And all of this because I can keep replacing shitty locks over and over for less than it costs to make a key for the old lock that actually works.
That was my morning.
Speaking of broken junk…
Everybody knows that fringe history has a recycling problem, but I swear that the internet is making the lather-rinse-repeat cycle shorter than ever. Back in November Micah Hanks published an article on Mysterious Universe in which he asked whether Edgar Allan Poe was a time traveler who had advanced knowledge of the cannibalism death of Richard Parker, which he then included in his novel The Narrative of A. Gordon Pym. (He didn’t; the details, aside from the name, are entirely different.) Today Hanks recycled the same material from his November article, and pretended to be shocked to read about the time travel conspiracy theory in a different November 2015 article (which he calls “recent”). He also admits that he is recycling the content, yet again, from the first time he wrote about it in Fate magazine back in 2007!
Now, there isn’t anything inherently wrong with revisiting stories. Heaven knows I’ve talked about the same material more than once. But what’s galling is that there is never anything new in a Hanks recycling job. He doesn’t revisit stories to update them, to provide new insights, or to correct errors; he just repeats the same topics at random, like a very limited iTunes playlist on “Shuffle.” The reader doesn’t learn anything, but Hanks gets paid again!
In fact, it would seem that fringe sites could cut out the middle man and save money by using bots to generate computer-written material by remixing content from their archives. It would probably save money without altering the quality at all. Indeed, there is a Canadian movie production company that is currently using a computer to generate the plot and aesthetics of a new horror movie based entirely on audience reaction to earlier horror movies.
“It not only evaluates and suggests plot twists and deviations,” said Jack Zhang, the founder of production company Greenlight Essentials, “it also suggests which type of actors and actresses would make the movie more appealing, specific plot and cast combinations and even helps us find the right target market for the film to be successful.”
The computer’s “potent data processing” determined, for example, that to reach women under the age of 25, the movie must feature a ghost, family relationships, a score composed for piano, and a scene set in a bathtub. I’m not sure how processing data and delivering a spreadsheet counts as artificial intelligence, but press releases are not known for their accuracy.
The trouble I see from this is that the computer can only process data based on what has already been done. It can tell you how to remix and recycle, but it doesn’t seem to be designed to generate anything new. After a few iterations, it will be spitting out the same story over and over based on the highest-scoring data and then filmmakers will wonder why audiences stop responding to it. It may seem “data-driven” now, but so did the brain trust that gave us wall-to-wall sequels. And eventually audiences caught on to that and stopped responding to it. At least some element of originality and surprise is necessary to make a movie worth caring about—and here I define that very loosely. The recent remake of Macbeth, after all, is about as familiar a story as one can find, but it applies the aesthetics of Game of Thrones to the story, which is at least a bit of originality, or at least interest.