How can I leverage artificial intelligence and virtual reality to create intelligent automatic story generation?

Question

How can I leverage artificial intelligence and virtual reality to create intelligent automatic story generation?

My idea is to come up with a system that is empowering the user’s imagination to create and experience stories that engage and immerse them. So they can have the control over the story creation/narrative?

A system that could enable the following use cases:

1. We sit down in our entertainment room. Select a genre from a menu. Pick the character types for the cast. Decide on a generic plot or give the system the option of picking a random choice. Sit back and enjoy.

2. Let's say, I want to watch a sci-fi movie set in Paris with Marlon Brando and Jim Carrey. Then the system instantly generates such a movie.

It has come to my attention that SoarTech developed a pattern-of-life capability for the U.S. Army Simulation: Agent-based Patterns of Life for Virtual Environments.

That the University of Twente in the Netherlands has been working on a project called the Virtual Storyteller.

Here is a research paper entitled The automatic generation of narratives.

And more recently that IBM has developed a speech recognition sandbox.

I know that the answer probably requires tech, AI, speech, VR and AR, and Watson integration specialists, but any help is welcome.

Watson probably won't be a necessary component of the solution, as long it is an efficient voice recognition software with speech to animation features.

Answer 1

The VR and computer speech aspects are entirely corollary. Adding them in would be relatively trivial in comparison to creating an algorithm that can dynamically generate stories of interest to humans. Essentially, aesthetic components not related to story structure (images, sounds, speech) are "window dressings".

A story generation algorithm would have to focus on the structural elements of storytelling, possibly in conjunction with human-user reactions, and generate narratives narratives dynamically.

Such a system would be validated by human reaction to the stories, and the validation would not require graphics, sounds or speech. Think of a movie or video game with a story as the expressions of stories that originally had textual form.

Game theory could certainly be applied to this process--creation of narrative in particular utilizes game theoretic concepts, consciously or unconsciously, in terms of strategies to engage reader/viewers and provide "emotional payoffs".

It should certainly be possible to develop an algorithm with these capabilities, even where the algorithm doesn't not really "understand" the content it is generation.

A similar endeavor might be found in Cameron Browne's work on algorithms that can create simple, combinatorial games of interest to humans: Evolutionary Game Design.

It's important to note that the game created by Ludi are based on pre-existing mechanics, and algorithms don't yet have the capability to make profound leaps in the way that human game designers can. (In other words, it will likely be a very long time before an algorithm can create a game like Magic: The Gathering where no such gaming system previously existed.)

However, this doesn't present much of an issue in the context of procedural narrative generation in that the most successful literary work in terms of audience/sales tends to be based on pre-existing formulas. So an AI that can merely reconfigure elements in a formula should be able to generate narratives of interest to the average human.