This is a kind of personal post about what I've been feeling about where I am with relation to the state of the internet.
My point here relies on recognizing a process which has happened over my lifetime. What isn't happening here is nostalgia. I understand that the state of the internet is what it is and that definitely not everything was better. A lot of things were indeed worse.
I have some "practical" suggestions below, but what I am really suggesting though, is to look back to help us ask the question about where we want to take this whole thing. In an age of internet giants it's hard to remember that the internet is the one place we can steer. I do believe, maybe naïvely, that what makes these virtual spaces is individuals, and not the "big players".
My suggestions at the end are highly individual and demand a certain level of dedication and technical expertise. I understand they are not for everybody.
With this out of the way:
I am old enough to remember the internet of the early 2000s. It was a very different place. For one thing: a lot of the conventions that we know now to be self-evident, did not exist at all. Websites would quite often be self-contained Flash apps and they would often have intros and other things which would be considered anti-patterns. What (relatively) few "apps" out there were all different in quite important ways.
But the most relevant difference here is how generally DIY everything was: if you wanted anything, you probably had to do it yourself, to some degree at least. If you wanted to host a forum to discuss your niche hobby, for example, you were going to have to buy hosting services and deploy a system (phpBB was a popular choice) or even roll your own. You could not turn to Reddit, or Facebook. The venerable WordPress would launch in 2003, but its hosted offering would not hit the scene until 2009. The assumption was that you had a typical LAMP-stack (Linux-Apache-MySQL-PHP) somewhere, and you had what little technical skill was required to install WordPress. This actually was not that difficult. A bored teen would have been able to figure it out on their own as yours truly can attest to.
The hurdle of the starting cost was not the end of it, however. After you were done launching your forum or blog - congratulations: you are now the owner of your own infrastructure. Updates are often manual and error-prone; any customization was on you to do (or not). More crucially: The very structure of whatever it was, it was entirely then up to you, including moderation.
This last bit is important. Your host would not be in the role of playing cop, unless you were explicitly breaking their fairly content-agnostic terms of service. And even if you somehow did disagree - there were other hosts, including just hosting your things yourself. All of this is to say - moderation was now up to you.
Compared to their streamlined modern counterparts, it was a very varied ecosystem. A lot of the great websites of the era really did feel like complete worlds in their own right. They looked vastly different (like I mentioned, the standards had not emerged yet), they had different features and values, and they were literally governed by different rules. Unlike their modern platform-hosted counterparts, they also resisted being aggregated and broken up into smaller chunks and experienced piecemeal: it was a website, and you had to visit it, its contents and its presentation were one and the same.
The ease of use when it comes to starting up in these platforms cannot be overstated. You need virtually no expertise, and you can be done within minutes. It doesn't even cost anything, not even the cost of a domain. Beyond that too: you do not need to persuade your prospective members to open an account, because they likely already are on said platform. Growth, if that is interesting to you in any way, is mostly taken care of by your host platform as well. It is easy to understand the allure.
The cost is that it makes for a very uniform experience that exists within very tight parameters, governed by a specific code of conduct. This is OK, and I'm happy that these options are available, but there are problems.
Suppose there is a disagreement between your particular community and your host platform. First of all, if the platform does not like your particular community, it will cease to exist. They are allowed to do that, that's part of the deal. If certain posts violate a platform's terms of service (but not your own internal moderation rules), they will be removed. That, too, is part of the deal. There is no recourse - it's not your website.
Generally speaking however, history has taught us that we are quite powerless to prevent the large platforms from making unpopular moves - such as what happened on Reddit over API changes. The result was some communities simply no longer existing. A lot of user comments have been actively deleted by their owners, so that information no longer exists either. At the time of writing, Wikipedia mentions that the outcome was "mass migration of subreddits to alternative platforms", but realistically speaking, where are you going to go if the one service you've used is the only game in town - as is true for a lot of communities? Suppose there had been an alternative. Relocating is incredibly difficult, if not virtually impossible. Subreddits cannot relocate all existing threads, comments, and users to a new service. They have to re-open in another platform, and instead rely on its good will to continue to exist in a way that is agreeable to them.
Twitter/X too, has experienced a similar exodus during its heel turn, except on an individual level. There, people were not able to simply take their accounts, their tweets, their follow lists and their followers and simply move them elsewhere.
Facebook used to be good. When it first launched it was a clean alternative to the shocking and wild MySpace. It was a chronological feed populated by the people you chose to follow. One could log in, see what's new, and eventually exhaust any new information it had to give: the end of the feed. It even had an API.
