Written in Glass: The Future of Memory and the Risk of Being Forgotten

In laboratories and data centers, a new form of memory is being built. Engineers at Microsoft have developed a method of storing data inside glass, encoding information with lasers into layers that can last for centuries, possibly thousands of years. The project, known as Project Silica, is designed to solve one of the most pressing challenges of the digital age: how to preserve information over long periods without degradation.

At a technical level, the achievement is significant. Traditional storage systems degrade. Hard drives fail. Magnetic tapes decay. Formats become obsolete. By contrast, glass offers durability. It can withstand heat, water, and electromagnetic interference. Information is etched within it, not written on the surface, creating a medium that is both stable and resistant to environmental damage. In controlled demonstrations, entire films and large datasets have been successfully stored and retrieved.

The promise is clear. Humanity now has the ability to preserve knowledge at a scale and lifespan that previous generations could not have imagined. Yet the question is not only how long information can last. It is whether it can be understood.

For most of human history, memory was physical and legible. Civilizations carved their stories into stone, painted them onto walls, and inscribed them into materials that could be seen and interpreted without machinery. In regions like Belize, the remnants of the Maya civilization still stand. Their glyphs, etched into temples and monuments, remain visible centuries after the societies that created them declined. Even before their meanings were fully deciphered, their presence was undeniable. They signaled that something had been recorded, something worth preserving. Across Mexico, Belize, and Guatemala, the Maya erected stelae that recorded dynasties, ceremonies, and moments of power in carved stone. These monuments remain legible centuries later, offering historians a direct line into the past.

Lamanai stela in Belize, photographed by Kristen Lanum/Getty Images. Lamanai was part of the Hyde family portfolio and served as a refuge during Spanish and British colonial periods. Other civilizations approached memory differently. The Aztec civilization used pictographic systems that conveyed meaning without fully phonetic writing, while the Inca Empire developed quipu, a system of knotted cords used to record information that scholars are still working to fully interpret. In each case, the record survives, but understanding exists on a spectrum.

Other civilizations approached memory differently. The Aztec civilization used pictographic systems that conveyed meaning without fully phonetic writing, while the Inca Empire developed quipu, a system of knotted cords used to record information that scholars are still working to fully interpret. In each case, the record survives, but understanding exists on a spectrum. Modern data does not behave this way. It is not visible. It does not announce itself. It exists within systems that require layers of interpretation. Hardware must function. Software must be compatible. Formats must be recognized. Without these conditions, the data may still exist physically, but it becomes inaccessible.

This introduces a paradox at the center of technological progress. The more advanced our methods of storage become, the more dependent they are on the systems that support them. Glass storage solves the problem of physical decay, but it does not solve the problem of readability. A future society may recover a piece of glass containing vast amounts of information and still be unable to interpret what is inside it.

This is not a distant or theoretical concern. Even within a few decades, data has been lost because the systems required to read it no longer exist. Early digital formats, once standard, have become difficult or impossible to access. Hardware disappears. Software becomes obsolete. What remains is not the absence of information, but the absence of access. The implications extend beyond technology. They touch on how civilizations are remembered. Archaeology depends not only on what survives, but on what can be understood. A carved inscription, even when partially eroded, offers a point of entry. A digital archive without context or compatible systems offers none.

Project Silica represents a shift in how humanity approaches memory. It reflects a belief that preservation is primarily a technical problem. In many ways, it is. Durability matters. Scale matters. The ability to store vast amounts of information securely over long periods is essential in a world that generates data at unprecedented rates. But preservation is also an interpretive problem. Information must not only endure. It must remain accessible, translatable, and meaningful across time. Without that, preservation becomes incomplete.

There is a deeper continuity between past and present that is often overlooked. Ancient systems of recording were constrained by materials and tools, but they were designed for persistence and recognition. Modern systems are defined by abstraction and efficiency, but they often assume continuity of infrastructure. That assumption may not always hold. The question is not whether future generations will inherit our data. It is whether they will be able to understand it.

Glass may outlast paper, magnetic tape, and silicon-based storage. It may carry forward an unprecedented volume of human knowledge. But without the means to interpret what has been stored, that knowledge risks becoming silent. In this sense, the future of memory is not only about endurance. It is about translation. It is about designing systems that can survive not just time, but discontinuity.

Humanity has always tried to leave a record of itself. The mediums have changed, from stone to paper to digital code. Each transition has expanded what can be preserved, while introducing new vulnerabilities. What is being built today will shape how this era is remembered, if it is remembered at all. We may succeed in storing everything. The more difficult task will be ensuring that it can still be read.