Conversational AI + Interactive Story Structure: The Basics

As a follow up to queries with regard to “how is the contentAI platform different?” – The basics are that the contentAI platform converges conversational AI (Natural Language Processing “NLP”), machine learning, pattern recognition, session management and personalization with a “STORY STRUCTURE.”

“Story” means that we start at “A” and end at “Z.” The path through may be non-linear, but, the User is guided to progress through the engagement and reach one (often of many) possible endings.

For mobile Users, this is ideal. The application has contextual awareness of the user (physical location; item they are holding, etc.) and there is an “awareness” of the “scene” that is likely to be played out between the User and the application. This means that the AI characters do not need to be “generalists” or be able to carry on random conversations or answer obscure questions — they need to “motivate” the User to get through the story/engagement.

This can be as simple as providing “item location” data plus providing a relevant mCoupon, or as complex as delivering transmedia story elements from a motion picture to the mobile engagement.

The Evolution of Virtual Characters – It’s About the Creative

Having received some interesting comments from Agency “Creatives” about how they didn’t like “bots” since they couldn’t be controlled, it got us to thinking about how the perception of ‘bots’ is extremely “dated” at this point.  People tend to think of them as the chatbots on AIM 5 years or more ago.  Chatbot technology has evolved, with a quest to auto generate responses, typically without any context or restriction of theme. We’re all for advances in conversational AI technology, but, we’re realists about what is a viable and valuable commercial product today.

Working with Brands, Agencies and writers, we create complex virtual characters who are “motivated” to progress the conversational experience within a story structure.

Now, that structure is different from a film or television structure (with similarities as well, staring with “beginning/middle/end”), as the conversations may branch and the user input can personalize the character’s responses, feeling improvisational.

But, end of the day, the virtual characters we create are very much like a film or television character.  Or, think of them as characters in animation, created by humans, yet, seemingly with a life of their own.

Part of the magic comes from the Natural Language Processing features and capabilities.  Yet, again, this is a converged effort of technology and subjective human creativity.  Yes, even writing Regex is a creative skill; where a writer envisions the “other side of the conversation” (the other Character, being the human/customer/user).

For “Creatives,” our platform offers opportunities to control and create stories that not only “touch someone,” but, which someone directly participates in.

Retail, Marketing and Entertainment – Storied Engagement and Virtual Characters

After a couple of weeks of being very focused on technology, platforms, compatibility, integration, pricing. . .

Well, it was refreshing to reach a LINKEDin Discussion today from @GavinJohnston a business anthropologist, who reminded us of “why” we created our platform and “why” we enjoy the work so much.

It’s about story and entertainment.  Sure, it’s about a lot of other things as well.  But, at it’s heart, we have a story and virtual character engine that’s, well, fun and entertaining.

These lines stood out:

“…entertainment and a memorable in-store experience probably have more to do with a sale than the product…”


“The retail environment is essentially an expansive, immersive media platform”


“The products may mean less than the underlying value of “story” being told.”


This is not intended to demean the value of the product – particularly, once it’s been purchased, used and enjoyed.   But, the in-store experience for customers is not always logical and rational.  It is emotional.

Emotions are best reached through story.

UPDATE: The discussion thread resulted in a follow up from @GavinJohnston that should be included here:

Increasing sales revolves around more than getting people in the store, it involves getting them to think of the store as a destination and thinking of it as a “Place” rather than a “Space.” Place comes into existence when humans give meaning to a part of the larger, undifferentiated space. One of the most affective ways to do this is to incorporate people into an experience and directly involve them in the story.



What’s in a name? Mobile Virtual Agents, Mobile Bots, Mobile Characters. . .

What do we create?

There are a number of Industry or common terms that are used to describe different features of our product.  Depending on your business, you may refer to an automated AI character by a specific name such as ‘virtual agent.’

Here’s the basic breakdown and difference the contentAI platform delivers:

* Our “characters” are motivated to help direct a User to various “end points” in a session

*  In many respects, a “session” is similar to a scene in a film.  Only, with different possible endings

*  Our “characters” are NOT “passive chatbots” waiting for User’s to guess what to ask.  Our “characters” take a more proactive conversational lead

*  Our “characters” can access deep repositories of information, data and even ‘character personality traits” to deepen the conversational experience

*  We refer to our “characters” as being part of an “Interactive Story.”  That refers to the Characters taking the conversational lead and driving the conversation forward.

