We need an open-source AGI governed like a public utility whose purpose is to process news and media content with a centrist, evidence-based, logical lens.
People will stop reading the NY Times or FOX articles and instead will read the AGI versions of them, which will also store all instances of bias, error, and propaganda into a publicly accessible database that will rank sources on several scores.
This database will also allow discovery of patterns and narratives.
I firmly believe this would be so successful that its datacenters would be under constant threat.
How would that even work? You'd have to train the AI as to what "a centrist, evidence-based, logical lens" would look like, and that training will inescapably have its own bias.
I mean, nobody can even agree on what "centrist" actually looks like.
People will stop reading the NY Times or FOX articles and instead will read the AGI versions of them, which will also store all instances of bias, error, and propaganda into a publicly accessible database that will rank sources on several scores.
This database will also allow discovery of patterns and narratives.
I firmly believe this would be so successful that its datacenters would be under constant threat.