generally useful maybe (though maybe also bad idk)
is to build a pipeline P that makes it easy to interact with large amounts of fairly legible media content. Specifically, things like
point P at an author / source (e.g. a blog, twitter account, youtube channel, forum user; or a person, organization, political party), and then P indexes all the stuff from that source
ask P what a given source says about X
ask P to search for contradictions (e.g. a pundit saying J6 is terrible and then later saying J6 is fine, or what have you)
ask P to collect all instances of X (e.g. instances of ICE officers behaving poorly)
ask P to collate videos or other media into fast-paced clip shows
given a bit of media (tweet, clip, etc.), fetch the full context
etc.
This has good and bad uses. It would make it easier to clip-chimp. However, it might make it easier to break into really intense filter bubbles that are clip-brained, using concentrated memetic injections.
One thing that I think might be
low hanging fruit
generally useful maybe (though maybe also bad idk)
is to build a pipeline P that makes it easy to interact with large amounts of fairly legible media content. Specifically, things like
point P at an author / source (e.g. a blog, twitter account, youtube channel, forum user; or a person, organization, political party), and then P indexes all the stuff from that source
ask P what a given source says about X
ask P to search for contradictions (e.g. a pundit saying J6 is terrible and then later saying J6 is fine, or what have you)
ask P to collect all instances of X (e.g. instances of ICE officers behaving poorly)
ask P to collate videos or other media into fast-paced clip shows
given a bit of media (tweet, clip, etc.), fetch the full context
etc.
This has good and bad uses. It would make it easier to clip-chimp. However, it might make it easier to break into really intense filter bubbles that are clip-brained, using concentrated memetic injections.