Looking for a 100-year-old photo or an obscure book from the 1800s? The Library of Congress has you covered. But with today’s media growing ever more complex and confusing, the library now has to grapple with an important question: What will researchers of the future need to know about our time? At a moment when historians and museum curators are scrambling to collect Black Lives Matter signs and Covid masks, a team at the Library of Congress is making hard choices about how to preserve the web pages, memes, tweets, and viral videos that have come to dominate our increasingly virtual lives.

Founded as a pilot project in 2000, the four-person LOC web department today has a vital role as official archivists of the nation’s internet culture, documenting, for instance, grassroots sites behind Black Lives Matter protests, changing mask guidelines on the CDC’s website, and the latest Urban Dictionary definitions (“Covid brain” is one of the most recent to merit LOC attention).

“I call them living, breathing archives,” say Abbie Grotke, who leads the team. “You can’t select [the materials] and walk away. Content changes, and we have to keep up with it.” Just think about social media, where significant missives can fly by at a dizzying pace. Until 2017, the library archived every single tweet (yes, even yours), but that practice was discontinued due to the extraordinary volume. Now it’s up to Grotke’s department to figure out which ones deserve safekeeping.

So how to decide? That falls to a much larger staff of roughly 200 subject-matter experts tasked with keeping tabs on what’s happening in their specialty areas—the 2020 election, say—and flagging content to save. The department’s permanent staffers then handle the technical details. The archive now has more than 23,000 entries, including every Trump tweet.