Archivists

Archivists still do most of their work by hand, but computers are helping with some tasks. For example, scanning and digitizing old records is partly automated today: special scanners and OCR software (optical character recognition) can copy paper documents into digital text. Official job guides list “copying records … to computer formats” as a task ^[1].

That means archivists often use machines to scan photos, documents or even audio. Once digitized, tools can also help organize and tag all that data. In fact, the U.S. National Archives has built a new exhibit where machine learning (a type of AI) tags and sorts over 2 million historical items so visitors can search them ^{[2] [2]}.

And industry surveys show this is a big goal: 45% of media-archiving professionals rate “automation, metadata enrichment” (another word for tagging and labeling) as a top challenge ^[3]. In short, technology is augmenting work like data-entry and classification.

However, many archivist duties still need human judgment. Tasks like finding new items to add, deciding what history to keep, and guiding tours rely on people. O*NET (a U.S. job database) even lists “coordinate… out­reach programs, such as tours, workshops, lectures” and “locate new materials … for display” as core archivist tasks ^[1].

Those roles involve creativity and personal skill, so far beyond AI. As one curator noted, AI tools are not writing history – they sort it. Archivists emphasize that they use AI only to organize records, not to create new ones ^[2].

Archives and museums are interested in AI but take it slowly. Big institutions with money can invest: for instance, the new archive exhibit cost $40 million and uses AI to help visitors ^[2]. A recent survey found many archives plan to increase their tech budgets – about 76% of media archives said they will spend more on technology soon ^[3].

That means they see value in faster searches and smarter catalogs. Also, libraries are digitizing huge collections for AI (for example, Harvard opened nearly one million old books for AI researchers ^[4]), so the data is becoming available.

But costs and caution slow things down. Advanced AI systems can be expensive compared to hiring staff. Smaller archives may not afford fancy tools yet.

Also, archivists must be careful about trust and ethics. They worry AI could mislabel or even "rewrite" history. In fact, the National Archives made a point of saying AI is used only to connect people with real history – not to invent or change it ^[2].

In the end, archivists expect AI to help with big data tasks (searching, sorting, tagging) but see human skills (judgment, teaching, compassion) as irreplaceable. This means adoption will be steady but cautious, with technology taking on the heavy lifting while people keep guiding the way ^{[2] [3]}.