The first spam email. The first MP3. The first recorded appearance of "LOL." The AOL dial-up sound in all its staticky glory. The Hampster Dance. Friendster. All relics of a nascent, more freewheeling internet, long since buried in the digital graveyard. Or were, anyway. Because you can now experience them all firsthand, thanks to one of the web's most original minds.
Internet Artifacts is the latest project from Neal Agarwal, the creative 25-year-old coder who launched neal.fun six years ago today. You may not recognize the name, but there's a good chance you've encountered one of Agarwal's viral creations. They include The Deep Sea, a mind-bending data visualization of the ocean's depths; Asteroid Launcher, in which you can hurtle mile-wide space rocks into your hometown; the self-explanatory Absurd Trolley Problems; and the sadistically addictive Password Game, along with a few dozen others.
Agarwal's work is a joyful call to arms. He has spent his adult life infusing the web with a throwback spirit of discovery, defying preconceptions of what a website can and should be. The internet is boring. Social media is dead. Being online today is doom-coded, it's gloom-pilled, it's a wasteland of algorithmically attuned sludgery. You may encounter sparks of viral fun — you'll find no Skibidi toilet erasure here — but very little joy.
It wasn't always so. The web used to be so much weirder.
"I grew up at the tail end of that era of the internet," Agarwal said. "There were just a lot of independent creators making cool stuff and it was clear that they were doing it for the fun of it, not to grow a business or anything like that. And then I kind of watched that all slowly die."
But within the corridors of Internet Artifacts, and through the work of Agarwal and a host of other creators dedicated to the web as a medium, you can see an urgent attempt at resurrection, echoes of an internet that was messy and unpredictable — and all the better for it.
How did we wind up with today's anemic internet? The pocket version goes something like this: A couple of decades ago, give or take, easy-to-use web-design software like Adobe Flash, combined with mainstream internet availability, catalyzed a golden era of online experimentation. Animated series like Homestar Runner, nihilistic absurdities like QWOP, and single-serving meme shrines like YTMND dotted the landscape. They spread not by algorithmic ranking but by word of mouth. They had little in common beyond an antic sense of discovery and childlike play. After all, the web itself was just a kid.
Over a few short years, two concurrent tectonic shifts would reshape the internet. The first is that Apple killed Flash; Steve Jobs exiled it from the iPhone for being a buggy, sluggish, proprietary security risk — he wasn't wrong — which made the once-ubiquitous software a liability to work with. Suddenly building things on the web wasn't quite so simple, and eventually a lot of cherished sites no longer worked. Tools like HTML5 came along to take Flash's place, but by then it was too late. Developers increasingly built for the iPhone, prioritizing apps over browsers.
The other transformation is that websites themselves became largely obsolete, displaced by platforms. First came the likes of MySpace, Facebook, Twitter, Tumblr, and Reddit; then came the image- and video-focused wave of Instagram, Snap, and TikTok. To their credit, those tech giants did democratize the act of posting. You didn't need to be a software engineer to put a photo on your timeline. With powerful mobile editing tools, the act of putting stuff online is now easier than ever. But the result is a little like giving a room full of artists the same size canvas and identical matte paints. There are only so many variations one can bring to the theme.
"Now in lieu of the unique URL, you have the unique handle," Jamie Cohen, a digital culture writer and an assistant professor at CUNY-Queens, said. People once created their own distinct worlds; now they mostly shout at each other in the same rooms. "It's not the same. It's a flattening experience."
Neal Agarwal rejects that looming sameness.
"When I was just a kid, the internet felt like a Wild West, at least more than it is now," Agarwal said. "I would always go down these long rabbit holes. Almost all the sites I visited were by solo creators or small teams of people. It felt like much more of an independent web. As I watched all that go away, I kept feeling that this probably isn't how it should be. There should be more people creating fun stuff on the web."
And so he did.
Unlike some origin stories, there's no single spark that ignited Agarwal. The first website he built, Kidcrash, was itself an act of curation, putting all his favorite Flash-based games in one place. He made it in 2006. He was 9. Like a lot of precocious young coders at that time, the Fairfax, Virginia, native soon found his way to Scratch, a programming website for kids, where he went by the handle awesomestickdude. His first game was a knockoff of the reality-TV competition show "Wipeout." Then 12, he offered to build a new game level for every 15 "love-its" he got from the community, not anticipating that he'd eventually receive 1,167. He bailed after level 10.
In high school, Agarwal made a single mobile app, a game called Toast Man, before pivoting back to the web. He set up shop with a site called Kamogo, where he hosted projects with the same inventiveness and absurdist energy — albeit less polish, and often not much of a point — that would later animate neal.fun. Much of it still holds up. There's a Silicon Valley Idea Generator, a tool for making vertical videos horizontal, and a "Text to Hodor" translator that turns whatever you type into, well, "hodor."
"There's a value to having somewhere on the internet, something that says, 'This was important to a lot of people. This existed. It mattered.'"
