To a certain kind of nerd, Ev Williams is the Forrest Gump of internet media. Williams helped write the software that made us call blogs blogs. He founded a podcast company years before most people listened to them. He sent Twitter’s 75th tweet, then ran the company. And now he’s the founder and CEO of Medium, the platform for online writing embraced by sportswriters, Silicon Valley executives, and the President of the United States.

Of all the American internet industry’s critical events (other than that fateful night in Mark Zuckerberg’s dorm room), odds are good that Williams was there or knew someone present. So at 9 a.m. on a Tuesday in March, as he sweeps down the stairs into the BART station at 16th and Mission—into the fast-transit artery of the country’s most technologized region—you might expect someone to recognize him. But as he gets on a downtown train, no one turns a head.

Despite serving as a board member at one of the five largest social networks, and a mainstay of the Bay Area tech industry for almost two decades, the kind of fame attached to the names of Mark Zuckerberg, Peter Thiel, or the “Google guys” has eluded Williams. He’s maybe even sub-Travis Kalanick now.

Yet his run near the top has been remarkably consistent. While other CEOs in his early-web cohort have left the industry, or have become writers or consultants, Williams has stuck around, leading companies. His startups have nearly all specialized in the same abstract medium: text boxes. He has dotted the web with these text boxes, and people have poured their souls into them, have argued and wept and whispered into them. Millions of people have had their worldview shaped by these text boxes, and the boxes themselves have, in turn, changed the Internet. They have also made Williams rich. Though few of his businesses have turned a profit, he is a billionaire.

I met him in a cafe on Valencia Street, an old punk and immigrant district in San Francisco now lined with spartan boutiques, ethical taxidermy shops, and other beacons of mass-appeal hipsterism.

Williams looks the tech-CEO part. He is tall, soft-spoken, with a constant air of chilled-out concern. His gray hoodie and black t-shirt are woven from some athleisure Star Fleet-issue textile, and he wears broad, squarish white glasses that I internally dub the Warby Mugatu. Within minutes of arriving, he has launched back into his endless theme, which he expands on across multiple meetings, on two different coasts, across three months: “The open web,” he says, “is pretty broken.” But don’t worry—he has a plan to save it, or, at least, sort of save it. And it involves text boxes.

The open web is the nickname for the internet as it should be—free, uncensorable, and independently owned and operated. According to the blog posts that hashed out most of its theory (and which themselves were published on the open web), the open web describes an internet where people mostly publish their writing (or music, or photos, or films) to servers that they own or rent, accessible via their own personal domain names, in formats that are themselves free or unrestricted. It is the web because the pages are written in HTML and CSS; it is open because anyone can access almost all of it, without special privileges, expenditures, or a user account. Above all, the open web is free—free like language is free, like consciousness is free. Freedom not so much as a right, but as a technical and inalienable fact.

This liberty has an end goal: to turn the web into the finest, coolest piece of media ever created, a library of libraries authored by all of humanity. This web encompasses novels and newspapers and scientific journals, all at once. Anyone can write for it, and anyone can read it. It is a to-do list, a logbook, a work of literature, and a communication tool so powerful that it could abort war.

This is a vision for the web that sounds both very similar to and very distant from the web that you and I use everyday. Our web, after all, contains unhappy news, garish advertising, unsympathetic grandstanding, and a lot of photos of other people’s kids. All this clutter reaches us after being shunted through social networks, which (the idealists lament) are effectively shut off from the rest of the network. The follow-on effect from these networks is even worse: Cookies tied to those same user accounts surveil us as we read across the open web, then a mysterious algorithm uses this collected browsing history to decide how to distract us with ads. The open web is pretty broken indeed, and this isn’t even getting into spam, mass harassment, identity theft, and digital espionage.

“There’s still a bunch of stuff on the web. The stuff we read everyday, the stuff you write, is on the web. And that’s great,” says Williams. (In fact, you are reading this very story on the open web—unless you found it on the Facebook app on your phone, in which case you are reading a copy nearly identical to the open-web version of the story, except that yours loaded much faster and lives on Facebook’s servers.) “There’s still the fact that anyone, at any time, can create their own website and start publishing, and they have a voice—I mean that’s the idea that I got really excited about almost 20 years ago.”

