How an AI system can write stories of Game of Thrones?

Recurrent Neural Network

How an AI system can write stories of Game of Thrones?

The final, six-episode season of “GoT” (Game of Thrones)on HBO will be written by an AI (Artificial Intelligence) system.

How does this AI system work?

Developed by Zack Thoutt, a Got fan and a software engineer, this AI system is actually a “Recurrent Neural Network” that is capable of predicting stories based on previously written materials. Thoutt fed the AI system with all the 5 GoT books (total 5,376 pages) and also some popular GoT fan theories. Based on those data, the AI system can generate predictions on what will happen next.

This recurrent neural network based AI system is not perfect. It makes many grammatical mistakes, and it also do not comprehend when one character in the story dies, it continues on with the dead character’s storyline.

But this AI system is a great help for writers, if they want to finish their work ASAP.

Leave a Reply

Your email address will not be published. Required fields are marked *