Looking for Highly Intelligent Individuals
In January 2012, a plain black image appeared on a message board with a short greeting and a puzzle hidden inside. Thirteen years and three rounds later, the people who got the furthest still can't say who was on the other end.
At some point on January 4, 2012, a post went up on 4chan’s /x/ board — the corner of the internet reserved for paranormal talk and conspiracy theories — that had nothing to do with either. It was a black rectangle with a short block of white text:
Hello. We are looking for highly intelligent individuals. To find them, we have devised a test. There is a message hidden in this image. Find it, and it will lead you on the road to finding us. We look forward to meeting the few that will make it all the way through. Good luck.
It was signed “3301.” No context, no explanation, no link back to a company or a cause. Image boards get flooded with this kind of thing constantly — copypasta, in-jokes, half-serious cult recruitment posts that go nowhere. Almost nobody would have looked twice.
Except the image really did have a message hidden inside it, and finding it turned out to be the easy part.
A cicada in a JPEG
The people who bothered to check ran the image through a piece of steganography software called OutGuess, which can pull hidden data out of a JPEG without visibly changing the picture. Buried inside was a string of text and a set of numbers. The picture’s dimensions weren’t random either: 509 pixels by 503, both prime numbers. Multiply them together, multiply the result by 3301, and you get 845145127 — which, typed into a browser as a URL, resolved to a website.
That site displayed a countdown timer. When it expired, it revealed a new image of a cicada and a fresh string of text, this one pointing solvers toward a Caesar cipher. Decoding it led to a subreddit, and the subreddit led to something considerably stranger: a book cipher built on Liber AL vel Legis, “The Book of the Law,” a short occult text written by the British ceremonial magician Aleister Crowley in 1904. Solvers had to match location codes — chapter, line, character — against specific editions of Crowley’s book to extract the next piece of plaintext. Get one detail of the edition wrong, and the cipher fell apart.
That pattern repeated at every stage: a genuinely obscure piece of specialist knowledge, sitting exactly where a solver needed it. Over the following weeks, the puzzle pulled in prime number theory, Mayan numerology, medieval Welsh poetry, classical music theory, and enough applied cryptography that university computer science departments started quietly paying attention. This wasn’t a scavenger hunt built by an amateur with a book of riddles. Whoever put it together had range.
Flyers on telephone poles
The digital trail eventually spat out GPS coordinates — fourteen of them, scattered across the world, including Warsaw, Paris, Seoul, Sydney, Seattle, Miami, New Orleans, and Hawaii. At each location, someone had physically posted a paper flyer bearing the cicada image and a QR code. Local solvers, recruited through the growing online community around the puzzle, went out with cameras to photograph them and report back what the codes revealed.
That’s the detail that turned “clever internet riddle” into something else. Whoever was behind this had coordinated real people, in real cities, placing real paper on real telephone poles, timed to sync up with an online puzzle unfolding in parallel. It required money, planning, and either a genuinely international organization or an extremely dedicated small group with friends scattered across several continents.
Joel Eriksson, a Swedish cryptography and computer-security researcher then in his mid-thirties, found the original post shortly after it appeared and worked the entire chain alone. It took him just under three weeks — cryptography, steganography, Mayan numerology, a voicemail clue, and finally one of those city flyers — to reach what solvers started calling the end. Almost nobody else got there that fast, and almost nobody got there working solo. Most solvers who made real progress ended up pooling notes in real-time chatrooms, because no single person had the full range of skills the puzzle demanded.
That turned out to matter more than anyone collaborating in those chatrooms realized at the time.
The door that closed on the people knocking
When solvers finally reached the puzzle’s final stage, in place of triumph they found a message. Whoever ran Cicada 3301 said they were disappointed: too many people had gotten there not by solving the puzzle themselves but by riding along on the discoveries other, more capable solvers had posted publicly. The registration window for the “winners” closed, and it closed on people who had done the real work too — Eriksson among them.
“It was quite disappointing,” he told a reporter afterward, “especially considering that the people who registered in time were mostly ones that had not actually solved much of the puzzles themselves.” People had been sharing solutions and collaborating, he said, a bit too much for the puzzle’s taste.
The few who confirmed they’d gotten further described being funneled into private, invitation-only spaces — some hosted through Tor — where the conversation turned philosophical: online privacy, censorship, freedom of information, secure communication. According to messages several solvers received and later made public, whoever was behind 3301 described themselves only as “an international group” trying to build “a society based on ideals of liberty, privacy, and security.” Beyond that, almost nothing. No solver who reached the inner circle has ever produced hard proof of what, if anything, came next.
Marcus Wanner got a version of that experience a year later, and he was fifteen years old.
