Interesting Novels Written By Artificial Intelligence
How good and comprehensible are they?
“He who does not weep does not see”
“It is nothing to die. It is frightful not to live”
These lines are from the famous Les Misérables. Technology is currently in the process of creating greats such as this. AI is Creative! Music, Art, Games, it is starting to do it all! With machine learning, creativity knows no bounds. Today we will be diving into another such creative domain which is writing.
The following is a list of 5 books created by or with the help of Machine Learning that might just bring up fear in authors as there is a book here which placed further in a competition that it’s human competitors. One thing to note going in is that the books don’t make much sense but the descriptions portrayed are of high quality.
1 the Road — Ross Godwin and Kenric McDowell
- “1 the Road” is taken from Jack Kerouac’s On the Road. Ross Goodwin drove from New York to New Orleans in March 2017 with an AI in a laptop hooked up to various sensors, whose output the AI turned into words.
- As writing is a task done in continuous sequences it was framed by LSTM Networks. The model had been trained on previous text data. Data was collected in forms of scenes, speech captured inside the car and coordinate values.
- As understood from data collection, the book has been written on the basis of the mood set by the surroundings and how people’s conversation changed with it.
- Though the book is quite ambiguous the scenes explained are quite a delight to read.
- The project was definitely a success as it was funded by google and there were good critic reviews. Brian Merchant from the Atlantic said, “there are some striking and memorable lines”.
- The novel was published in 2018 by Jean Boîte Éditions. This is also one of the first novels completely written by AI.
The Day a Computer wrote a Novel — Kimagure Artificial Intelligence Writer Project.
- As the name suggests it is about a computer who writes a novel. The computer portrays a story with “itself” as the protagonist. Yes, the sentence is right, “itself”.
- This novel was made mostly by humans but the AI restructured it. The main aim was to see a standpoint of creativity by AI.
- The project also made key progress noting the liabilities in attempting such a task.
- The book is based on a system known as the Ghostwriter system which is explained here which is the official print that explains the story and how it was formed.
Dinner Depression — Julia Joy Raffel
- Another book completely developed from Machine Learning. From the cover art to the price which is put forth on amazon.
- This book is on the list solely due to its mysterious feel which one gets while attempting to read it.
- From the blurb alone it is quite hard to interpret what the story is trying to convey but what one understands is the vivid descriptions given from all the scenes.
- Though not mentioned, such projects are always based upon the LSTM algorithm.
World Clock — Nick Montfort
- World Clock generated on twitter when Monfort announced the open-sourced project. It was developed for the NaNoGenMo (National Novel Generation Month).
- The story just goes about various timestamps around the world with no certain connection as such.
- What this story provides is a path where one can verify that a sense of time can be instilled while making a computer write code.
- Looking into the code, one can see few parts of it are hardcoded and the majority of it is done with the pytz package in python.
- The source code and the information can be found here. The book is free to read for all.
Irritant by Darby Larson
- Irritant is a story conveyed in one paragraph which goes on for 624 pages. This book twists fiction like nothing else.
- Darby generated the text with a mix of text mining and constrained text structures. The text structures made sure that the chosen text made sense.
- Though not completely based on AI this does give a good base for someone to approach this certain task. For a feel of the book go ahead and interpret the following paragraph-
“In something of red lived an irritant,” it begins. “Safe from the blue from the irr. And this truck went in it. Safe. Something of red in it back to the blue to the red. This truck and something extra. Listen. The nearby something extras in front of the truck. The man in front of the truck trampled from front to back safe from the blue. And all this while the man scooped shovels of dirt and trampled from front to back. The other and the clay sighed for something of red. The irritant lay in something of red and laughed.”
With the foundation of GPT-2 and now the revolutionary GPT-3, the writing industry can expect a major involvement of AI in their work. Creative artists often spend quality time discarding scenes or creations from the final cut. If a decision system is created, the whole manual editing process can be taken away. Another great implementation is seen in the first book of our list. Many find it hard to create a scene in which one can feel the surroundings. The software used in “1 the road” really puts one into perspective! It literally converts the scene into words. With improvements, it could be a challenge for an author to match such a feat.
As seen from the above collection, there sure is a prospect for artificial curated stories. One such example is a poem written by google’s AI. Books by AI is another curated list of books which are all managed and gotten from AI. It goes as far as AI-generated human beings comment on the books there. They act as the critics on the webpage. The beauty presented here is the possible creation of a society that consists of only AI. Who knows, 100 years from now when state of the art AI functions well and needs their own peer groups when they feel bored. Maybe one-day machine learning will tell us the right way to learn machine learning?