*  Because humans are humans, they have a certain curiosity about “bots” and try to trick them or test them.  Our “characters” are all designed to humor and entertain humans in this regard — But, then, get them back to “story.”

We’re not particularly partial to any of the standard names for “bots” (and, don’t like “bots” much either).  “Virtual Agents” or “Virtual Guides” works well for many commercial applications.  But, we do think of each “bot” as a “character” that is constructed in the same way film and television characters are constructed.  Only, they have to improvise, but, always “in character.”

Converged Media – Interactive Narrative, Video & Images

The original impetus for creating an interactive narrative engine, was to bring User Interactivity to the story side of internet television.

We’ve expanded upon that original vision — a lot.

With the emergence of mobile Apps and the iPAD, along with our stand-alone interactive narrative work, we’ve started taking the steps to integrate video, imagery and interactive narrative where the personalized “user preferences” are shared between those diverse elements.

While this is in Alpha, without the sharing infrastructure in place — we are able to get a strong sense of the User Experience vis a vis a User’s manual engagement with converged media elements.

We’ve opened up a page, based on our favorite example of “CapnWag,” here:

http://contentAI.com/capnwag

Watch the video, start the interactive narrative, and venture into the imagery. Now, imagine when this seamlessly happens within an App…STAY TUNED…

31 March 2010

Platform Launch – Interactive Story Demo(s) Alive and Kicking

Our bright new shiny story-engine came out of QA polished and ready to go.  The new features are remarkable, but, instead of talking about them, it’s best to take it for a spin.  We’ll be adding SMS and Twitter integration soon, but, for the short-term we’re using Gtalk and AIM IDs and chat clients; with a roadmap to quickly add WAVE to our network (which will allow the ability to push files within story conversation).  Please contact us by email for access to the demo(s).

22 January 2010

New Year – New Interactive Story Platform Release

We had to make some tough decisions in Q4/09 — We saw the continuous need to quickly integrate with 3rd Party partner platforms; as well, we needed a path to incorporate 3rd Party Speech Recognition and text-to-Speech capabilities sooner than later. We needed our platform to anticipate integration, not struggle through adaptation time and again. We needed to provide Clients with the ability to scale instantly.

That meant undertaking a “Page One” rebuild…which meant a lot of transitions…

New team members, new hardware, code, and the Cloud – Coupled with fresh thinking — to do a round-the-clock, over the Holidays, entire rebuild that provides amazing stability and scalability.

Welcome to the New Year. While we had to set aside certain Content development during that time-period, but we’re now in full-swing and will be releasing new Demo content in the coming weeks…and, beginning our integration with Partners for new delivery channels.

Happy New Year to All…

8 January 2010

Conversation is Content

Conversation is Content.

That has become the “elevator pitch” opener for us.  We’re in the “conversational content” business.

We create new ad formats based on conversation.  We can provide eLearning tools through conversation.  We engage users through conversation and 50% of the “content” is User Generated (otherwise our AI Characters are just talking to themselves; which, actually occurred yesterday during a re-boot, as our Twitter Character thought “she” should respond to the “temporarily unavailable” message on her account — while curious, we cut off her “talking to herself” after a few exchanges).

We’re not the only folks seeing this as an important and emerging space, noting “conversation as content” posts  HERE and HERE.  The background Big Picture is nicely presented HERE, which presents this as a near epoch making shift in human::machine interaction.

It’s a fascinating area to be constructing and facilitating Content, where value is added to all sides and terrific User Experiences are paramount.

24 July 2009

welcome to contentAI studios

Why “content AI?”

The term “Content” has become synonymous with film, video, audio and written media materials.

We chose the URL and Brand “contentAI studios” to expand the definition of “Content” to include conversation.

Conversation between humans (User Generated Content, “UGC”) and digital characters powered by Artificial Intelligence engines.

Welcome to the intersection of UGC and digital characters. 

Where the User becomes part of the story making process.

8 July 2009