Much of the experience of browsing the web is actively hostile towards users.
Web pages are stuffed to the gills with obnoxious advertising. Spammers and scammers are constantly trying to trick us into revealing our passwords, or into infecting our computers with malicious software. Cookie banners interrupt us every time we visit a new site. The news article we want to read is obstructed by pop-ups and overlays begging us to “subscribe now!” or “download the app!” Opaque algorithms bombard us with content we don’t want to see. Unelected robber barons at big tech companies try to shield us from ideas they’ve deemed too “dangerous” for the commoners to discuss, and from material they’ve deemed too prurient for adult viewers.1
Beyond just the web, we have credit card companies controlling what kinds of products we’re allowed to buy, and from whom. Media conglomerates control when, how, and where we’re allowed to access the movies, music, and ebooks we’ve purchased. Device manufacturers control what software we’re allowed to run on the computers in our pockets and on our wrists. Foreign and domestic governments mine our video calls, financial transactions, and phones’ GPS locations for mass surveillance. Email providers’ spam filters control with whom we’re allowed to correspond.
Instead of empowering us, technology is being used to take away our own agency, to shackle us in walled gardens controlled by people who think they know best. Our dystopian future is already here, and it’s pretty widely distributed.
A web browser is supposed to be a user agent2. In other words, it’s
supposed to act on behalf of the user, with the user’s best interests in
mind—not the best interests of the website owner, or of the advertising
company who funds the browser’s development, or of the government, or of woke
activists demanding someone be canceled.
An individual’s web browsing history is incredibly personal. It can reveal our biggest fears and our most private secrets. It can reveal what products we’re likely to purchase, or for whom we’re likely to vote.
Attempts to use machine learning to customize the web experience have brought us the Facebook newsfeed and YouTube’s video recommendation algorithm, among others. But these are once again at odds with users’ interests. Facebook and YouTube are funded by advertising, so they use dark patterns to trick us into clicking on ads, and to make it difficult to click away onto a competitors’ site. There’s immense power in controlling what news we read and what videos we watch, so populists and politicians are forcing companies to shape their algorithms in ways that are hostile to the user—and often in ways that are hostile to scientific discourse or a healthy democracy.
The digital assistant is a common trope in science fiction. But today’s so-called “assistants” assist our digital masters way more than end-users. Alexa, Siri, and others are sealed boxes that are not under their owners’ control; the warranty is void if the sticker is removed. The always-on microphone that’s uploading our voice to Amazon’s servers has an end-user license agreement which prohibits any reverse engineering. Apple won’t even let us run Mozilla’s or Google’s web browsers on the iPhone we supposedly own; why do you think they’d let mere users have any influence over Siri’s search results?3
Insights gleaned from our behavior should not be in the hands of advertising companies, to be sold to the highest bidder. Our online history shouldn’t be stored in the cloud, where it can be collected in bulk by intelligence agencies with secret court orders. We alone should hold the keys to the safe with our personal data in it, and we should be able to use our own data for our own benefit.
What we need is an actual user agent, running on a trustworthy computer. Instead of mining our behavior for ad targeting purposes, it could learn from how we interact with the web sites we visit and the apps we use, and filter out the hostile aspects which lead to a bad experience. It could scour the web on our behalf, seeking out content based on your or my preferences, not the preferences of marketers, SEO experts, or political spin doctors.
-
In newspeak that would make George Orwell proud, Steve Jobs famously said he wanted to give iPad users “freedom” from being able to access pornography. ↩︎
-
The term dates back at least as far as the original specification for HTTP, probably longer. ↩︎
-
Along with banning porn and “objectionable” content, Apple will reject apps which use a non-WebKit rendering engine. “Chrome” and “Firefox” on iOS devices are actually just wrappers around Safari. ↩︎