Understanding
genres and their conventions have enabled people to make software that generate
genres. These “genre generators” use a certain genre’s conventions and create
an example of that genre. Patterns of some genres have become so predictable
that they could be programmed in a computer. The programmers who developed
these “genre generators” know how certain genres are set up and used that
knowledge to let internet users create works of that particular genre.
A genre as sophisticated as
Computer Science research papers could be broken down and replicated by a
machine. SCIgen, the automatic CS paper generator, allows the user to create a
computer-made CS paper just by entering a name of an author and then pushing a
button. All of the papers generated in SCIgen follow the same format: title,
authors’ names, abstract, introduction, body, conclusion, and references. For “genre
generators” like this to work, it would require consistency. If the paper has
information that makes sense, the rest of the information on the paper must not
only make sense, but be coherent as well. However, having a machine write a valid
paper would require more than just patterns. So to solve this problem, everything
that is written on the paper must be complete nonsense. The point of this
generator is not to create a legitimate CS paper but to show what a typical CS
paper may look like, or “to auto-generate submissions to conferences that you
suspect might have very low submission standards” as mentioned in the “About”
section of the SCIgen homepage. To give some credibility to the paper, academic
language is used. They also included graphs and figures to make the paper more
believable. To make the graphs and figures appear true, the paper uses the
first-person word “we” as if to show that personal research had been done to
obtain the data. I personally searched the references used in the generated
papers and I learned that they were also fake. One of the authors were actually
“agjhk”, the made-up author name that I entered.
Another “genre generator” is Pandyland,
a comic generator. Every comic is comprised of images and text inside speech
bubbles. Pandyland makes exactly that. As it says on its website, Pandyland is random,
so like SCIgen, some of the comics generated might not make sense. The comics users
can make only have three panels: beginning, middle, and end. Because of that,
the stories created are not too intricate and don’t make much sense. It does
however give room for the viewer to use their imagination and make sense out of
the comic. It is easy to assume that Pandyland’s intention is to provoke humor
as the panels contain silly elements to them. A generator that produces random
and nonsensical images with the intention of comedy would have a better chance
of receiving a positive response from the audience than with the intention to
produce serious comics. With Pandyland, preset images with no apparent context are
shuffled to make a somewhat sensible comic.
Instead of having prewritten text,
Memegenerator just provides a template and lets the users create the text. These
memes that users create consist of an image with text on the top and bottom of
the picture. The top text is usually the intro to the joke and the bottom would
be the punchline. The generator provides different meme images for people to
use. Each of these images have their own context and creators use jokes that
fit with the context of that image. One example of a meme image is Bad Luck
Brian. The picture shows a boy who someone could say appears unlucky and the
textual jokes used with the picture involves Brian getting into very unfortunate
events. These individual meme images are a genre themselves.

A “genre generator” that I found
called Chaoticshiny, specializes in writing and gaming for the game Dungeons
and Dragons. For people that run out of ideas for their story, they could use
the generator to make stuff for them. It has multiple generators to choose from
including a civilization generator, name generators, and even an RPG Drinking
Game generator. One that stood out to me was their tarot card generator. It
lets you choose the quantity of tarot cards you want to create and which
information from description, meaning, condition, and the back of the card you
want to include. The format of the tarot card is simple; it starts with what
image is on the card, then says what is associated with it and what it
represents. A description of what’s on the back of the card could also be
included. Using conventions of a tarot card, the program can randomize the
stereotypical information on a tarot card and combine them to make a new card.