Network email was born in 1971, when Ray Tomlinson sent a series of messages from one computer to another. “The test messages were entirely forgettable and I have, therefore, forgotten them,” he later wrote. The two terminals were sitting side by side, but had separate central processing units. The only connection between them was the ARPANET, the computer network that laid the foundation of the global internet. Tomlinson had adapted a program for leaving messages locally on a computer by combining it with an experimental file transfer protocol that he had developed for the network — “mostly because it seemed like a neat idea,” he said.
Looking back after 50 years, it’s clear that Tomlinson’s neat idea has proved strikingly resilient in the face of services like Slack, Facebook, Google Hangouts, WhatsApp, Snapchat, Skype, et al. Amid an internet landscape filled with “creative disruption,” email has managed to withstand the forces of rapid change that have rendered other technologies obsolete, failed, or co-opted.
This is largely because email is decentralized — its underlying protocols are neutral and universal. The fact that no single company or entity controls email has made the technology accessible to a vast number of people worldwide. Today, some four billion email users send and receive approximately 3.5 billion legitimate emails per hour, dwarfing social media usage. Revenues for all segments of the email market are estimated to total more than $46.8 billion in 2020 and grow to more than $84.2 billion by 2024.
Email has had such strong staying power because its decentralization spurs competition and variation. Far from being a technology on its last legs, as has repeatedly been forecast over the years, email should be seen as a model for other technologies. In fact, web-based social platforms and services should work more like email.
Email’s adversarial interoperability
The outward effect of email’s decentralization is what some call adversarial interoperability. “That’s when you create a new product or service that plugs into the existing ones without the permission of the companies that make them,” writes Cory Doctorow, special advisor to the Electronic Frontier Foundation. “Think of third-party printer ink, alternative app stores, or independent repair shops that use compatible parts from rival manufacturers to fix your car or your phone or your tractor.” Without adversarial interoperability, users have limited agency and innovation is stifled.
On its surface, email has changed a lot since 1971, but its infrastructure hasn’t changed much at all. Anyone on the network can send an email to anyone else without permission, with providers and relays distributing the messages worldwide. The simplicity of the design was devised before the existence of personal computers and email marketing, and its inefficiencies and lack of protections have led to serious problems, such as spam. A whopping 45 percent of emails were classified as spam in 2019. But this decentralized design has paradoxically helped email become more durable and innovative.
Over the years, as spammers grew more sophisticated and emails became larger in size, it became increasingly difficult to run email providers, which are responsible for storing and maintaining the messages that are sent to recipients, performing a service comparable to a digital answering machine. Google radically shifted the email market by offering storage capacities at orders of magnitude larger than anyone else for free, in exchange for personalized ads, keyword search, and tracking voluminous amounts of user data. In 2019, 1.5 billion Gmail users tolerated this surveillance, but thankfully, because email is decentralized, you don’t need a Google account to send an email to a Gmail user. You also don’t need Google’s permission to develop a competitive or complementary service based on something as simple as email forwarding.
With a centralized authority on a platform like Facebook, user interface innovations are either nonexistent or limited to whatever the company deems acceptable. Using legal tools such as provisions in the Digital Millennium Copyright Act, Facebook can basically deem certain competitive practices illegal — enforcing that only Facebook posts can be written or read on its application or website, for example. As every platform maintains it’s walled fiefdom, users are required to create separate accounts on Facebook, Instagram, Yelp, Uber, TikTok, etc., and none of these accounts work well with each other. Applying this same property to email would strike anyone as absurd. It would mean that in order to send an email to firstname.lastname@example.org you would need to sign up for an AOL account.
The underpinning architectural qualities of email provide for the adversarial interoperability that can help generate the productive competition everyone is calling for — and it’s possible for us to integrate these features into the platforms that dominate markets today.
The ability to freely send and read messages between email providers is a key distinction that stokes innovation. For example, Basecamp’s Hey is an email provider that claims to “fix” email with a better user experience and no ads in exchange for a $99 annual subscription fee. (Hey made headlines last year when Apple demanded a 30 percent cut of all proceeds as part of its in-app purchases policy.) Protonmail, an email provider that hosts data in Switzerland, boasts more than 10 million users for its secure, end-to-end encrypted platform (although it is only encrypted between other Protonmail users). These relatively small competitors wouldn’t be able to grow without the ability to interoperate with a giant like Gmail.
The stability of email’s underlying protocols makes it a low-risk technology. Compared to the Facebook API, which changes multiple times per year, email’s consistency is attractive to developers and companies. The same libraries and tools that large providers use are all open source and available for free, with implementations in multiple programming languages, providing a solid foundation that reassures developers that the ground won’t move out from under them. This makes it more lucrative to invest in integrated services like file sharing, scheduled sending, automated calendar synchronization, threads, contact synchronization, and suggested responses — knowing that these services can conceivably continue to work with minimal changes for another 50 years.