All of this is unthinkable today. Years later in the days of the algorithmic feed it became less of the stream of updates from my friends of a decade prior and more of a machine carefully tuned to extract my outrage, joy and addictive tendencies - known in aggregate as "engagement". It would do this by attempting to lure me into debates and disagreements often with strangers and sprinkle in some ads. 1
It is a pale shadow of what it was. There used to be discussion groups. The Arab Spring was basically organized on Facebook. For me, in college, Facebook was the tool of choice to coordinate anything with my peers, and it was used to great effect. Later it continued to be a useful resource (marketplace, for one thing, certain groups as well). But through a process known as enshittification, it seemed as though the price to pay for these was simply increasing with every shitty upgrade. By the time 2019 rolled around I started periodically deactivating my account until in 2021 I left for good - i.e. I proactively deleted my account, and it no longer exists.
I did not miss it either. More of these services started following suit: I no longer browse Reddit very much, and I've already scaled down my involvement with Instagram. 2
I used to love Spotify. When I first started using it in 2017 I discovered so much new music which I was able to access freely and at any time. It felt like a cool friend who introduces you to new music. Sure, not everything was available, but it was enough for me to more or less abandon the music collection which I had meticulously curated over my teens and up until then.
Then things started disappearing, such as when Joni Mitchell and Neil Young pulled their music from Spotify, in protest over Joe Rogan's weirdly popular podcast. Or more recently, when a lot of artists pulled their music out of the Israeli market. Sometimes it's not even that, but just corporations fighting over intellectual property (do I care? no). I ended up not listening to certain things, not because I chose not to, just because it was "difficult" and I had to step out of Spotify's walled garden. It's not wrong to oppose Joe Rogan or the Israeli government, but this non-ownership model has created the infuriating option of removals and geofencing to serve as some kind of financial, political or some other kind of tool. So whenever Warner Bros. or Neil Young have some kind of grievance, Spotify is going to stop being your cool friend and turn into your conservative dad who will confiscate your CDs. 3
So in case you wanted to listen to this weird trip, good luck! Because a higher power has decided to not afford you that option (except track 5 in particular): 4
A similar fate has befallen Netflix. When it first picked up steam it saved us from the tyranny of cable television. It was cheap, available, and had quality shows — in stark contrast to linear TV which was expensive, generally not on demand and had the worst programming imaginable interspersed with (unskippable, if you can imagine) ads. This would not last of course, and as soon as intellectual property-holders realized that streaming was a viable model for turning a profit they would pull their shows and movies out of the service — and launch their own service. Now we're back where we started: television has re-balkanized, and an average TV-watcher would have to get 5 separate subscriptions, each with their specific geofenced selection. It didn't even take ten years.
My "consumption" of whatever media or lack thereof is a mere consequence in a bigger battle waged above my head in faraway courtrooms, like some medieval peasant in a backwater village that occasionally changes hands due to complex courtly politics to which they are not privy.
Nowadays we would marvel at "the old internet". If we go back to the scene that this piece opened with: Facebook vs. MySpace. MySpace, for all its many, many flaws gave us a lot of freedom in how we were able to choose to use it. Pages were infinitely customizable: as in, they let users inject their own HTML onto the page. The result was auto-playing music and barely readable pages, but to be fair it also afforded a lot of possibilities. Your band's Instagram profile will never get close to looking like this:
MySpace was very typical in the genre of "personal websites". The pre-platform
internet was an eclectic medium, rife with <iframe>s and other types of
embedded content. In fact, YouTube owed a lot of its early popularity to exactly
this: it was possible to upload a video to YouTube and see it show up on a
different corner of the internet.
It was weirdly easy to make websites on the early internet. Website builders were easy enough for GeoCities to launch in 1994 and give a good amount of control. I argue that it is generally harder to make a website today than it was in the year 2000. The reason for that is that people expect your website to work not only on their desktop computer at 800×600, but also on their ultra-wide monitor, in darkmode, on their mobile phone, and their screen-reader. None of this is negative, but it certainly didn't make things easier, now that we have to be considerate of these requirements.
As a result the internet back then still had the qualities of a print medium in some ways (print was king), as well as those DIY free-layout qualities that FrontPage or a good old slice table could give. 5
Deferring to platforms makes this impossible. They by their very nature have to impose certain rules on what a "page" can be and what can be customized, and precisely how.
Before deleting Facebook, I did the prudent act of downloading my data. This left me with an archive of individual HTML pages and some media weighing in at 900 MB, containing excessive, creepy or useless information about me: every single message I ever sent, every comment, every post, IP addresses and locations spanning two years, days on which I have looked at marketplace ads. The list is very long. In preparation for writing this, I actually looked through it again which made me realize that this was not so much the photo album I'd run into my burning house to save, so much as it was the stack of incriminating documents I rather should burn.