The kid in Copper Hill
Wanner lived in Copper Hill, a small community in rural Virginia, when Cicada 3301 returned on January 4, 2013, with a second puzzle. He and a handful of collaborators he’d never met in person — coordinating over IRC under the name #decipher — worked through a new gauntlet that included a custom Linux distribution nicknamed “Cicada OS,” another round of the Crowley book cipher, audio files hiding data in their waveforms, and another set of worldwide GPS coordinates. Wanner, without a driver’s license or independent transportation, had to talk a relative into detouring to a coordinate near his home so someone could photograph what was posted there.
He and his team were among the few who made it through that year — young enough, and far enough from any tech hub, that the story wrote itself when journalists came looking. Rolling Stone profiled him not long after; his university later put him on the cover of its alumni magazine as the student who’d cracked one of the internet’s most famous puzzles as a teenager, working the whole thing out of a bedroom with a dial-up-era sense of rural isolation and a laptop.
What Wanner and the others who solved the 2013 puzzle actually got at the end is, again, mostly secondhand and unconfirmed: private contact, a look at deeper material, the same vague language about privacy and secure communication — and then quiet. None of the credited solvers from any year has ever produced conclusive documentation of an ongoing relationship with whoever runs Cicada 3301, and the group itself has never publicly claimed any solver as a “member” of anything.
The book nobody can finish reading
The 2013 puzzle ended by pointing solvers toward something called the Liber Primus — “the first book” — a short excerpt of text written in an unknown alphabet of runic-looking symbols. It read, at the time, like a coda. It turned out to be a preview.
On January 4, 2014, Cicada 3301 returned with its third and, so far, final full round — and this time the entire puzzle was the Liber Primus itself: a roughly seventy-four-page manuscript, released as the puzzle’s culmination in May of that year, written almost entirely in a 29-symbol invented alphabet that the solving community named the Gematria Primus. A handful of early pages could be transliterated directly, symbol for letter, once solvers worked out the alphabet’s values. Everything after that got progressively harder — shifted ciphers, reversed alphabets, Vigenère-style keys layered on top of the runes, prime-number streams determining which symbols to read and which to skip, and stretches that nobody has cracked the method for at all.
More than a decade of sustained, worldwide, entirely volunteer effort — forums, wikis, shared spreadsheets, custom-built solving tools posted to code-sharing sites — has produced only partial progress. Something like a fifth of the manuscript has been convincingly decoded. The rest sits there, unread, its cipher method unknown, by design or by accident nobody can say.
The text that has been decoded doesn’t resolve into a manifesto or an instruction sheet. It reads more like philosophy — fragments about the nature of consciousness, self-knowledge, liberation through understanding. Whether that’s the whole point, or just cover for something more specific still locked inside the unsolved pages, is exactly the kind of question the Liber Primus was built to keep unanswered.
A warning, and then silence
Cicada 3301 kept a light, occasional presence online for a few more years — cryptographically signed messages, used to prove that a given post genuinely came from whoever had run the original puzzles, rather than from one of the many people now attempting to impersonate them. The last verified one arrived on April 4, 2017. It warned solvers to “beware false paths” and to check the signature on anything claiming to be an authentic Cicada communication before trusting it.
Nothing verified has come from them since.
That’s the detail that makes the story land differently than a typical unsolved-mystery writeup. Cicada 3301 spent five years building a reputation for rigor — real cryptography, real coordination across continents, real consequences for the people who solved it — specifically so that its signature would mean something. Then, having established exactly why solvers should trust only a verified message from “the real” Cicada, it stopped sending them. Every claimed sighting since 2017 is, by the group’s own stated rules, indistinguishable from a fake. If the people who built this ever come back, there is no longer any way for a solver to know it’s really them — and neither is there any way to know, at this point, whether the “them” who might come back are the same people who started it in 2012.
Who was ever on the other end
The obvious guesses arrived almost immediately and have never gone away: a covert recruitment pipeline for the NSA, the CIA, MI6, or a similar agency, using a skills test that would flag exactly the kind of independent, obsessive problem-solvers those organizations want and can’t easily find through job postings. No agency has ever confirmed anything of the sort, and the philosophical tone of the group’s own communications — heavy on privacy and liberty, light on anything resembling institutional language — sits awkwardly with that theory.
The alternative theories are just as unproven: a loose, well-funded collective of cryptographers and cypherpunks running an elaborate, deliberately pointless test of skill; an art project functioning as an alternate-reality game, except that no one has ever tried to sell tickets, merchandise, or a documentary tie-in off the back of it, which is unusual for anything built purely as a marketing exercise; something closer to a closed society that genuinely does recruit, just not for anyone official.
What’s not in dispute is the caliber of the thing. Steganography, layered cryptography, an invented writing system, physical logistics spanning at least four continents, and a decade of discipline about who gets access and why — all of it aimed at an audience that, by definition, would only ever be a few hundred people capable of following the whole chain. Nobody has ever monetized it. Nobody has ever taken credit for it under their real name and had that claim hold up. And the people who got the furthest, the ones journalists actually found and interviewed — a Swedish security researcher, a Virginia teenager — describe the same thing: a door that opened once, briefly, onto something they still can’t fully explain, and then closed.