“I think that will continue. I think the openness of voices is not going to consolidate back to the old days of media,” he told me. “I think the distribution points are going to consolidate.”

The distribution points are the search engines and the social networks: Facebook, Google, Twitter, Snapchat, and the messaging apps. Also on that list are YouTube (owned by Google), Instagram (owned by Facebook), Whatsapp (also owned by Facebook), and Facebook Messenger (ditto). By linking the web together, or hosting normally data-heavy content for free, these distribution nodes seize more and more users. And because each of the nodes is more interesting than any one individual’s personal site, people who used to go to personal sites wind up at the nodes instead.

As Williams puts it: “Primarily what we’ve seen is that the social networks have gotten really, really big, and they drive more and more of our attention.” With this size, they also collect more revenue: 85 cents of every new dollar in online advertising went to Google or Facebook in early 2016, according to a Morgan Stanley analyst quoted by The New York Times.

“That could be bad,” says Williams, in his low-key way.

The open web’s terminal illness is not a story that he alone is telling. It is the common wisdom of the moment, espoused by Times columnists and longtime tech bloggers. The developers who wrote Drupal and Wordpress, two important pieces of blogging software, both recently expressed anxiety over the open web’s future. Since so many of these social networks are operated by algorithms, whose machinations are proprietary knowledge, they worry that people are losing any control over what they see when they log on. The once-polyphonic blogosphere, they say, will turn into the web of mass-manufactured schlock.

Something like this has happened before. Tim Wu, a law professor at Columbia University, argues in his book The Master Switch that every major telecommunications technology has followed the same pattern: a brief, thrilling period of openness, followed by a monopolistic and increasingly atrophied closedness. Without government intervention, the same fate will befall the internet, he says. Williams cites Wu frequently. “Railroad, electricity, cable, telephone—all followed this similar pattern toward closedness and monopoly, and government regulated or not, it tends to happen because of the power of network effects and the economies of scale,” he told me.

Williams and his team at Medium say they are working to resist this consolidation, though they are not doing quite what anyone else would recognize as resistance. The truth is that they themselves want to consolidate some of the web, too; and then—with that task done—govern as just, beloved, and benevolent despots. Josh Benton, a media critic at Harvard, once described Medium as “YouTube for prose,” and that’s an apt summary of what it feels like to use. But as I spend more time with Ev, I catch him thinking of Medium as a project philosophically akin to the “Foundation” novels by Isaac Asimov. The heroes of those books sought to centralize all the learning across the galaxy before a dark age set in, knowing that though they cannot stop the shadowed era, they may be able to preserve scholarship and therefore shorten it. Ev’s ambitions, though not as grandiose, follow similar lines. Medium seeks to replicate the web’s old, chaotic hubbub on a single, ordered site—because, ultimately, Ev values the chaos.

Alex Fine

In the spring of 2000, a developer and designer in San Francisco named Meg Hourihan was surveying the city’s swelling ranks. New coders had come from all around the world to her city, to work on internet projects, and they were crowding her favorite haunts. She loved the web, and she was excited to see it catch on with a broader public—but she could not get excited about the hordes descending on the city.

“I realized there are dot-com people and there are web people,” she wrote on her blog at the time. “Dot-com people work for start-ups injected with large Silicon Valley coin, they have options, they talk options, they dream options. They have IPOs. They’re richer after four months of ‘web’ work than many web people who’ve been doing it since the beginning. They don’t have personal sites. … They don’t get personal.”

She continued. “Web people can tell you the first site they ever saw, they can tell you the moment they knew: This, This Is It, I Will Do This. And they pour themselves into the web, with stories, with designs, with pictures. They create things worth looking at, worth reading, worth coveting, worth envying, worth loving.”

At the time, Hourihan was co-founder of a small company named Pyra Labs. Her co-founder was Williams. They were both web people.

Born in 1972, Williams grew up on a farm about 90 minutes from Lincoln, Nebraska. For a long time, he didn’t stray far. He stayed in-state for school, going to the University of Nebraska. But sensing the internet’s enormous potential, he dropped out, preferring to try his luck with tech ventures funded by his parents’ money. One of his companies sold a CD-ROM with information about that year’s Cornhuskers team. Another distributed a video about how to connect to the internet.

