There is a pattern in how things survive, and it runs opposite to intuition.
The object someone actually touched is almost always the first thing to disappear. What tends to survive is a copy, and then a copy of the copy, and eventually a description of a copy. The original is fragile precisely because it was used, and copies exist because they were made to be handled.
Consider photographs. A family's paper prints from sixty years ago are usually damaged, scattered, or in a box nobody has opened in a decade. The versions that persist are the ones somebody scanned, which means the surviving record is a compressed approximation of a physical object that may no longer exist. The scan is not the photograph. It is a good enough substitute that most families stop noticing the difference.
Consider film. The original camera negatives of most silent-era cinema are gone. What survives is often a print, sometimes a duplicate negative made from a print, sometimes a version assembled from fragments found in different countries. The version people now watch as the definitive restoration of a 1920s film is a reconstruction, and its claim to the word "original" is a claim about fidelity rather than about provenance.
Consider documents. The letter a person wrote by hand is the only object that carries their actual handwriting. It is also the object most likely to be thrown away, because it is small, unremarkable, and stored in a drawer. Historic archives are full of transcribed copies and relatively short on originals, because transcription was how things were preserved before photography, and transcription loses exactly the thing that made the original valuable.
Digital formats have their own version of the same failure. In 1986 the BBC produced the Domesday Project, a survey of British life released on two interactive laserdiscs to mark the 900th anniversary of the Domesday Book. It was read through a BBC Master computer and a special laserdisc player. Within about fifteen years the hardware had failed and the discs could no longer be read; the data was recovered only by an academic project that reverse-engineered the format, and even then the recovery was partial. A national institution had created what it intended as a permanent record, and it became unreadable faster than most paper.
The same thing has happened to recordings. The original slow-scan television tapes of the Apollo 11 moon landing were lost; a formal search in 2006 concluded they had most likely been erased and reused during a tape shortage in the 1970s. What exists today is a restored version made from broadcast copies of lower quality. The most significant television transmission in history survives, in effect, as a copy of a copy.
The same logic applies to buildings and objects, though the timescale is longer. The house a family actually lived in is renovated, subdivided, or demolished; what is preserved is a photograph, or a plan, or in the best case a museum reconstruction. Preservation almost always means converting a used thing into a described thing, and description is where the loss happens.
This is not a technology problem. Scanning, storage, and backup are all effectively solved. The reason originals vanish is that keeping them requires a decision by a specific person, repeated across decades, at a moment when nothing seems urgent. Someone has to look at a box of letters and decide, this year, not to throw it out.
That decision fails far more often than it succeeds, and when it fails it fails silently. Nobody knows what was in the box.
The practical consequence is that the window for preserving a person's actual material is narrower than it feels. It is open while the material exists and while someone who cares is still present. Those two conditions overlap for a generation or two at most, and after that the material is either somewhere stable or gone.
The Chinese project 3026 is built around that window. It does not attempt to interpret anyone, summarise a life, or produce a narrative. It keeps the original words, in the form they were written, with the identity of the writer attached.
That is a deliberately modest goal, and it is the hardest one to actually achieve, because it requires acting while there is no visible reason to act. A record that gets made this year is unremarkable. The same record, had it failed to be made this year, would be irreplaceable in thirty.
https://3026alive.com