing children inappropriate content. Worse still, ads were running on these videos, spooking big brands. In response, YouTube changed its monetisation policy, prioritising 'ad-friendly' content. This incident was dubbed the 'Adpocalypse'.

Of course, being 'ad-friendly' is not inherently harmful, I guess, but if it demands a response, or a change in procedure and the scorn of some human actions, it can become problematic when decisions about what is, and what is not, acceptable are taken unilaterally. YouTube and others have built their brands off emphasising community and individual involvement. But these stakeholders get little say when advertisers are dissatisfied. Thus, rather than building a more inclusive internet per se, corporate interests sanitise these spaces, dictating the broad rules for inclusion. Inevtiably, interesting content--the type of stuff that has important and unpredictable effects on online culture and beyond--tends to suffocate.

Sometimes these dynamics play out slightly differently, but they are there all the same. Consider VRChat, probably the only 'successful' metaverse project. While, as of writing, the metaverse is a dead (or at least dying) idea, the lure of a completely privately controlled virtual world is too tempting for Silicon Valley to abandon. But Meta's Horizon Worlds project is a flop. There are probably several reasons for this, though a major one must surely be that it just doesn't look fun. It looks safe, boring, and corporate. The gaming journalists People Make Games have an excellent video about VRChat, which covers much of this ground. VRChat is not advertiser friendly. It is in many ways antithetical to corporations, with essentially no respect for copyright or intellectual property, either. These are precisely the features that make this metaverse popular. As one user, quoted in the Financial Times stated: "It's [VRChat] much like the early days of the internet, where it's oddballs and weirdos coming together to find their tribes." But, as above, such features antagonise the current advertising business model of many platforms.

One word which should probably be included in this discussion is 'enshittification.' As the writer (and coiner?) Cory Doctorow defines it, "the term describes the slow decay of online platforms such as Facebook." But this is just another word to describe a trend; a brand for a campaign, so to speak. It is more interesting to think about the dynamics of enshittification and platforms--how does it happen, and why? Doctorow must be interested in this, too, as his keynote at 2023's DEF CON focuses on precisely on this topic.

Doctorow's argument is quite simple. Firstly, that much of the internet has developed by one service subverting the restrictions of another, more popular servce, through exploiting interoperabilities between online services. Secondly, the exploiting service becomes popular, before itself becoming more restrictive (or, perhaps, simply less innovative and more cluttered and sluggish). Thirdly, the new dominant service is undermined by a new exploiter. Except, for Doctorow, major platforms like Facebook and Google have used their online hegemony--and economic capital--to change laws to prevent future interoperability (not to mention simply buying would-be rivals out).

If we follow this logic, VRChat (and other would-be 'rewilding' efforts, to borrow Farrell and Berjon's term) will either a) be bought, and slowly sanitised as it is transformed into a commercially-friendly product; or b) destroyed through legal efforts from a consortium of commercial interests. Both would lead to an enshittified online space.

DIY Online
As above, I am more interested in thinking about strategies for revitalising the internet (the titular 'how to save the internet'), or at the least, engaging with the internet and the platforms which dominate it in a more conscious way (for someone interested in this topic, much of what I have said above will not appear especiallt original). Doctorow's enshittification thesis is, for my immediate purposes, a sufficient theory of the decline of online diversity (again, to borrow some language from Farrell and Berjon).

Ivan Illich was a controversial social critic. He rallied against various institutions that most people consider important and essential, such as healthcare and education. More broadly, he criticised industrial society and the industrial mode of production, arguing that in many ways industrial society takes more freedom from us than it gives us in return. In fact, to Illich, industrial society does not give us freedom, it gives us stuff, which is really just a by-word for further wants (I plan to write much more on Illich, who certainly deserves more general discussion and study).

