Juan
Benet

The Man Who Built the Library of Babel - and Then Asked What Comes Next

He turned down Yahoo. That's the sentence that explains everything. In May 2013, Yahoo acquired Loki Studios - Juan Benet's location-based gaming startup - in one of those acqui-hires where the company buys the team and quietly shutters the product. Geomon, the creature-collection game that Benet and his Stanford classmates had spent two years building, was killed. The users were left. Benet walked away from a secure role at one of the world's largest internet companies and went back to his apartment to rethink the fundamental architecture of how the web stores information.

Three years later, Geomon's concept became Pokemon Go, a global phenomenon. By then, Benet was already presenting IPFS - the InterPlanetary File System - at conferences, explaining why HTTP was the wrong design choice for a durable, open internet. The irony was not lost on observers. Yahoo's $4 billion mistake, one headline called it. Benet never said a word about it publicly.

That silence is instructive. Benet is the rare founder whose ambition is architectural rather than personal. He is not building a product with his name on it - he is building the pipes that other builders use. The protocols he has created (IPFS, Filecoin, libp2p, IPLD, Multiformats) are the kind of infrastructure that disappears into the internet the way TCP/IP disappeared - present everywhere, attributed to no one at the moment of use. The fact that libp2p, his modular peer-to-peer networking stack, now powers Ethereum 2.0 is not something he promotes. It's just what good infrastructure does.

Six Protocols. One Idea. A Different Internet.

IPFS stands for InterPlanetary File System, a name that signals the ambition plainly. Where HTTP asks "where is this file?" (on which server, at which URL), IPFS asks "what is this file?" - identifying content by a cryptographic hash of its contents rather than its location. The practical consequence: files become permanent. A webpage stored on IPFS cannot be taken down by shutting off a server. It cannot disappear when a company goes bankrupt. It cannot be censored by blocking a domain. The content exists as long as any node in the network holds a copy.

The January 2015 initial release of IPFS was quiet, almost accidental in its reach. TechCrunch called it "quickly spreading by word of mouth" by October of the same year. No ad budget. No product hunt launch countdown. Just a whitepaper, a GitHub repo, and engineers who recognized a good idea when they read one. That is how Benet's work has always traveled - through credibility, not marketing.

Decades, Not Quarters

Benet gave a talk at the Long Now Foundation in 2018 titled "Long-term Info-structure." The Long Now Foundation runs on centuries-long thinking - it maintains a 10,000-year clock - and Benet fit right in. His argument was simple: the infrastructure humanity builds for information storage and transmission will either be open, durable, and owned by no one - or it will be owned by corporations whose incentives will not, over the course of a century, align with humanity's needs.

This is not a fringe position in the history of the internet. It is the argument Tim Berners-Lee made when he gave HTTP away for free. It is the argument Vint Cerf made when he designed TCP/IP as an open standard. Benet is reaching for the same level of permanence - the kind of infrastructure that outlives the company that built it. Protocol Labs' own stated mission is to improve the internet as infrastructure for human knowledge.

This is the context for a decision that baffles conventional investors: Benet explicitly chose open-source for all of his protocols. No proprietary lock-in. No moat built on closed APIs. The revenue model for Protocol Labs comes through Filecoin - a token economy that rewards participation rather than a subscription that rewards dependence. Traditional VCs, he has noted publicly, don't fund this kind of work. The timelines are too long. The returns are too indirect. That is why YC's Summer 2014 check mattered, and why the $257 million ICO in 2017 mattered even more: it funded decade-scale research without the quarterly pressure that would have killed it.

From Cuernavaca to Codebase

Juan Batiz-Benet was born on March 16, 1988, in Cuernavaca, in the state of Morelos, Mexico - the same city that inspired Malcolm Lowry's "Under the Volcano." He moved to San Diego as a teenager and arrived at Stanford for computer science, eventually completing both a B.S. (2010) and M.S. (2012) there. The research focus: distributed systems, networking, and cryptography. Which is to say, the precise set of disciplines required to build what he would go on to build.

The Stanford years produced Loki Studios, co-founded with classmates Ivan Lee, Rico Roque, and Brian Laub. Their game Geomon was clever - creatures spawned by real-world GPS coordinates and weather data, a model for location-aware computing years before it became mainstream. At its peak the game had roughly one million users. Yahoo acquired the team in May 2013. Within months, the game was gone.

What followed was Athena - Benet's attempt to build an open-source platform for human knowledge. Think of it as a package manager and cloud for ideas: structured, versioned, and freely accessible. The project failed. Coursera and Udacity had more funding, clearer paths to revenue, and better timing. The Athena writings survive online, preserved - appropriately - exactly as Benet intended his protocols to preserve everything: permanently accessible, regardless of what happened to the company that made them.

The ICO That Changed the Rules

August 10, 2017. Filecoin opened its token sale at 9 AM Pacific. In the first 30 minutes, $200 million was committed. The final tally, when the sale closed: $257 million. It was one of the largest initial coin offerings in history at the time, and it was conducted legally - using the SAFT framework that Benet had co-developed specifically to make compliant token fundraising possible in the United States.

