Is it some sort of program that was written? How does it process those words and scenes sensibly and then reproduce them semi coherently?
Is this some inside joke that I missed?
In: Technology
As another commenter said, they’re mostly parodies written by people.
In the cases that are done by computers, the data is usually in word or still photo form. They pass that data to a machine learning algorithm that tries to break down the data in a meaningful way so that it can replicate it.
A good example of this is [This Person Does Not Exist](http://thispersondoesnotexist.com) where they gave a machine learning algorithm a lot of pictures of faces and the algorithm tries to make new faces based off of the data it was given.
There’s also [Talk to Transformer](https://app.inferkit.com/demo) (sadly, it looks like the demo is down for the time being) that was trained with a bunch of news articles and Reddit posts/comments in order to get the algorithm to try to replicate real human writing.
There’s [Harry Potter and the Portrait of What Looked Like a Large Pile of Ash](https://botnik.org/content/harry-potter.html) which was generated by a machine learning algorithm that was given the Harry Potter books as data to replicate. This is the one that got famous and caused people to create parodies of their favorite shows and movies under the guise that they were generated by a machine learning algorithm.
It’s just a way to set up a parody script. That’s all. None of them that are shared are from any kind of artificial intelligence.
The majority of the “I had my computer watch X hours of Y and here’s what it wrote” are parodies written by actual humans. There was a “I had a computer read Harry Potter and try and generate a new story based on that” effort a few years ago, which was legitimate, but it inspired a bunch of copycats.
Latest Answers