The Internet Is (still) the Platform
It started with a post from Mark Zuckerberg. It wasn't the content of that post that drew my attention – that was just the Metaverse pitch recycled with "AI" swapped in – it was the form. It felt quite surprising to see plain text in Times New Roman following a good old H1 tag. But something felt off. The page didn't load instantly - something one would expect from a minimalistic page like this. There was some weird layout shift in the middle of the page rendering - something one would not expect from a minimalistic page like this. And if you opened it without Meta cookies already in your browser, up came the consent dialog.
A peek at Safari's Web Inspector Network tab explained a lot. In order to show roughly 4 KB of text, the page loaded:
- 7 CSS files
- 17 JavaScript files
- 3 external HTTP requests
- 6 tracking pings
- 1 HTML document
In total: about 1 MB of data, of which 0.4% was the content and 99.6% was the overhead.
It’d be easy to shrug this off as just Meta being Meta — of course they’re going to track the hell out of you on every page. But the thing is, that single page felt like a perfect snapshot of the Internet today. We keep turning to platforms like Facebook for things they are really bad at. In Zuckerberg’s case, it makes sense — he can’t exactly skip his own product without looking bad. But more and more, I see people and organizations using these platforms by default, even when better options exist.
And honestly, the size of the transfer itself doesn’t really matter. Hardly anyone cares if a page takes 4 KB or 1 MB to load. Our networks are (mostly) fast enough, our devices (mostly) powerful enough. But that technical overhead is a pretty good metaphor for the other overhead you take on when you use platforms like Facebook.
Because you never just get the content. You also get the cookie banners, the trackers, the suggestions for what else you “might like,” the constant background hum of a system built to collect as much as possible about you. Every click, every scroll, every second on the page is feeding a machine that isn’t really about the thing you came to read — it’s about keeping you in the ecosystem and showing you more ads.
That 99.6% overhead in the network tab is the same kind of overhead you feel as a person: the distraction, the manipulation, the subtle drain of giving up a little more data and attention than you intended. The content is there — but it’s buried under everything else.
All of this made me realize something: we’ve somehow forgotten what the Internet itself is. The Internet isn’t Facebook, it isn’t Instagram, it isn’t TikTok. The Internet is the ultimate platform — open, flexible, and built for content, communities, and communication without the built-in overhead of corporate surveillance and attention engineering.
There are better ways to do the things we rely on these centralized platforms for.
Blogs (like the one you're reading!) let you read and share ideas without the constant nudges to “like” or “share”. You can subscribe to an RSS feed to follow anyone without sharing your e-mail address for yet another newsletter.
Smaller virtual communities built on web forums or Slack groups can host real conversations without tracking your every click.
Adult-supervised virtual communities for children's sports club or other hobby groups don't need to rely on Meta services. Signal groups, private forums or other independent tools work just as well – without the baggage. Why is it such a big deal if a child has a Meta account? Because as soon as they sign in, the tracking isn't limited to services owned by Meta. Thanks to technologies like Meta Pixel – and various half-criminal tactics, millions of websites across the web feed identifiable browsing data back to Meta. Good luck just trying to find out when your next soccer practice while being bombarded with personalized ads which know exactly what your weak spots are.
Using these better tools removes the invisible overhead that slows down your attention, manipulates your behavior, and treats your personal data as a commodity.
On top of that, many of the big debates we see today — around content moderation, age verification, content rating, misinformation, and harassment — are really symptoms of this same worldview, where a few huge platforms dominate how content is distributed. When billions of people are funneled through the same systems, every problem gets magnified. Centralized moderation teams are tasked with sifting through horrors at a global scale, while users feel powerless and communities feel impersonal.
Contrast that with smaller, independent communities. These spaces are naturally more manageable: the communities are humane-sized, participants usually share some common interest or context, and internal norms and moderators emerge organically. Problems still exist, of course, but they’re far easier to handle, and the social incentives actually work. There’s no need for a single monolithic company to try to moderate the entire world.
At the end of the day, the 1 MB it takes to load Zuckerberg’s 4 KB post is just a small reminder of a bigger picture: the Internet was designed to be the platform, and even after 30 years, it still is. We can reclaim it — one blog, one forum, one thoughtfully run community at a time — and in doing so, reduce the invisible overhead from content, attention, and human connection today.
You could say this is just the same old argument — we’ve been here before, with the Web swallowing the Internet, then Google swallowing the Web, and now social platforms swallowing everything. But that’s why it matters: every decade the Internet drifts toward centralization, and every decade it needs a reminder that it was built to be open. And the next wave is already here — the LLM gold rush is on, with billions of dollars poured into a race to become the ultimate centralized & final destination for all Internet users. All the more reason to remember what possibilities the Internet really provides to us, at least for now.