Fixing the Web, Part I
Three Things #101: January 7, 2024
I’ve written about a huge array of topics here (and in older blogs) over the past few years: technology, politics, culture, deep tech, media, family, travel, economics, and of course Spacemesh, just to name a few. It may seem like there’s no connection among these disparate topics but in the back of my mind I always felt like there was one that would reveal itself one day. To be clear I didn’t set out with the intent to tie it all together, but I’ve always had the sense that there was a common denominator. For the first time I’m beginning to see what that is.
I don’t think I can clearly articulate that common denominator quite yet—it’s not yet that clear in my mind—but if you forced me to I’d say it has to do with reimagining software, the Internet and the Web in a human-centric fashion as a newer and better medium for human connection and communication. (I realize that’s quite vague. I’ll work on improving it.) One of my goals for this newsletter over the coming year is to narrow the focus a bit and start to home in on this topic.
Let’s start with media and content. Over time I’ve become more and more obsessed with content—with creating it, consuming it, collecting and curating it. The entire bundle of these products and experiences feels extraordinarily broken today in some very profound and important ways. This, like so many things, has a lot to do with broken incentives.
You may wonder, if I care so much about media and content why am I building a blockchain? These ideas are connected: in brief, the blockchain is the necessary infrastructure to reimagine media and content. At the moment Spacemesh is a bit like Bitcoin: the main thing it can do is mint coins via mining and allow coins to be sent, and not a whole lot more. But I have grand visions for Spacemesh, several of which I laid out recently (and surely many others have visions at least as grand as mine). One of these visions, though I haven’t been able to fully articulate it, is for Spacemesh to power a new content-based information economy.
This is a big topic and a big problem space. Before trying to fix something it makes sense to lay out the problem. I’d like to use this space for a time to consider the most important ways in which content, and the web in general, is broken today and what might be done to address those problems using fresh technology and fresh ideas. These things are very broken and I can’t even cover all of the biggest problems here in one week, but this is a starting point and I intend to continue the list.
Here are three of the biggest and most important ways in which content and media, and the web more generally, are broken today.
Thing #1: Broken Links ⛓️💥
It sometimes feels like half the time I click a web link the link is broken or the connection hangs. Sometimes the page loads very, very slowly and sometimes the formatting loads incorrectly. This is by design: the web is of course decentralized and servers, sites, documents, and the underlying resources that power them (Javascript scripts, CSS stylesheets, databases, API backends, etc.) come and go. But this leads to a “swiss cheese” problem where pages and documents that still exist link to ones that don’t, and pages sometimes link to scripts and stylesheets that can no longer be loaded. By design, there’s no way to verify the integrity of a link without attempting to visit it—and two users that do this in two different places and/or at two different times may get completely different results.
Web sites, pages, and these other resources disappear for lots of reasons. Sometimes people just lose interest and stop maintaining them. Sometimes a bill isn’t paid. Sometimes companies and agencies disappear, sometimes they change policies or leadership, and sometimes people intentionally remove content that they no longer want to share. Sometimes people just get pissed off and mic drop.
I think the web would be more functional if it were immutable. Once a page, document, or other resource is published, you should never be able to remove it. In practice it sort of already works this way thanks to the Internet Archive Foundation’s Wayback Machine and other, similar tools, but the Wayback Machine can only archive a small percentage of all sites and content and it’s not able to archive all of those underlying resources. It certainly can’t archive every API backend and every row in every database!
In my opinion the web should be like a blockchain: an immutable, append-only ledger. You should be able to “revoke” a document, and most frontends should respect that revocation, but the original should still be visible in the history—like a git repository or a wiki. The web, like a blockchain, should be unopinionated and very consistent and available. Different frontends and services built on top of this underlying infrastructure should of course be free to filter, interpret, and display this underlying data as they like. This immutability would prevent broken links and provide accountability. It would also enable the use of content-addressable storage, where the address used to fetch a document is intrinsically bound to the document itself, so that changing the document in any way changes the link to it. This is simply a better system than the one in use on the web today and it would solve some of the issues outlined here such as two people clicking on the same link and getting two different things.
