The web doesn’t have a centralization problem, it has an imagination problem
The opinions stated here are my own, not those of my company.
The mythos of web3 goes like this:
Web 1 was created in the 90s and was made up of static webpages. Web 2 came with the advent of Amazon and the like where you can buy things on the web. This caused a massive wave of capitalist funding. Facebook centralized social, other companies centralized other things, and the web became a mess. Companies dependent on ad revenue prioritized clickbait and attention over creating real value.
web3 with its focus on decentralization and peer-to-peer payments (blockchain) creates a more equitable web that resists centralization and thus the ruining of the Internet.
Yet this is not an accurate history of the web nor a strong analysis of its current state. It glosses over much of the secondary aspects of what has made up the web and prioritizes an approach that doesn’t offer meaningful solutions.
A better history of the web
Web1 was not static webpages. Many were interactive even in the Internet’s early days. Services like Usenet were around even in the 1980s to allow for communication. Users shared content and discussions in various public message boards known as newsgroups. It wasn’t really that different from Reddit.
By the 90s people could interact online using message boards. These bulletin board systems allowed for communities to be created and moderated around any topic or set of topics.
The web was interactive, at least kind of. At the time there was no direct way to send content to a server and wait for a response. Submitting data through a form on a website would require a page load.
When I add a post to a topic on a message board, my browser would go into a loading state. The data had to go to the server, record my post, then load a new set of page content ideally with my included post, send that back to the browser and render it.
With slow connections and slower browsers it was often painful, yet that was the way it was back then. This makes me feel a bit old to admit using Internet Explorer, but this was a perfectly reasonable way to use the web. It was interactive.
When people talk about Web 2.0 it’s easy to talk about the rise of platforms and social media, but they miss out on the web’s richness.
I would say the advent of AJAX is really what distinguishes Web 2.0. It is a method that allows a webpage to make an Internet request or send some data without necessitating the entire page to be reloaded. This means that the data being passed back and forth can be smaller. It also means that you don’t need to just call your server but others.
This is really what opened up the web to create the ‘interactivity’ that is often mentioned. Being able to load any number of requests on the same page led to novel applications like Gmail.
This really improves the user experience and it was useful as the basis to create social applications. People could create profiles on MySpace or upload videos to YouTube.
But let’s not say that this is when venture capital took a look at the web. Amazon was founded in 1994, using “Web 1.0” technologies. There were countless companies who saw the commercial potential of the web right away and rushed with cash to build the future. The dot-com bubble came and went by the time Web 2.0 was being discussed.
Host a website
People point to YouTube, Facebook, Twitter, and a few others at the entirety of the web. While it’s true that they may make a large portion of web traffic, the web has a long tail. Consider that I’m writing this blogpost on Medium.
For example, I saw Valley live in New York about a week ago.
What I like to do at the end of each concert is create a Spotify playlist to recapture the feeling of being there. My memory isn’t good enough to have every song remembered in order, but I can turn to the web to do this. There is a niche site setlist.fm which just hosts setlists from concerts.
That’s one small example of a useful website that isn’t necessarily going to capture a lot of users but can exist thanks to the web.
“Now Nick,” you’ll say next. “Your cute little story involves you writing on Medium and listening to music on Spotify. You’re using these massive corporate platforms.”
Yes, I choose to do things for the convenience. I could host my own blog like many do, but I don’t want to. If you are interested, there are a lot of different tools out there.
Digital music may be a bit harder to move away, but not that hard. I could always get digital downloads of the songs and then host it on a Plex server on my desktop.
“Okay, so there are a few alternatives. But how do they get funded? Setlist.fm uses ads, and it’s technically funded by Live Nation. It’s not an independent site.”
Maybe this isn’t the best example, though it is one of the millions of niche sites out there.
However, there are numerous websites being created and hosted everyday. Wordle was one that just appeared and became popular. Most don’t, and that should be okay. Proponents of web3 talk about a better web while looking to replicate the exponential growth of Web 2.0 over the sustainability of the web.
Solution looking for problems
When these proponents look at a better web, their first look is to blockchain. It represents some global state for application developers to use. Events in these applications are blockchain transactions, requiring each action to use at least some fee for gas.
How does this gas fee affect my ability to use an application? Does every song I add to a playlist have to be a separate charge? Do they get bundled together? How expensive is it going to be? With Spotify I can create as many playlists as I want, each with as many songs as I want, and I don’t need to be charged numerous microtransactions.
The web provided us with abundance. Wikipedia provides countless knowledge. Spotify provides a large music collection, as do Bandcamp, Soundcloud, and even any website if they want. These things are often free as the marginal cost is insignificant.
Moving back towards a world of nickeling and diming me doesn’t feel very good. Is it better than having an ad flash in front of my eyes? Maybe. But even existing payment models can be more aligned with marginal costs.
I sponsor a few different groups on Patreon. With a monthly fee I’m able to access their content while still having an abundance mindset. Maybe it’d be cheaper to go with microtransactions if I just wanted access to one post or another, but why does this access need to be recorded on the Blockchain rather than just having Patreon know it?
Despite what some may claim, the global availability of NFT resources aren’t really going to be as broad as you may like. A service like Movies Anywhere is going to be more than adequate for what anyone would want.
If a system doesn’t work for the average person, it’s not going to scale well. This is why AJAX made the web more ubiquitous through improving the user experience.
And despite what proponents might suggest, web3 applications are not that decentralized. NFTs in Metamask are pulled from OpenSea, so if OpenSea bans an asset it largely disappears from your collection despite the data still being there.
But with OpenSea accepting credit cards, the user experience becomes better yet more opaque. People aren’t going to care if it’s on the blockchain or not, since OpenSea is largely a central authority.
It’s easy to claim that ads will become less necessary, that content creators can earn royalties for each sale of their items. Yet already it seems like this is resulting in copyright theft and weird rent seeking behavior. So the money isn’t going to the right people.
“Now Nick,” you’ll argue. “These are a few unfortunate exceptions. What’s clear is the old model isn’t working and we need a new model.”
So what should this new model be?
Imagining a better web
This is not the first time web3 has come up. I recall it distinctly being brought up during my high school years.
What was called Web 3.0 is also called the Semantic Web, or the machine-readable web. This was meant to create and define a variety of standard schemas which would give a browser greater context for exploring content between websites.
What is more surprising is that this largely has come to pass. In 2013, it was estimated about 4 million web domains contained Semantic Web markup. I really don’t know what that number is today.
It shows up in some places, but not enough. A few different markup types can be pulled in by the Google Assistant and search results. Yet unfortunately this potentially rich set of data is not used enough by app developers to browse and make sense of the web.
Not all of this content is going to be public. Sometimes it will require the user to be authorized to access it. OAuth 2 is an existing industry standard for accessing content across decentralized servers and clients. Payments can be one way content owners can manage payments, which could be a credit card, a cryptocurrency, or anything else.
I agree that the web is imperfect. I don’t want to browse dozens of different websites for event details, so I created a way that does it for me and converts the data to a standardized calendar format.
I do want to spend more time exploring schema.org types and seeing how to similarly make greater contextual sense of the web in a way that works for me. Tools are available to do it which exist today and have existed for quite a while.
There’s not really a need to break everything down. It requires us to be a bit creative in deciding what we want the web to be, and using standards to get to that point. But as we approach it, we shouldn’t forget what makes the web so distinctive: it’s ability to minimize marginal costs and allow us to browse and interact with content from a variety of standards-compliant hosts.
Because the web is a diverse ecosystem, much more diverse than is often given credit. If we just imagine the web to be Facebook, all we’re going to build are more Facebooks.