But by the time he was 24, Williams realized he would have to leave the plains to work on the net. He moved to Sebastopol, California, to work at O’Reilly Media. O’Reilly publishes dead-tree books—programming manuals and standards guides—that held biblical importance to ’90s coders. “When viewed from Nebraska, Sebastopol looks like it’s in exactly the same spot as San Francisco,” he would write later. “In actuality, it’s about an hour away and feels like a very different place.”

He stayed there for several years nonetheless. But he was right about San Francisco. It was there that he met Hourihan. They discovered their mutual admiration for the web, briefly dated, and ultimately founded Pyra Labs together in 1999. Pyra never actually shipped its namesake software, a suite of office collaboration tools, but in the offing it managed to build Blogger, the first simple web-journaling software to find a massive user base. Blogger also helped popularize the word blog.

Williams and Hourihan had tremendously unlucky timing. Blogger got big just as the first dot-com bubble popped. The company wasn’t expensive to run, but with VCs going bankrupt right and left, no one could find the money to fund it. It missed payrolls. Its leaders fought about the right path. It laid off employees. In January 2001, Hourihan resigned, and everyone else at the company walked out. (Hourihan later founded Kinja, Gawker Media’s blogging software.)

Yet Pyra didn’t die. Williams kept it alive by knocking out small contracts to keep the corporate name afloat, while finishing long-planned product updates. Two years after the bubble’s collapse, he shipped a premium version of Blogger that cost money to use. He hired a few more staff. In February 2003, Google bought Pyra. “We had a million registered users,” Williams says now. “And that felt big.”

It’s worth dwelling on this moment. Blogger’s story contains all the contradictions that would eventually dissolve the open web. For all the talk of their radical openness, blogs had mostly been the domain of those with hosting space, programming experience, and the time to write them. The blogosphere was dense and complicated, with many writers posting dozens of times per day; those who had the power to blog (like Andrew Sullivan and Stereogum’s Scott Lapatine) could shape conversations in politics, culture, and music.

Blogger’s great innovation was to supply writers with an easy interface and a free domain name, blogspot.com, where they could host their journals. This latter feature fueled the site’s growth. It allowed blogging to graduate from the dominion of a tech-savvy elite to something that anyone with a computer and web connection could do—and more people than ever, motivated by the national-security anxiety and intense politics of the 2000s, were eager to take part.

But even in the blogosphere’s early days, growth was synonymous with consolidation. Expanding the web’s power to more people also centralized it—there was no difference between the two. It foreshadowed what was to come.

Williams stayed at Google for six months before moving on. In the fall of 2004, he co-founded Odeo, an early podcasting company. Odeo wanted to be to podcasts what Blogger was to blogs, but internet audio was still too disorganized for a business to succeed.

“The entire idea of podcasts came from realizing you could do a hack to pull stuff down from the Internet on your computer and put it on your iPod,” he told me when we met in New York. “And that was cool, but it was a pain in the butt.”

By early 2006, some of Odeo’s employees began playing around with a software doodad they had developed. It was a digital megaphone, basically: If you sent it a short SMS text, then it would broadcast that message to all of your friends. That product, separate from the core podcast offering, debuted in March and formally launched in July. By December, it had more than 60,000 users. By February 2007, Odeo had rebranded itself Twitter.

The next month, a subset of technology and media elite glommed onto Twitter at the South by Southwest technology conference in Austin. They loved it, they blogged about it, they started dropping daily witticisms and one-liners there—tweeting when they had written a new blog post, for instance—and the service exploded. By April 2007, Twitter had 8 million users. It had grown more than 13-fold in five months.

This period—the fall of 2006 to the spring of 2007—was the most heated the aughts ever got in Silicon Valley. In this period, Google acquired YouTube, an 18-month-old company, for $1.6 billion. Facebook opened to all users, not just college students. TIME declared “You” the Person of the Year, a silly gimmick that nonetheless initiated the era of social-media hype. And Apple debuted the first iPhone.