(As an aside, people who don’t want to share content publicly and permanently already have the option of publishing it somewhere other than the clearweb, on apps with ephemeral content like Signal or Tiktok. The web definitely is not the right medium for every sort of content.)
Of course, this would also require that someone store all of this content, which would be prohibitively expensive if done in a centralized fashion. And it would be subject to something of a tragedy of the commons: everyone would potentially benefit from such a resource but no one would want to pay for it. It would be very difficult to monetize without running into all of the other problems described here.
Enter blockchain and cryptocurrency. The best way to achieve this is with a decentralized protocol and good mechanism design. More on what this might look like in a bit.
Thing #2: Paywalls 🧱
When the web was new there was no such thing as a paywall and all of the content was free. There was a strong ethos of making as much information accessible as possible and no one was really thinking of business models, which isn’t surprising given the academic origins of the Internet. That state of affairs clearly wasn’t bound to last. Pre-internet publications such as newspapers and magazines began to colonize the early web, and in parallel web-native content such as blogs began to appear. Several content-related business models emerged, some from the pre-web era (subscriptions, advertising) and some native to the web (social media, freemium, sharing economy, online marketplaces, crowdfunding, decentralization a la Wikipedia and open source, etc.).
As is so often the case, pre-web business models didn’t translate terribly well into the web era. It may have made sense to subscribe individually to newspapers and magazines when there weren’t that many to choose from in the first place and when an individual subscription included an enormous amount of diverse, curated content. The NYTimes is the canonical example of this: it includes current events, local, national, and international news, investigative journalism, sports, entertainment, cooking, product reviews, travel, real estate, etc.. Today, however, in our modern, fragmented, digital world, we tend to consume content in a much more piecemeal fashion. While the earliest websites did attempt to do this (remember the days of AOL, Compuserve, and “portals”?), today it’s impossible to imagine a single website or property giving you all of these things. It’s an outdated model of content creation, aggregation, and distribution! The very idea seems untrustworthy: how can one company or one platform do so many different things, and do them all well?
In fact the period of mass aggregation that took place in the middle of the twentieth century was a historical aberration, and moving to a more fragmented model is really a return to the way things always were (Balaji Srinivasan refers to this phenomenon as “history running in reverse”). So there’s nothing wrong with this on the face of it, but in practice there is a problem: each individual content provider wants to charge you separately for their content! The aggregate model has one huge advantage: you pay a single monthly or annual subscription and get an enormous amount of diversity—it’s an all you can eat buffet, a one stop shop. Both the NYTimes and cable TV provided such a model, where they curate and bundle the content for you. You don’t get to pick and choose precisely what you want access to since these packages always come as a pre-selected bundle, but it severely reduces cognitive overhead.
Let me be clear: I don’t mind paying for high quality content. The problem is that I don’t want to manage dozens or even hundreds of subscriptions. Every digital newspaper, magazine, Substack creator, Onlyfans creator, Patreon creator, podcaster, YouTuber, and X content creator wants me to subscribe. This is on top of “the apps”: Apple TV, Netflix, Paramount Plus, Disney Plus, Max, YouTube Pro, Spotify, Apple Music, etc.. The list of content subscriptions we’re expected to manage has gotten out of hand. Subscriptions are dangerous because you pay for them even if you’re not using them, and it’s too easy to forget you’re paying! I really do wish I had the time to comb through every one of my credit card and bank statements each month. I don’t. No one has time for this much complexity or for managing this many accounts. It’s gotten to the point where I’ve just given up trying to subscribe to anything else, or to consume any paid content whatsoever. Goodness knows there’s enough free content out there to last several lifetimes.
Which brings us back to paywalls. No matter how much I want to read an article there’s basically zero chance I’m going to subscribe to get past a paywall. I either find the same content for free somewhere else or else I give up entirely. Huge, important web properties like WSJ, Financial Times, and Forbes simply don’t exist to me. It’s not that I can’t afford the money to access these; it’s that I can’t afford the time or mental energy to manage the subscriptions.
