Charles Frederick Page and the airship

Banned Histories of Race in America

Earlier this week Washington Post AI reporter Gerrit De Vynck spoke about just how little AI cares about facts. Like, any facts at all. De Vynck talked about Googling what time it was in London and Google’s AI getting it wrong. Yikes. His explanation was that Google’s AI acts, “as a people-pleaser and if it doesn’t have information, if it doesn’t know the answer to what you’re asking, it just makes something up that generally sounds pretty good, but might not be accurate.”

While this problem is presented as something specific to AI, it’s not. In fact, this flaw is actually at the core of how we, as Americans, understand information. Our collective desire for knowledge has always been more about reinforcing narratives, identities and hierarchies than fact finding. I mean, this country has established multiple fields of science, entire forms of government, the police and countless careers, laws, philosophies and just general things all for the sake of creating and maintaining white supremacy. It shouldn’t be controversial to say that you can’t behave in that way while also valuing facts.