In this environment, Twitter was growing explosively, though under the aimless leadership of Jack Dorsey. In 2008, Williams was named Twitter CEO.

Even the internet of 2008 can seem distant. That year’s presidential election was famously waged via web blogs. By 2012, much of the conversation had moved to Twitter. Speaking now, Williams sounds contrite about this centralization, from many news sites and blogs to a single platform. “In general, structurally, it’s probably bad if all our media and communications are going through services that are controlled by profit-driven corporations,” he says. (A similar sentiment sparked the creation of public broadcast media in the 1970s.)

The dangers of corporate consolidation dominate his metaphors. A favorite idea is that the web’s current state resembles the factory-farmed food system. “If your job was to feed people, but you were only measured by the efficiency of calories delivered, you may learn over time that high-calorie, high-processed foods were the most efficient ways to deliver calories,” he says. They would be the most margin-friendly way to deliver calories. But the food still wouldn’t be good—because the original metric didn’t take into account “sustainability, or health, or nourishment, or happiness of the people.”

I proposed that Medium is trying to be the Whole Foods of content. He laughed.

“Maybe we are,” he said. “Not that Whole Foods is perfect, and we’re not perfect either, but we are trying to figure out how to optimize for satisfaction and nourishment, not just activity or calories.”

Williams and his team have devised alternate metrics to account for those more holistic virtues—namely, “time spent reading,” which measures how long Medium users collectively spent reading a story. And instead of garish display ads, much of its revenue (right now) takes the form of native advertising, or brand sponsorship of certain series.

And Medium’s marketing position isn’t far from Whole Foods either—it wants to be the big corporation that upscale customers trust. For even though Williams may express suspicion of the big profit-driven networks, Medium vies to join them. Weeks after we met, the company debuted a tool to suck up WordPress blogs and drop them into Medium. Publications that would have previously lived at their own domain name—like The Awl, Pacific Standard, and Bill Simmons’ new site, The Ringer—now live exclusively on Medium. (The Toast also considered moving to Medium, but chose to shut down for other reasons.) Each of these sites still lives on its own domain name, but in terms of design and function, each is essentially a Medium page. Their stories also live on Medium’s servers.

Alex Fine

While he was CEO of Twitter, Williams spoke to a small product team about what the social network needed to become. The Internet was transitioning from the web of archipelagos to the web of continents, he said. The archipelagos—think of email and the blogosphere—constituted many small, independently owned atolls that could communicate. But their disjointedness also made them nearly impossible to update.

A new form of organization was supplanting the archipelagos, he said: the web of continents. Facebook was a great continent, of course, but so was any other site that absorbed its users into a great centralized morass. If Twitter hoped to survive, it needed to do more than serve and connect the archipelagos—it had to become continental.

Williams did not quite take it there. His tenure at Twitter was marked by fast growth, but the company never found its business footing. In 2010, he stepped down as CEO, though he remained on its board. Two years later, he founded Medium, describing it as a place for content that was too short for Blogger and too long for Twitter. The next year, in the autumn of 2013, Twitter had its initial public offering. Williams’s 12-percent stake in the company made him a multi-billionaire.

Which is funny, because talking to Williams, you get the sense that—to paraphrase the joke about Obama—if things had really worked out for him, he could have been a journalist. What seems to excite him most about Medium or any of his other ventures are that he helped give voice to the afflicted. He remembered “I, Racist,” a Medium post adapted from a sermon by John Metta. It found tens of thousands of readers on his site.

“He, as a person, just had the right way to say something,” he said of Metta. “And he wasn’t necessarily someone who said I’m gonna be a publisher, I’m gonna start a blog, I’m gonna have a voice. We gave him a canvas, what was in his brain came out, and it found people who needed to see it. That’s a better world, when that’s happening all the time.”

Williams still comes off like a cheerleader for this better world. He told me that a Medium user wrote an open letter to him, saying that though they had posted to the site every day for a month, they had not gotten more than 100 “recommends” on their post yet. (Every social network has its atomic unit of dopamine-like recognition: Facebook has likes, Twitter has hearts, Medium has the recommend.) He said he wanted to reply and tell the guy to step back.