An aspect of Illich's thought which I think has utility as a strategic idea for the present discussion is the idea of forgoing our intellectual independence and will to autonomous living. The result is often a society in which actions are undertaken for the service of institutions, rather than institutions serving those for whom they are, apparently, established. The best place to look on this topic is Illich's Tools for Conviviality and his discussion on science, which shares quite a lot of overlap with Feyerabend's work:

"Overconfidence in "better knowledge" becomes a self-fulfilling prophecy. People first cease to trust their own judgment and then want to be told the truth about what they know. Overconfidence in "better decision-making" first hampers people's ability to decide for themselves and then undermines their belief that they can decide... People who have unlearned how to decide about their own rights on their own evidence become pawns in a world game operated by megamachines. No longer can each person make his or her own contribution to the constant renewal of society." (p. 86-87)

While the above focuses on science, and the nature of the public expert, this is simply a more specific statement of a broader criticism that industrial society tends to lead people to lose knowledge (and tools) which allow them to operate autonomously. This is to say, we are less free to choose how we live our lives. To be sure, this is both a simplification of Illich and a perspective I am not entirely in agreement with. My disagreement arises because there are many things I do not know how to do, but would neither want to know how to do. Self-dependence is lonely, and so willingly allowing oneself to be dependent on another is often agreeable--indeed, according to some, this 'collective brain' constitutes human culture. The simplification, though, is that Illich fully recognises that we are interdependent creatures, and argues that this interdependence would remain a crucial feature of 'convivial' society. His objection is that industrial society coerces us into states of personal infantilisation, and more so, into a state of dependence on machines rather than each other. Anyway...

The relevant bit of Illich's philosophy is an encouragement of rediscovering self-dependence. This comes in two parts.

Firstly, in Tools for Conviviality, Illich argues we should be self-limiting. He does not denigrate human ingenuity, but cautions against the often unforeseen consequences of fully pursuing the material and physical possibilities afforded to us. For instance, faster cars create possibilities for fewer, bigger stores, eroding the highstreet and one's sense of community. My interpretation of Illich's notion of self-limitation is that it invites us to think about the social trade-offs which come from technology, and the illusions contained within.

A good example is AI art. Anyone can draw, paint, sculp, and so on. Many of us would not do these things with particular skill or grace. But, there's nothing wrong with that. AI art generators represent a technological 'solution' to a perceived 'problem' that most people are bad at art. Of course, this isn't true. It's just something that we've been manipulated to believe (in the words of Illich), or a belief that's been manufactured (in the words of Galbraith). Still, so long as we believe it, we choose the technological 'solution' and ignore the real problem that most people limit themselves by industrial expectations (e.g., if you do not produce professional quality art, you are a bad artist).

Secondly, in Shadow Work, Illich emphasises building shared domains--public spaces, common resources, free knowledge, and so on. For Illich, only when everyone has access to public commons can we choose how to live our lives, because only then might we each have the freedom to study, to debate, or ignore apsects of our society. We would not all be fully informed of the institutions and operations to which we subsume ourselves from time to time, but we would be ignorant by choice, and should that choice change, so too would our relationship with those subsuming forces. To continue the AI art example, the lack of common resources for someone to produce art, and to learn about art if they wanted to, reinforces the false binary between good and bad art, which--as above--creates the conditions for a technological solution. Access to common resources mean we are bound only by limitations we have consciously accepted, and probably had a say in setting. Absent these resources, we are bound by limitations either subconsciously or coercively (both of which Illich considers 'manipulative').

As immediately above, I think these ideas can be applied to understand some dynamics of current technology. Through using these ideas to think about the current internet, I suggest we can also arrive at a perspective which achieves the 'rewilding' of the internet; the saving of it.

The Town Square is Everywhere
I think an interesting and valuable place to start applying these ideas is with the philosophy of the early internet, and 'web1' as it is now called. The backbone of the internet is HTML--a language so basic, some don't regard it as programming. With a default text editor and an internet browser, you can play with HTML. During the early internet, many did just that (with the aid of a domain hosting service). Thousands of websites emerged from people playing with HTML. There were guides and courses, and certainly, some corporate involvement from parties who were seeking to establish the internet as a 'thing' and thus (eventually) profit from it. Nevertheless, for many, the internet was a shared domain to do-it-yourself. It captured part of the magic that makes VRChat appeal to many today.

JavaScript turned web1 into web2. Well, it was a key component insofar as it enabled (and enables) many features which are now crucial to the archetypical web2 service, those being social media websites. JavaScript works with HTML, and allows for greater website functionality by providing dynamism in the code. This is to say, it can respond to conditions, such as those given by a user, and thus create a kind of feedback and responsiveness which the 'static', 'flatter' websites of web1 could never really achieve. The downside of JavaScript, in my opinion, is it adds to the learning curve for website programming, and thus immediately makes website design more alienating; more specialised.

