Ep. 38: Spoiler Alert! AI Predicts Next Chapter in Game of Thrones Saga
Sep 14, 2017
auto_awesome
Joining the podcast is Zack Thoutt, a data scientist from Boulder, Colorado, famous for using AI to predict outcomes in the Game of Thrones saga. He dives into the intriguing use of neural networks to analyze narrative paths in the series. Zack shares quirky results and unexpected character fates, like Cersei's demise and Ned's survival. The discussion also highlights AI's role in creative writing, exploring how it can help generate endless narratives while facing challenges in coherence. Plus, there's a fun look at fan theories surrounding the series.
A data scientist successfully employed AI to generate text for the unfinished Game of Thrones book, highlighting AI's potential in creative writing.
Despite its capabilities, the AI's inability to maintain long-term plot consistency underscores the challenges faced in machine-generated storytelling.
Deep dives
Using AI to Write the Next Game of Thrones Book
A data scientist has utilized a neural network to generate new text for the unfinished Game of Thrones book, The Winds of Winter. By training the model on the existing text from the A Song of Ice and Fire series, the AI learned to recognize patterns and structures in the writing. For instance, the developed system uses unique identifiers for around 32,000 distinct words found in the series, allowing it to produce text that mimics the style of George R.R. Martin. Although the generated text lacks perfect grammar, it reflects the characters and tone of the original works, offering a novel exploration of AI in creative writing.
The Challenges of AI Text Generation
Despite the impressive ability of the AI system to generate coherent text, it struggles with maintaining long-term plot consistency and context understanding. For instance, while the text generated may reference characters like Jon Snow and Sansa Stark, it does not accurately denotate their significant plot developments, such as character deaths. The neural network's primary function is to predict the next word based on previous sequences, leading to amusing yet disconnected narrative outcomes. This limitation highlights the gap between AI-generated content and the intricate storylines crafted by human authors.
Future Applications of AI in Creative Writing
There is potential for AI models to assist writers and artists in various creative endeavors beyond merely generating text. One possible application is utilizing AI to identify plot holes or inconsistencies within stories, thereby helping human authors refine their narratives. The integration of AI in artistic creation may foster innovations where human creativity and machine-generated ideas coexist. Although this technology is still in its infancy, its growth might change the way narratives are constructed in the future, ultimately enhancing the writing process.
This might be the first time we've been able to say this, but we might have spoilers ahead on this episode. Joining us we have Zack Thoutt, a data scientist and a developer from Boulder, Colorado. Zack has done something we've all wanted to have done: he's working on finishing the books behind HBO's Game of Thrones - a Song of Ice and Fire - by putting an AI system to work with sometimes comical results.
Get the Snipd podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share & Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode