Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Breaking Bad and Coming Good: Computer-Generated Stories of Change


Published on

Stories move us emotionally by physically moving their protagonists, from place to place or from state to state. The most psychologically compelling stories are stories of change, in which characters learn and evolve as they fulfil their dreams or become what they most despise. Character-driven stories must do more than maneouver their protagonists as game pieces on a board, but move them along arcs that transform their inner qualities. This presentation describes the Flux Capacitor, a generator of transformative character arcs that are both intuitive and dramatically interesting. These arcs – which define a conceptual start-point and end-point for a character in a narrative – may be translated into short story pitches or used as inputs to an existing story-generator. A corpus-based means of constructing novel arcs is presented, as are criteria for selecting and filtering arcs for well-formedness, plausibility and interestingness. Characters can thus, in this way, be computationally modeled as dynamic blends that unfold along a narrative trajectory.

Published in: Technology
  • Be the first to comment

  • Be the first to like this

Breaking Bad and Coming Good: Computer-Generated Stories of Change

  1. 1. Tony Veale, @MetaphorMagnet Modeling Metaphor Blending and Change for Novel Ideation
  2. 2. * * see also EC coordination action: PROSECCO: PROmoting the Scientific Exploration of Computational Creativity This work forms part of the WHIM project.
  3. 3. I’ve only been a bug since 8am … … and I’m already very annoying! “As Gregor Samsa awoke one morning from uneasy dreams, he found himself transformed in his bed into a monstrous vermin.” So starts Franz Kafka’s novella of transformation, titled Metamorphosis, in which the author explores issues of otherness and guilt by exploiting a character’s horrific (if unexplained) change into an insect.
  4. 4. I used to be a “people person”… … now I am just an animal lover! Authors from Ovid to Kafka show the value of transformation – physical, spiritual and metaphorical – as a tool of character development. Metaphorical contrasts become literal changes!
  5. 5. Change happens in stories – good stories at least – for a reason. It is typically an external expression of an internal conflict.
  6. 6. From caring husband & teacher to ruthless drug-baron: Walter White is an unknowable unstable blend of categories 1. 2. 3. 4. 5. 6. 1.
  7. 7. To understand how people & things change over time, we need a rich property-level & behaviour-level model of world concepts. For we want to deconstruct & pull our concepts apart, to reassemble them in new ways that create interesting contrasts. P1 P2 P3 P4 P5 P6 P8 P7 P9
  8. 8. Linguistic constructs like the simile frame “as ADJ as a NOUN” allow us to harvest knowledge of stereo- typical properties from the Web.  These are from the Google n-grams.
  9. 9. We can use metaphors to understand how properties align and change, so that one concept can virtually become another. Creative metaphors are commodities that computers can generate at Web Scale via CC Web Services!
  10. 10. Metaphor Magnet suggests metaphors on demand for given topics or topic pairs. It finds conventional metaphors in large Web corpora and elaborates them via rich stereotypes.
  11. 11. Metaphor Magnet will even generate poems (ahem) for your metaphors on demand.
  12. 12. Metaphor Magnet uses Web n- grams to appreciate which concepts are best imagined as the start state of a transformation, and which are better suited to being the end state. The goal is to produce character arcs that are jarring for the right reasons!
  13. 13. businessman millionaire tech geek pauper If we take @MetaphorMagnet’s metaphors seriously, as literal statements of becoming, we obtain What-If scenarios in which our story characters undergo surprising but apt changes. IN
  14. 14. { } Property-level transformations: dashingdrab happymiserable richpoor richskint ostentatiousunpretentious pamperedpoor privilegeddestitute privilegedpoor richbroke richdesperate richdestitute richmiserable spoiledpoor wealthybroke wealthydestitute wealthypoor So what if a millionaire … … became a pauper? A Lexical Database of Antonyms (e.g. WordNet) allows us to align the contrasting properties of two stereotypical representations.
  15. 15. Millionaire maddened by mansions, walks into workhouses Millionaire rejects richness, hungers for hopelessness Millionaire goes from being apparently dashing to increasingly drab Millionaire's fake happiness masked true hardship Millionaire goes from living in mansions to living in shacks and slums Millionaire goes from living in ornate and luxurious mansions to living in wretched and Spartan slums Millionaire goes from living in magnificent and swanky mansions to living in miserable and wretched slums Millionaire goes from ostentatious to unpretentious Millionaire's private pampering masked by outer poverty Millionaire's artificial privilege masked genuine deprivation Millionaire's public wealth masked pervasive poverty Millionaire's false wealth masked genuine deprivation “Story pitch” Summarization via Automated Headlines
  16. 16. A twitterbot using the Metaphor Magnet Web-service to tweet a new hard- boiled metaphor every hour!
  17. 17. A chain of causal triples can yield an interesting scenario when it shows how a familiar concept has quite unexpected consequences.
  18. 18. Causal chaining allows @MetaphorMagnet to blend the norms of two concepts with differing affect profiles to achieve a surprising what-if whole. Consider Artists: (I used to be an artist!) By chaining KB triples together, a machine can observe that artists typically cause a positive effect on the arts. Seeing that critics can have a negative effect on arts, and dictators have a negative effect on their critics … … @MetaphorMagnet reasons that dictators can have the same positive effect on the arts as artists! This is sophistry as ‘critic’ is used in two differing senses, but in this what-if blend, Artists become like Dictators!
  19. 19. When two concepts cause the same effects (after a little sophistry), we can easily imagine a story in which one becomes the other, deliberately or unintentionally, with dramatic consequences. I love the smell of narrative in the morning!
  20. 20. We are currently evaluating the packaging of ideas in @MetaphorMagnet using Crowd-sourcing (against a non- creative baseline @MetaphorMinute).
  21. 21. For more comix on creativity: