Martin Kunze wants to gather a snapshot of all of human knowledge onto plates and bury it away in the world’s oldest salt mine.

In Hallstatt, Austria, a picturesque village nestled into a lake-peppered region called Salzkammergut, Kunze has spent the past four years engraving images and text onto hand-sized clay squares. A ceramicist by trade, he believes the durability of the materials he plies gives them an as-yet unmatched ability to store information. Ceramic is impervious to water, chemicals, and radiation; it’s emboldened by fire. Tablets of Sumerian cuneiform are still around today that date from earlier than 3000 B.C.E.

“The only thing that can threaten this kind of data carrier is a hammer,” Kunze says.

So far, he has created around 500 squares, which he allows anyone to design for a small donation. Many preserve memories of the lives or work of people involved in the project. Around 150 of the tablets showcase items from collections in Vienna’s museums of National History and Art History. Some local companies have been immortalized. One researcher’s CV now lies in the vault.

But Kunze aims to expand the project, to copy research, books, and newspaper editorials from around the world—along with instructions for the languages needed to read them. For this, the clay squares he’s currently using would take up far too much space than could be set aside for such an audacious undertaking. So Kunze also has conceived of a much thinner medium: He will laser-print a microscopic font onto 1-mm-thick ceramic sheets, encased in wafer-thin layers of glass. One 20 cm piece of this microfilm can store 5 million characters; whole libraries of information—readable with a 10x-magnifying lens—could be slotted next to each other and hardly take up any space.

“We can have the whole Harry Potter series on two ceramic microfilms,” Kunze says.

The goal of the project, which he calls the Memory of Mankind, is to build up a complete, unbiased picture of modern societies. The sheets will be stored along with the larger tablets in a vault 2 km inside Hallstatt’s still-active salt mine. If all goes according to plan, the vault will naturally seal over the next few decades, ready for a curious future generation to open whenever it’s deemed necessary.

To Kunze, this peculiar ambition is more than a courtesy to future generations. He believes the age of digital information has lulled people into a false sense that memories are forever preserved. If today’s digital archives disappear—or, in Kunze’s view, when they do—he wants to make sure there’s a real, physical record to mark our era’s place in history.

* * *

In 2013 alone, humans produced 4.4 trillion gigabytes of data, according to the International Data Corporation. This number doubles every two years. By 2020, it’s predicted that 6.1 billion people will use a smartphone, and that “connected things”—everything from self-driving cars to artificially intelligent toothbrusheswill constitute almost a third of the digital universe, which will have grown by an order of magnitude from today.

Much of this information goes into digital storage—ranging from servers on personal computers to colossal data centers, like the NSA’s facility in Utah. Such sites are growing to accommodate the surge in data; one currently under construction in Nevada will cover nearly 6.5 million square feet when completed—113 football fields of server space.

But this method of storage has inherent problems. Digital space is finite and expensive. Digitally stored data can become corrupted and decay as electrical charges used to encode information into binary bits leak out over time, altering the contents. And any enduring information could be lost if the software to access it becomes obsolete. Or a potent, well-timed coronal mass ejection could cause irreparable damage to electronic systems.

“There’s no getting around the risk of catastrophic loss in our culture,” says Robert Darnton, the librarian emeritus at the Harvard University Library. “Digital texts are much more fragile than printed books.”

In the face of such concerns, collaborative projects have been popping up around the world with the explicit intention of preserving history. In 2013, Darnton helped launch the Digital Public Library of America, an initiative that links digitized collections from museums, libraries, and historical societies across the country. In the United Kingdom, an organization called the Digital Preservation Coalition provides digital-preservation guidance to a growing network of members, from the BBC to the Bank of England.

But these efforts are still digital. Even if societies manage to keep adequate records, Kunze contends, we’re bound to hit a wall: Either storage space will run out or the environmental costs of maintaining it will grow too large. In the latter scenario, our data would have to be ruthlessly removed to reduce energy use from digital storage, he says. “This will be executed by algorithms, and therefore the pictures our grandchildren will have of our time will be dictated by machines.”

Kunze believes an inevitable path toward environmentally friendly architecture will lead to a decline in our physical presence, too. Design movements such as “cradle-to-cradle,” in which construction materials are clean, renewable, and completely recyclable, will mean that buildings no longer remain as relics. “We have to become a green society, otherwise we will not survive on this planet,” Kunze says. “But future archaeologists will only find traces until the 21st century. After, there will be just pure earth.”

* * *

Kunze’s quest began more than decade ago, when he was studying art and art history at the University of Arts and Industrial Design in Austria. He realized the storage potential of clay, and as his concerns grew in the years following, he came to the conclusion that humanity needed a ceramic time capsule. In 2012, he says, he contacted the company that extracts salt from Hallstatt’s mine, and they agreed to provide space for him to house his vault inside.