“Think about what you’re doing,” he says. “You’re playing this game for attention that half of humanity is playing. And you’re competing for not only the thousands of people who publish on Medium the same day, the millions of people who publish on websites that have ever published, the billion videos on YouTube, every book in the world, not to mention what’s on Instagram, Facebook, Twitter, Vine, everything else, right now—it’s amazing any people are reading your stuff!”

That this can still happen—that any subset of readers can still find and read an amateur writer’s work—is what excites him most about Medium. Talking about the centralization of the web, he continually returns to the “bad world.”

“The worst world, the scary version, is if the tricks to get attention are a skill developed and owned primarily by profit-driven companies,” he told me. “I’d go back to the food analogy. What are people going to be consuming most of the time? They’re optimizing for clicks and dollars. Can a person who has a unique perspective play that game? Are they just going to get trounced?”

This is Medium’s reason for existing: to protect individual writers in the fierce and nasty content jungles. Resistance to the centralization generally is futile, he believes, citing Wu. “That’s the way the Internet works, and that’s the way humans work,” he says. “Efficiency and ROI and economies of scale and user experience—they’re all going to drive more things to consolidate. I kind of look at that as a force of nature. But if things consolidate, does that mean that everything is shit?”

That is the Medium appeal, in a nutshell. Keeping everything from being shit. It wants to do so by adopting many of the tics and habits of the original blogosphere—the intertextuality, the back-and-forth, the sense of amateurism—without being the open web. It will use its own custom metrics, like time-spent-reading, to decide who sees what stories; and it will tend to show your friends something if you “recommend” it. Medium, yes, will just be another platform, but it will run the open web in an emulator.

“I understand the skepticism, that we’re a venture-backed corporation that is saying those things,” Williams says. “I think you can still be optimistic that something good can be created, and you can at least get behind the fact that there shouldn’t be one platform that everything centralizes to.”

And by one platform, he means Facebook.

Facebook. Of course it would end with Facebook. The web people have always been suspicious of it. As early as 2007, early bloggers like Jason Kottke called Facebook “a step sideways or even backwards” for the web. They compared it to AOL, another platform that intended to centralize the net before it flamed out. (A year earlier, Kottke had married Hourihan.)

Except Facebook has succeeded where AOL failed. An April report from the web-analytics company Parse.ly found that Google and Facebook, just two companies, send more than 80 percent of all traffic to news sites. (No wonder they make 85 cents of every digital-ad dollar.) And since few people today use RSS readers like Feedly or visit homepages directly, publications like The Atlantic essentially depend on Facebook and Google to send them their regular readers. Forget that thriving blogosphere: If their authors didn’t move to social media years ago, then their readers did. The web of 2008—the web that helped elect President Obama—has already withered.

All of this can make Williams’s memories of the web sound elegiac. I once met Williams in a hotel lobby in midtown Manhattan, early in the morning. We looked out across Columbus Circle and the late autumn ruddiness of Central Park. In a freak of city planning, Trump Tower was the only thing obstructing our view.

Williams’ flight the day before had been exhilarating, in a comfortably ordinary way, he told me. “I was having one of those rare moments where you appreciate that you live in the future,” he said. “From having called my Uber to get there, to having my boarding pass in my iPhone wallet and scanning it.”

“And something about that, like—everything worked! And that was amazing. And that’s an everyday occurrence, and there’s wifi in the airport, and I can use my phone and laptop the entire time, and there’s wifi on the plane. That was our dreamed-of future.”

And it was. But the thing about dreaming up a future, and making it real, is then you have to live in it. Back in San Francisco, coming out of the BART station on Market Street, he admits that the web game has changed since he came up. His glasses have been switched out for ruby sunglasses, polarized, reflective, and movie-star dashing.

“There were always ecommerce startups,” he says. “I was never part of that world, and we kind of looked down on them when the whole boom was happening. We were creating businesses, but ours had more creativity, ours weren’t just for the money. Or maybe ours were even for utility but not just money, whereas clearly there are ways for both.”

He laughs. “Even the Google guys—they were trying to create something really useful and good for the world, and they made all the money.”

Now the internet works differently, he says: “It’s in general no longer about the creativity, it’s about the business.”