There must be a better way. I think blockchain and cryptocurrency gives us a better way. Lots of folks have talked about micropayments for years. I wouldn’t mind at all paying $0.10-$1.00 to read an individual article, especially if the transaction is seamless and basically invisible, i.e., it’s something I don’t have to think about much. And I’d like it even better if I could access a bundle of content curated by someone I trust. Unfortunately, I don’t trust any of those apps I mentioned above, and I certainly don’t trust the NYTimes.
This would unlock an enormous amount of latent demand for content creators and curators. There must be millions of people like me out there happy to pay for quality content and curation, if only there could be a better, easier way to do so. What’s missing is both a platform to deliver all of this, and a much better crypto wallet and payment UX.
Thing #3: Ads 👁️
While we’re on the topic of business models let’s talk about the most popular business model of them all: advertising. As discussed above advertising is a pre-web technology that hasn’t translated terribly well into the web era. It’s still enormously popular on apps and platforms that serve digital content but I think the main reason is risk aversion (it works, even if not terribly well), laziness and a lack of creativity to try other, more novel, more promising business models.
Advertising worked well enough when it meant static ads in the pages of newsletters and magazines. In fact, ads in places like magazines can work quite well when they’re well designed, nonintrusive, and relevant to the reader. There’s no fundamental reason ads couldn’t work this way on the web! Obviously digital ads can be as well designed as, or better than, their analog predecessors since digital designers have more degrees of freedom at their disposal (more mediums, crisper, cleaner displays, multimedia including images, video, audio, and text, better tools, better access to archives and samples, and simply more dynamism). And it should be clear that digital ads can be more relevant—since as we all know online platforms have orders of magnitude more data about us than old-school newspapers or magazines.
Intrusiveness is sadly another story. Static ads on print media are almost by definition nonintrusive. Most print ads are relatively small and, well, boring. Even full-page ads are great since they don’t compete for space with any “real” content. The web is another story, of course. There’s been an arms race in digital ads for many years along at least two dimensions. The first is loudness. The first web ads were simple static banners or sidebars, usually with plain text or at most grainy images. The day the animated GIF was invented everything got worse, and it’s all been downhill since. Ads have gotten bigger, brighter, and louder—in short, more and more intrusive—over the years since. As users have become inured to ads, advertisers have made them more and more obnoxious in order to grab attention.
The second dimension is the browser arms race. Popup ads became a huge thing over a decade ago. Browsers and ad blocking browser plugins responded by blocking popups. Then pop-under ads started appearing. Every time browsers and extensions got smart enough to block a type or class of ad, a new, even more clever, obnoxious, intrusive one would emerge. It’s an arms race that can never end because browsers and extensions can’t go too far in blocking ads or they’ll block useful content as well—an impossible balance that advertisers continue to exploit. In my experience even the best, most effective ad blockers, such as the one built into the Brave browser, still only manage to block around 90% of ads. There are always a few that find a way through. And ad blockers are often paid by advertisers to let some ads through.
I don’t know enough about the history of advertising to know why this never happened in offline content like newspapers and magazines. Maybe it’s because print media never had to compete as fiercely for eyeballs and attention as digital media. Maybe it was just as chaotic in the early days of those mediums. In any case by the time I was reading them advertisers, publishers, and consumers had long since reached an equilibrium (or a stalemate) because print advertising was never as annoying as web advertising. By contrast, I find the web more or less totally unusable today without extremely powerful, up to date ad blocking software. Every time I have to turn off the “Brave shields” due to some incompatibility (another way web publishers and advertisers get you) I feel completely smothered in ads. The horrendous experience of the web today really makes me miss print media!
As I said above: surely there’s a better way. Surely it’s possible to reach some detente, to have ads that are respectful, well designed, nonintrusive, and relevant—and are therefore far more effective than the garbage ads we have today. It’s in everyone’s interest. As with paywalls, I have nothing fundamentally against ads if they’re done right. As with the above things I think what’s missing is a better system of incentives, and maybe some better technologies too.