So far, Kunze has created all the large tablets himself in his studio, and a few private companies are helping with the production of the microfilms. Some prototype samples of these have been produced, now he just needs content to fill more. As word about the project has spread, a handful of researchers from universities in countries as far afield as Australia have reached out to Kunze to learn more about it. One professor at the University of Vienna is looking into how music and sound might be stored on ceramic, too.

As the project slowly starts to take shape, some are worried that its own place in collective memory may ebb over time. “The thing I don’t like about the time capsule is the sense that it’s frozen,” says Richard Ovenden, the director of the Bodleian Libraries at the University of Oxford. “Information is much more likely to be kept if it’s used. The danger is that [Kunze’s project] will end up being forgotten.”

To avoid this, Kunze plans to distribute ceramic tokens around the world to everyone who either funds, contributes to, or advises on the project. Every 50 years, starting in 2070, he says, holders will meet to keep the memory of the capsule alive and to discuss if it needs to be reopened. The location of the mine will be carved onto each token, and it will require geological knowledge similar to our own to find it, especially as land shifts with time. This would be a safeguard against unwanted discoveries if for some unpredicted reason—nuclear war, say—human civilization disappears or regresses to the Stone Age.

As to the question of what information goes into the vault, Kunze is planning a seminal conference this year entitled “Zero footprint: What shall we leave behind?” Here, he hopes to gather researchers from various fields across science and humanities to devise a framework to decide how to choose. Popularity doesn’t necessarily mean value, he asserts: “The most read articles on Wikipedia aren’t about scientific research, but about Justin Bieber.”

Kunze has teamed up with the Human Document Project, another preservation scheme, and University College London’s Heritage Futures project, to co-organize the event. By doing so he wants to garner more interest from the scientific community to establish more credibility for the project.

* * *

For Cornelius Holtorf, an investigator for Heritage Futures, Kunze’s work adds an intriguing angle to contemporary archaeological discussions. “Archaeology is not about reconstruction of the past,” he says. “It is about communicating meaning through material culture.”

Holtorf is an advisor to the project, but he maintains a degree of skepticism. In time’s relentless march, “it’s probably right that we will lose a lot, but who cares?” he says. “There’s so much noise out there, we should be grateful that it’s gone.”

Indeed, the extent to which Memory of Mankind has actual value depends on getting the balance right between quantity and quality. Leave too little information and the message could be lost or misunderstood; leave too much and the risk is that any salient or critical knowledge is drowned. The world is already littered with unread historical documents, buried within giant archives and libraries.

More practically, enthusiasm alone can’t make this project succeed. Creating a time capsule on the towering scale Kunze is aiming for of course would need huge amounts of funding and practical support, not to mention content. Without recognition and contribution from a wide network of global institutions, obtaining any or all of those things will prove difficult. Convincing people to add in their own stories is one thing; persuading businesses, newspapers, and governments to donate information, money, and time to what is essentially a one-man mission seems less plausible.

Also, institutions of memory already exist all over the world. “Libraries are time capsules,” says Ovenden, “if they’re well run and see themselves in this long-term role”.

Nevertheless, projects like this could nudge the issue of digital preservation further into the public consciousness. “It is one of the critical problems of our age,” Ovenden warns. “If we fail to grapple with it, we will regret that in generations to come.”

One surprising area where Memory of Mankind shows promise is toxic waste. The 35 countries that make up OECD, the Organisation for Economic Co-operation and Development, collectively produce more than 300 million tons of toxic waste each year—nuclear and otherwise—which is buried in the ground in depositories all over the world. Some of these sites will remain dangerous for many thousands of years. Finding an appropriate way to mark them for distant generations is an issue that the nuclear industry has been grappling with for decades.

The Nuclear Energy Agency (NEA) has proposed coming up with some sort of durable marker, but as languages and symbols fade and evolve over time, it’s conceivable that the meaning of such warnings would be lost. Claudio Pescatore, an expert in radioactive-waste management who has worked for the NEA’s Preservation of Records, Knowledge and Memory across Generations initiative, says MOM could form part of a wider, “systemic approach” to this problem.

Pescatore’s an advisor to the project, and the person who came up with the idea of the ceramic tokens for participants. With enough tokens distributed around the world, he believes, toxic-waste information is more likely to be found and preserved. And by providing the tools for future generations not only to be aware of warning markers, but to interpret them, the tokens could “give enough information to make informed decisions about whatever is there,” he says.

For Holtorf, the Heritage Futures investigator, the extent to which information can be successfully passed from one generation to the next depends on the meaning and stories that cultures attach to it, so that it is kept relevant, and understood. “Maybe what is more crucial to survive in this project is not the object, but the memory that there was this guy, who put these things down inside the mountain,” he says.