LLMs are context-less. #promptengineering is nothing but making sure that uses provide enough context in their questions so that the LLMs actually knows what’s what.
And that’s how Subtxt works.
One of the first things I teach when it comes to narrative is that context is meaning. Or, from another angle: life is meaningless without context. Same with stories.
If you want a meaningful story you need to set the context.
And that’s what Subtxt does in the background with its “prompts”: converts author’s intent into context.
(And most importantly: holds that context)