In 2016, Agarwal got his first glimpse of mainstream success when Insider wrote up everysecond.io, an early data visualization. But it wasn't until he was enrolled at Virginia Tech that Agarwal committed himself more fully to making his so-called web toys.
"Programming for me was always a creative expression," he said. "It's like creative writing. I program because I want to have the ideas in my head become a reality." Formal computer-science studies felt like the exact opposite. When professors required students to use certain programming languages, Agarwal would often go his own way regardless — nearly failing a few assignments in the process.
But the abject boredom Agarwal felt during lectures had a silver lining: the creation of neal.fun, which he worked on while he should have been taking notes. One of his first entries, Spend Bill Gates' Money, remains his most popular, with over 80 million page views since it launched in 2017. It took three days to make.
By the time he graduated, Agarwal was making something like a full-time living off ads on neal.fun. He gave himself a year to see if it could actually become a career. Two breakout visualizations — The Deep Sea and The Size of Space — suggested that it could, largely because he licensed the former to Ripley's Aquarium. After a detour working at MSCHF, a collective known for large-scale pranks and subversive art projects, Agarwal decided to go it alone for good.
"He is very, very impressive," said Josh Wardle, the creator of Wordle (which he sold to The New York Times for more than $1 million last year), and who overlapped with Agarwal at MSCHF. "There is a joy to his work, and a levity, which I think reflects how Neal sees the world."
Virality has seemingly come easy in the years since, including most recently with Password Game, which has amassed over 10 million page views since it launched in June. It's wholly original but also quintessentially neal.fun, taking a mundane part of internet life — password policies — to an inconceivable extreme.
The creative process varies from project to project. Password Game started as a germ of an idea five years ago but took a few weeks to build once he got going. Internet Artifacts has taken closer to three months. He does his own research but has been increasingly open to soliciting outside feedback before going live; he estimates 20 people have given input on his internet museum.
Internet Artifacts takes several touchstones of the anteplatformian internet and places them on literal digital pedestals. Homestar Runner makes the cut, as does Zombo, a joke web page that was once anointed "the least useful website." Agarwal has lovingly recreated his references, down to a faithfully emulated Internet Explorer browser window for entries on the original MySpace and Wikipedia pages, and a Napster simulation that lets you download real MP3s. The exhibit unfolds chronologically, starting with a 1977 map of ARPANET, the government-funded precursor to the modern internet, and making dozens of stops across the next three decades.
Think of neal.fun as a splintered reality, an alternate dimension where the web never stopped being weird.
The most surprising thing about Agarwal's undertaking may be that there's nothing really like it out there. (The closest thing may be Brandon Chilcutt's indispensable Museum of Endangered Sounds, which focuses on aural artifacts both on and offline: the whirring clicks of a rotary phone, a Tamagotchi's plaintive bloops.) "There's a value to having somewhere on the internet, something that says, 'This was important to a lot of people. This existed. It mattered,'" said Jessa Lingel, the author of "The Gentrification of the Internet" and an assistant professor at the University of Pennsylvania's Annenberg School for Communication. "It really shouldn't just be internet studies scholars who do that."
As delightful as Internet Artifacts is to click through, it also provides valuable context for Agarwal's larger ambition. Think of neal.fun as a splintered reality, an alternate dimension where the web never stopped being weird.
Agarwal is hardly the only person making fun things online. He meets regularly with a group of six kindred spirits in New York to compare notes and test out ideas. You can find complementary energy in the work of Nicole He, Matthew Rayfield, and Brian Moore. Immaculate Grid and Wordle (along with its infinite imitators) provide a daily fix outside the confines of the major platforms. The Useless Web curates a dazzling selection of silliness. Even Homestar Runner has mounted a comeback, in the form of occasional YouTube videos.
But it's the variety of Agarwal's work, and his dedication to the web as a medium for artistic expression, that stands out.
"As the internet develops, there's a bunch of things that tech companies want to do around apps and push notifications. Neal had this focus on not doing any of that," Wardle said of their time as coworkers. "It made me excited about the web."
As for what's next, Agarwal has maintained a list of potential experiments, sorted by a five-star rating system that helps him prioritize his road map, since he was 18. The list now numbers somewhere around 1,300. He still adds one or two each day. "At this point, even if I don't think of any more ideas, I have enough things to work on probably for the rest of my life," he said.
Internet Artifacts goes up to 2007. The last entry is a video of Steve Jobs announcing the iPhone at that year's MacWorld conference, played in an emulated QuickTime window. It's a fitting capstone for a certain era of the web, one that was open and weird and absurd. You're left with a sense of nostalgia, but also loss.
There's probably no going back to the web that was. Apps and platforms have fragmented but still dominate and will continue to lure this generation of creatives with the promise of monetization. But Agarwal's work suggests if not a return to the old ways then at least an alternate path forward. The web can be fun. It can be silly. It can be joyful. It can be anything you want. And isn't that the point? An infinite canvas, forever in search of an artist.
Brian Barrett is the former executive editor of Wired and author of The Art of the Minifigure. He has also written for The New Yorker, The Atlantic, New York, and numerous other publications.