Hyperlinked books and other interconnected information—imagine the AI potential of a truly trustworthy national digital library system. But beware if you want AI companies to be part of the effort. So far they’ve balked at the idea of using only trustworthy content.

That was one of the most important points that Tony Marx, President of the New York Public Library, made in a recent podcast that Reid Hoffman and Aria Finger hosted on LinkedIn.

Marx apparently had in mind material that was trustworthy in the usual library sense. But what to do when lying authoritarians like Donald Trump run the country and can order librarians to offer propaganda, including racist content? Trump and his crew are lawless and eager to flout traditional library norms. Of course, omission might be an even greater risk than commission. For example, the Trump administration is especially keen on censoring certain positive mentions of minorities in American history.

In a comment, I’ve reproduced an AI-generated summary of the podcast. Meanwhile, let me emphasize that the thoughts on propaganda and censorship are my own, not necessarily Tony Marx’s, although I suspect he would share my concerns. Also note that the podcast covered other topics, especially library access for low-income people.

Podcast found via Bob Snyder.