Today, professional programmers, rather than 'amateurs,' run the internet. Many people now depend on large platforms for news, entertainment, and their careers. Through these services built and sold to us by 'expert' web developers and computer scientists, we do not need to know how websites work. Instead, it is easier to let 'professionals' build the internet, and increasingly, professional website building services. The average netizen is no longer an agent in the online space, contributing to the internet, but a user, a line in a database, simply experiencing an internet.

For Illich, the question would be whether this is all worth it. Websites today have many features that simple HTML sites did not. In some ways, the internet is enhanced by these features. But in other ways, we have forgone our online autonomy. It is now somewhat embarassing to have a personal website; instead, we have social media pages. What matters less is that people are enjoying what we are putting online, or even that we are enjoying what we are sharing with the world, but rather, that people are engaging with our 'content' in an artificial, arbitrary, and ahuman manner--through likes, shares, followers, and so on. Is the modern internet really worth the autonomy we have had to give up for it? Is a slightly more 'professional' online presence (whatever that means) really worth the sacrifice of our ability to customise said presence with fonts, colours, sounds, amongst other more socially and politically important aspects? It is really worth it when some find themselves despairing that their favourite social media site is collapsing, and there is nothing they can do about it? It is really worth it when online creators find themselves mercilessly at the will of an algorithm and the billionaire who owns it?

To me, this Illichian problem is summed up by Elon Musk's assertion that Twitter (now X) is the 'town square of the internet.' For many, Musk's ownership of Twitter represents a defacement of the 'town square'. But in my mind, this is wholly backwards. The sin comes when we accept that there is a town square to begin with. Why must Twitter be the town square? Why couldn't we enjoy a vibrant, interesting internet through a series of independent blogs each linking to one another through HTML's most important component, the hyperlink? Why does online discourse need to be centralised and packaged in a corporate blue which, over time, we have come to see as the 'way the internet should be?' It doesn't have to be like this, though the emergence of rivals to Musk's Twitter--such as Mastodon and BlueSky--demonstrate that those 'experts' who build so much of the internet's infrastructure lack an imagination to build anything different. In the absence of an autonomous online space, in the absence of each of us being able to contribute, in Illich's language, to the 'renewal' of online society, the destiny of the internet is stagnation and death. We are experiencing this presently.

Something which I love about HTML, and about the WWW protocol, is that it is a public common. It is something that anyone can play with. We should embrace this. If Twitter and the rest of the monoliths of web2 are dying, let them die. We should embrace a kind of dorky, embarassing, but crucially autonomous and expressive internet of blogs and dumb web forums. I think some share this ethos. Jaron Lanier is an interesting talking head on technology, whom I do not necessarily agree with. But his website is great, and given his rejection of social media, I have to think he approaches his online presence through a similar philosophical outlook I am advocating here--do it yourself (also check out wiby.org). My decision to build this website stems from this ethosv(and my desire not to be a hypocrite). Substack and other such services might make blogging easy; but on my own website, with lines of HTML I have cobbled together in a text editor, I can make this website whatever I want. In doing so, I am rejecting the false idea that Substack looks 'better' or 'more professional'; I am rejecting the false relationship between the value of my writing the number of 'likes' or 'followers' I have. I do not want to be followed, and I do not want my thoughts and opinions to be dictated by whether others 'like' what I have to say.

I also think it is dangerous to say 'burn it all down' without articulating a positive message for what should be built on the internet. I have mused on the idea of a wholly new internet--what would that be like? And my firm answer is: terrible. If the internet ceased to be tomorrow, within a couple of weeks, half a dozen would exist. We would have Meta's internet, and Google's, and Twitter's, and so on. All would be walled gardens, with the tools available to the individual controlled and limited by unelected, unaccountable corporations (readers of Pondsmith's cyberpunk work will note this is exactly what happened when Rache Bartmoss destroyed the internet in that fictional world). web2 is dying; but the internet is full of possibilities. And so, yes--burn it all down--but only so that we have space to do something better with it.