The fundraise was unusual in another respect: it was conducted through CoinList, the platform Benet had co-built with AngelList, which required investors to be accredited and imposed actual compliance infrastructure on what was otherwise a largely unregulated market. The message was deliberate. Benet was not trying to exploit regulatory ambiguity. He was trying to create a precedent for how this should be done. The SAFT became a legal template used across the industry.

Filecoin mainnet launched October 15, 2020. Within one month, the network had exceeded 1 EiB (over 1.1 million terabytes) of total storage capacity. By 2024, that figure had grown past 15 EiB of raw byte capacity. The economics work on a simple principle: storage providers around the world pledge their hard drive space, prove they're holding the data cryptographically, and earn FIL tokens for doing so. The result is a decentralized alternative to AWS S3 that is - in theory - owned by no one and censored by no one.

Quietly Serious, Genuinely Strange

Juan Benet's public presence is a study in understatement. His GitHub account (jbenet) has 286 public repositories and 2,600+ followers. His bio: "beep boop." Location: "Earth." He maintains a personal site at juan.benet.ai - simple, minimal, linked from his GitHub profile without fanfare. His Twitter bio describes him as interested in "Knowledge, Science, and Technology." Which, in 2024 terms, covers just about everything he has done, is doing, and is probably planning.

Those who have worked with him describe a founder who is genuinely long-term in his thinking - not as a posture, but as a practical framework for decisions. When he makes choices about open-sourcing protocols or refusing to add proprietary lock-in, it is not altruism. It is strategy on a century-scale timeline. The internet's current centralization is a recent phenomenon; the open-source protocols underneath it (HTTP, TCP/IP, SMTP) date from an era when no company was large enough to own them. Benet is trying to create the next layer of infrastructure in the same mode - before any company gets big enough to own it.

He donated $1 million in Bitcoin in 2021 to Bitcoin and financial inclusion initiatives - quietly, without the press tour that typically accompanies a $1M donation from a crypto founder. His code is preserved in the GitHub Arctic Code Vault in Svalbard, Norway, designed to last 1,000 years. Whether this was something he requested or simply the result of his open-source work meeting GitHub's archival criteria is unclear. Either way, it fits.

Neurotech and the Upgrade of the Human Mind

In April 2026, Juan Benet launched a podcast. The subject was not IPFS. It was not Filecoin. The first episode featured Max Hodak, CEO of Science Corp, discussing retinal prosthetics and biohybrid brain implants. The series is called the Juan Benet Podcast, and it is described as starting with neurotech before expanding to AI, computing, and human intelligence augmentation more broadly.

The arc makes sense if you follow his logic. Benet has always been interested in knowledge systems - how humanity stores, retrieves, and transmits information. IPFS addresses the storage layer. Filecoin addresses the incentive layer. The neurotech pivot addresses the interface layer: how information gets into and out of human minds. His talk at Devcon SEA in late 2024 was titled "Neurotech - humanity's next frontier." At Vision Weekend Europe the same year, he spoke on "Accelerating Neurotech & Fast Grants."

"Today, neurotech stands to restore not only vision, but also hearing, speech, and mobility for millions of people," he said. For a man who spent the previous decade building infrastructure for how machines talk to each other, the next decade appears to be about how machines will talk to us - and how we will talk back. This is not a pivot away from Protocol Labs' mission. It is a natural extension of the same obsession with information infrastructure, extended from silicon to neurons.

What the Details Reveal

There is a version of Juan Benet's story that is told in billion-dollar fundraises and Fortune magazine lists. That version is accurate but insufficient. The more revealing version is the one told in small decisions: turning down Yahoo in 2013. Failing with Athena and publishing the writings rather than burying them. Building Filecoin's legal framework before building Filecoin, because doing it right mattered more than doing it fast. Cutting 21% of Protocol Labs' staff in 2023 with a public, honest letter about what the crypto winter had done to the industry - no euphemisms, no spin.

He is, in the terminology of Silicon Valley, a "technical founder" - the kind who writes papers and code rather than decks and press releases. His research output (the IPFS whitepaper, Proof of Replication, Distributed Web Applications with IPFS, Scaling Proof-of-Replication for Filecoin Mining) is cited academically. His GitHub commit history is extensive. He is not a CEO who happens to have an engineering background. He is an engineer who happens to run a company.

The ambition is genuinely unusual. Most founders optimize for a 10-year exit. Benet is explicitly designing for 50-100 year durability. That requires a different kind of decision-making: less about market timing, more about technical permanence. Less about competitive advantage, more about avoiding the failure modes that have doomed previous attempts at open internet infrastructure. The Athena failure, in this light, was not a detour. It was a data point in an ongoing experiment about what kinds of internet infrastructure survive and what kinds don't.

He is now 38 years old. The protocols he built a decade ago are still running - still being developed, still being adopted, still expanding into the infrastructure layer of a web that is slowly, unevenly, but genuinely shifting toward decentralization. The podcast is new. The neurotech interest is new. But the core obsession - knowledge, preservation, access, the long-term architecture of how minds connect to information - has been constant since the Athena days.

Yahoo got the game. Benet got the internet.