Somewhere in a data center, there were these just-under-900-MBs pertaining to this one guy who happens to be me. The company formerly known as Facebook cared enough to collect this information (to sell ads).
I hate to be that person, but why do we think that's OK? As a practice, and as a business model. If we're letting companies sit on this data, we're effectively incentivizing companies to manipulate users, because that's the one way you can extract money out of data. So all they have to do is just sit on whatever bits of information and try to build a model smart enough to capitalize on it. So the dump contains this wonderful gem for example:
You can and should make websites. This is why I'm posting here and not on Medium or wherever. I mentioned earlier that it's hard - and it is. But one thing that did change, is that we now have LLMs, and with all their flaws they sure can help you make a website.
For my part, I'm going to see what can be de-platformed. One big example from my own life: I still post Instagram stories. It's the one place I still have a connection with real people. I want to have some place where people I know can get a glimpse of what I'm up to. For what it's worth, the solution may honestly be an old-fashioned blog (something separate to my main website which is for "serious things"), or some variation of it.
If you have a small business, maybe you should make a tiny website. A lot of smaller businesses have been forgoing websites altogether for other solutions: Instagram and Google Maps seem like a common combination. I feel like a lot of their problems could have been solved by having a very understated website. These platforms are good at making your presence known, but terrible at handling large chunks of information, especially permanently. What if you want to host your restaurant's menu? A common hack on Instagram is to put it in a story and pin it to the profile, but that is no way to browse a menu. Or what if a bakery does custom cakes, and needs to give specific instructions for how to order that (and maybe pricing)? A web page solves this kind of problem quite easily.
I think there's an obvious gap here in the skills required to create an Instagram profile vs. making a tiny website. It seems to me that there is an opportunity in all of this: offering good tooling to generate reasonable websites in a way that doesn't leave the user beholden to the platform nor the framework: a kind of FrontPage for our present day.
Looks like it's time to revive (or start) our music and movie collections - the ones that were actual files written on an actual medium. I'm getting mine out of storage and buying up some new music that isn't available to stream (or isn't available anymore)
The reason here is not necessarily to be completely independent of streaming services, so much as bridging gaps left by the streaming services' internal intellectual property struggles. Your listening habits don't have to change just because this or that company has a monetary dispute, and taking charge of your own media diet is as easy as just saving a file on your hard drive.
If more of us did that, they would find it harder to rely on high switching costs to remain in business. There is also the benefit to some artists who will sell media directly. There is no way that it is not preferable to whatever current predatory model Spotify or similar would offer them.
There is an obvious cost to this. For one thing: you are now the owner of a new problem, organizing and storing files. Secondly: the lack of a streaming option is just not as convenient. I have personally set up a Navidrome server on a Raspberry Pi I keep at home. It works, for the most part and I am able to listen to my saved music. Spotify isn't going to tell me I can't have weird things and life's good.
Video is a bit easier as most of our watching happens either when we're at home in front of a TV that is probably "smart" anyway, or on airplanes for which we can prepare ahead. Setting up a shared drive and just accessing that on a TV is, all things considered, relatively easy. Android TVs have a VLC app that will work fine in my experience.
These solutions unfortunately require "real" servers, at least to the extent of having some kind of physical computer with a hard drive attached to it. This has become a tall order, and a lot to demand. That said, Symfonium appears to support reading from some consumer cloud storage solutions, e.g. Google Drive. Somehow I doubt Google would just let you host a bunch of audio/video files for very long if they suspect they are not entirely above board, but I have been wrong before.
With audio this is quite doable. Converting CDs to digital files (i.e. ripping) is still an option, as is buying digital music outright (Bandcamp being the obvious example here).
Movies/shows can be more difficult unfortunately. I doubt a lot of films or shows that came out after ~2017 would actually have bothered even publishing physical media that you could rip. This could mean resorting to piracy, which of course you should not do.
There was an incentive structure involved that caused this. Cory Doctorow (who also coined the term enshittification) wrote about this far more extensively and far more eloquently than I ever will. This did not stop me from writing about this here also here. ↩
Instagram I mourn in particular. There's something very troubling about the machine-curated reel-based horror it became. ↩
On the subject of Spotify, there's a whole book about how Spotify weaponized its market near-monopoly and its recommendation algorithm to do things like push AI-slop. I have not read it so I do not know if it's good or not ↩
Somehow I was not the only person who noticed this absence, and I found a Reddit thread from 2025 discussing this exact thing. ↩
For the younger folks in the audience: the way we used to design websites for a while, well into the 2010s in some places, was designing them in Adobe Photoshop, and using the "Slice Tool", one would make little "regions", or slices. Photoshop would then export an HTML page with a table whose cells were at those sizes, and images to fit the slices. This was absolutely frowned upon by 2010-ish and still definitely done regardless. ↩