National Archives Suggests Artificial Intelligence Could Erase ‘Inherent Bias’ From Nation’s Historical Records

An official with the National Archives said that her organization wants to ‘recognize and mitigate’ in America’s historical documents.

AP/Michael Dwyer, file
The OpenAI logo is seen on a mobile phone in front of a computer screen displaying output from ChatGPT. AP/Michael Dwyer, file

As the artificial intelligence race picks up across the world, one government entity wants to use that processing power to remove or rewrite certain aspects of American history in order to delete what officials say is “bias” in the material.

In a newly released video, the chief innovation officer of the National Archives and Records Administration, Pamela Wright, says future AI could help find new narratives about American history by removing bias, whether it be based on race, sex, or class. 

“We are not focused solely on the technical possibilities,” Ms. Wright said during a panel discussion, “but also on the ethical use of AI, which is critical to our successful adoption of these emerging technologies.”

“As you know, NARA is the country’s record-keeper, and given the fact that many of the permanent records of the federal government are inherently biased — the majority of the records were created by people in power,” she continued. 

Ms. Wright said that her organization wants to “recognize and mitigate that bias.”

Congressman Thomas Massie, an ardent libertarian, raised concerns about Ms. Wright’s comments on Wednesday, calling her stance “chilling” and agreeing with one commenter who said NARA’s deployment of AI would be the equivalent of “digital book burning.”

Ms. Wright did not immediately return a request for comment. 

According to the market research firm Precedence, the global AI industry was valued at about $120 billion in 2022 and is expected to grow to more than $1.5 trillion by 2030. 

The technology’s ability to restructure society — and the rate at which it is being developed — worry many about its potential for replacing jobs both blue collar and white. Many tests have found that AI and machine learning technologies can pass post-graduate licensing exams like the bar. 

The CEO of OpenAI, Sam Altman, admitted in an interview with ABC News that even he is a “little bit scared” about what he and his team are developing. In March, Mr. Altman’s firm received a private valuation of $30 billion, making it one of the largest AI development firms in the world. 

The current interactions of AI and machine learning apply to relatively simple things like grading homework, tracking user data on social media, or developing advertising campaigns. Using the technology to rewrite historical documents or to analyze them through the lens of an author’s power status is a new front, especially for a government agency. 

The digital records of NARA have grown exponentially just in the last decade. In 2015, the agency held just over five million “digital objects” — scans of records, maps, photographs, videos. By 2020, that number had swelled to nearly 125 million, a growth rate of 2,400 percent.


The New York Sun

© 2024 The New York Sun Company, LLC. All rights reserved.

Use of this site constitutes acceptance of our Terms of Use and Privacy Policy. The material on this site is protected by copyright law and may not be reproduced, distributed, transmitted, cached or otherwise used.

The New York Sun

Sign in or  create a free account

By continuing you agree to our Privacy Policy and Terms of Use