Big Tech’s big breakup?
Because proprietary digital platforms like Facebook and Uber aren’t incentivized to foster adversarial interoperability, it’s very challenging or nearly impossible for new services to compete or build off of their platforms. If an app somehow miraculously becomes competitive, the platform usually buys them. Last December, the Federal Trade Commission sued Facebook for illegal monopolization, specifying its acquisitions of Instagram and WhatsApp as well as anticompetitive conditions for developers on its platform. “This course of conduct harms competition, leaves consumers with few choices for personal social networking, and deprives advertisers of the benefits of competition,” the FTC said.
Indeed, despite Facebook owning Instagram for almost a decade, you still can’t see Facebook images on Instagram or vice versa. The US Department of Justice also filed a civil antitrust suit against Google last year, criticizing the company for “anticompetitive and exclusionary practices in the search and search advertising markets.”
A broader approach to data portability and interoperability could address a wide scope of competition challenges in the tech industry.
The prospect of US government regulation of web platforms and services has become a contentious topic over the past year. As the right calls for the repeal of Section 230 and the left calls for the deplatforming of violent extremists and misinformation campaigns, the future of the social web as we know it today seems unclear. Previous laws have already shaped this dynamic; for example, US export law requires most web platforms to block users connecting from Crimea, Iran, North Korea, Syria, and Cuba. But it’s unclear which platforms comply with the law. Iranian users were banned from Slack in 2018, Amazon Web Services and GitHub in 2019, and GitLab in 2020. Edward Snowden suggested that Amazon, “which runs basically half the internet on their cloud platform,” had overstepped its bounds and defeated the original purpose of the law by “cutting off the lifeline of Iran’s liberal opposition.” GitHub was able to secure a license to offer services to developers in Iran in January 2021, but not all companies have gone through this process.
This sets a dangerous precedent where knowledge can disappear or be inaccessible permanently and without warning. The resulting power dynamic creates information security vulnerabilities, and is especially dangerous for organizations with sensitive or mission-critical information. Needless to say, web services are a powerful force that can shape politics and economics around the globe. Even the slightest updates to regulatory law in the US, where many of these platforms are based, can result in profound global reverberations that are difficult to predict.
Despite these unknowns, think tanks, academics, and the EU have released reports that explore breaking up corporate platform-empires. “These firms have too much power, and that power must be reined in and subject to appropriate oversight and enforcement. Our economy and democracy are at stake,” declared a recent congressional report on competition in the tech industry. “We’ll be pressing Congress to update our antitrust laws to address more of the harms of Big Tech monopolies,” wrote the Electronic Frontier Foundation in response to the Google antitrust suit.
The internet’s commercialization has left us at the whim of large corporations that are often unaccountable and indifferent.
As calls to update antitrust laws are made, the practicality of doing so remains unclear. Compared to a typical US-based social platform company like Facebook, users in an interoperable and decentralized network like email are able to choose a provider or client that’s based in any legal jurisdiction that suits them best. This borderless interaction may affect the global impact of any US legal action on particular companies, but it’s a quality that has significantly contributed to email’s stability over time. On the other hand, reassessing laws like the Computer Fraud and Abuse Act and Digital Millennium Copyright Act could help open the field to more competition.
“We have created these legal tools that allow firms not just to dominate but to maintain their dominance not by making better products but by establishing a kill zone around their products, where it becomes so legally fraught to make compatible products that nobody even tries,” says Doctorow.
We can apply the lessons learned from email’s ability to flourish over the past 50 years to the next generation of the social web. The underpinning architectural qualities of email provide for the adversarial interoperability that can help generate the productive competition everyone is calling for — and it’s possible for us to integrate these features into the platforms that dominate markets today. “We need to regulate Facebook and others to force them to open up their APIs to make it possible for users to have access to each other across platforms and services,” writes data specialist Irina Bolychevsky, a board member of the Open Knowledge Foundation.
It’s worth noting that Facebook and Google both previously integrated a decentralized and interoperable protocol (XMPP) for their respective chat features, allowing users to communicate across services in a manner similar to email. But once Facebook’s market dominance was complete, it killed this feature in April 2014, forcing everyone on the service to use the Facebook messenger web interface or app.
Because email is globally accessible, stable, and interoperable across borders, it has been increasingly used to verify online accounts, devices, and social interactions. Similarly to how HTTP is the digital infrastructure for websites, email has become the digital infrastructure for social identity. What would it look like if social platforms were required to integrate with an interoperable social infrastructure, or even used email itself as this standard? We could imagine new interfaces that mix and match social messages, ride-hailing, room rentals, or classifieds. The wealth of all of our social interactions would be multiplied and combined across platforms, resulting in a better experience for everyone.