Ted Goranson - Personal Blog

I hear that a lot. This is absolutely the case. It is not a negative quality and one I want to embrace. In our pitches, we sometimes mention that the roots of the technology go back to the eighties and that the original challenge was posed (in the intelligence research community) in the 60s.

There is a very long legacy for what we are doing; there has been a whole lot of thinking, testing, failing on this — all out of the mainstream, all in the black. What we have are a combination of actual novel never-been-seen-before inventions, combined with some hard-won engineering approaches and knowledge that reduce the possibly of failure in ambitious applications. I'll diagram them at the end of this post.

We can apply this work in a large number of domains. We welcome those who want to use it wherever they wish (within moral and constitutional constraints).

But for me and my crew, I want to work with cinematic narrative because I love it and it gives me energy.

And we want to work in the biomedical research area, because it gives my life purpose.

Beyond that, we can adapt for business realities and similar constraints, but we have to always have fun.

This is absolutely the case. And that's a good thing.

Our story is that we can do stuff no one has ever done before. We know — or at least we know better than anyone alive — how to do some of those things. The normal pitch addresses some market need. We don’t.

We’ll invent the market and no one at all fully knows what will emerge when we open possibilities. The idea here is that if we succeed in allowing people to do something enriching that they couldn’t do before, we can find a place in commerce.

We do have two niches we want to address.

• there has to be a better model of social than what we have today. We want people to be able to collaborate (even with strangers) to create something of value.
• there will be an exciting future for cinematic narrative, as technology influences creation and consumption and the inevitable revolution in the business models force change.

I think what will happen is that different editions of a 'plan' will be generated, triggered by one potential partner or another. If you are watching from the sidelines, don’t take any of these as the business plan.

This is absolutely the case. I hardly understand it myself.

That’s part of the thrill, the absolute high you get by working with these ideas. The approach we are taking is to not put the underlying technology in front when we talk to potential friends. What we present is what can be done. Sometimes we talk to people who are astonished at the possibilities; others have heard multiple hucksters make similar claims.

This is why we start with the user interface ideas and what it allows you to see and work with. Situated reasoning is hard to grasp until you do it. So let's just do it.

I don’t believe we are, and am willing to bet the farm on it.

Let’s talk about that a bit. Useful wisdom is that things need to be simple to engage with and intuitive to understand. Some of our illustrations look cluttered and massively hard to understand. Fair enough. Some of the papers we have presented assume a different type of user than we will support in the web business, or native app for the second screen.

And some of the use cases are fairly rare and require skilled users. This is all by way of us describing capability to some user community or another. After all, some of this work was done for people who have nearly impossible jobs and they spend all day in front of our tools.

Our Strategy

We will rely on our initial users; we had a vital community of advisors in our earlier filmsfolded work. We’ll recreate that group and empower them with guiding what works for initial users and widespread use without compromising on the future. There is no better way than to trust your engaged users, while keeping the vision.

We offer greater power and new possibilities. We can find a balance that works.

So-called elevator pitches are problematic. I suppose they are necessary; people with resources get bombarded with all sorts of opportunities. Most of these are business ideas and most of those can be sufficiently evaluated in a few minutes.

But for us in our current situation, they're tough.

In my past lives, I'd be the one to blow someone off if they didn't have the time and skills to invest in understanding what we do. By definition, it is different or I wouldn't be doing it. That means rules of thumb don't apply. And you can't say something like "Oh, I get it. It's OkCupid for dogs."

We've decided to not talk about the technology, and instead talk about what it can do.

And then we've decided to not show the multi-world outliner, because to understand why it is cool, you have to understand what it is showing that is novel behind the scenes. Instead, if you were going to see one thing, you'll see the scrubber.

Now the problem is that until we get a real working setup, all the pictures look cluttered and complicated. It looks like we are serving a need that few people want and even then it will be too hard to work with, compared to say sports annotation software (with its scrubber).

Three Components

One is a frame preview, showing the current frame. The novelties here are that we can draw directly on it to highlight objects, environments and cinematic effects and have those serve as anchors in the film. We also have some novelty in how we can optionally place the preview image in the space-time scrubber to save space.

A second component is an ordinary scrubber with the only improvement that it has a shrunken image of the space-time scrubber on it.

In our simplest layout, the ordinary scrubber is at the bottom, the preview frame at the top and our space-time scrubber between. All of our advertised novelty is either explicit or implicit in this space-time scrubber.

The job of that scrubber is to allow a user to quickly move through a film and see content of the film, added drawings and movement on the film, and a variety of annotative anchor types on the film itself. These include dynamics as ephemeral as you want, or choreography as explicit as can be.

I'll put all the images and demos we make in the growing post for that purpose.

We do have a challenge in that so many different types of anchors can be displayed, and the film itself has a lot of information depth even without adding distractions. We intend to meet that challenge by very aggressive situating of the user. Some of this will be our system assembling the information in the way we know our user prefers, showing the annotations he/she is interested in. Part of it will be fluid control by the user over what is seen. There's a fine line between clutter and insight and we hope to allow each user to find his/her own balance.

In the meantime, to a prospective partner in an elevator, it will look like needless clutter.

Elsewhere in this blog, we’ve started publishing bits and pieces. Some — even some friends — have looked at one or two of these and think they understand, but they cannot without looking at the big picture. Here is a helpful overview in outline and chart form. You can see the chart below or download a largish PDF.

A Situated Reasoning System

We have a situated reasoning system. The system itself is an engineering solution that has to be implemented in and interface with what the world uses, so overall it is ordinary in how it puts pieces together.

Some of those pieces are internally engineered based on really hard-won engineering experience, but aren't particularly radical breakthroughs in the science. Those have to do with abstraction, representation and messaging.

On the other hand, some elements of this are big advances in the science and what makes it all possible. This idea of using explicit category theory to implement situation theory using narrative dynamics is one thing we really are selling. Another is using the symmetries of the dynamics for an approach to to difficult large-scale problems in heterogeneous reasoning.

There is no reason not to make the difference between innovation and engineering clear to partners. Some of this is real invention, some is relatively unique engineering knowledge and some is just ordinary (presumably) good execution.

A Video-centric User Interface

The novel bits here are:

• Considering elements of a film as situations, including the spatial and cinematic vocabulary. In the context of the scrubber that is just a value, until they are tied to real representations.
• The Space-Time scrubber's ability to display many types of annotative anchors in a situated context. This is inventive.

The rest is just engineering.

A Situated Knowledge Editor/Browser

(We'll fill this bit in later. It will have to wait until we have some explanation in the main post.)

A Diagram of All the Major Bits

(Image) Which bits are novel and which are informed execution.