You’re comparing two very different things there. Institutional libraries and government archives have records going back centuries, many of which can’t be found anywhere else. Even your local library might have some obscure local newspapers on microfilm from 100 years ago, that haven’t been digitized yet (we’ve admittedly made progress on this problem over the last decade).

To answer these “who am I” questions posed in the article, of unidentified historical photos, we need institutions like the Library of Congress digitizing their records, uniting them with other institutional datasets, and organizing them in a way that we can run facial recognition algorithms on them. The answer to “who are these musicians circa 1930” probably comes from a local newspaper that ended publication 50 years ago.

In case it isn’t abundantly clear, this is going to take decades to accomplish. These institutions operate on timelines measured in centuries, not weeks. They care deeply about problems like bit rot when replacing physical archives. If it is lost to the world in mere decades, it is not useful to them.

The Internet Archive is of little help here, beyond sharing technical information for efficient archival and retrieval of truly massive public datasets.