Long ago, and not terribly far away, I worked for a company that made some pretty cool software. The software was so cool, it was appealing to certain organizations.
:: TIA ::
One of those organizations was headed by a fine, upstanding citizen named Poindexter, who was embarking on a new project that had been dubbed Total Information Awareness, TIA for short. It was run out of DARPA’s Information Awareness Office (DARPA = Defense Advanced Research Projects Agency):
When public outrage over the TIA project caused the Congress to eliminate the TIA budget, the scuttlebutt among the sales team implied that another funding source would likely be secured. I was not privy to any information beyond generic lunch-room speculation, so cannot I say for certain what happened. I can, however, conjecture and make some educated guesses. The acronym for one of those educated guesses is RKF. More on that later.
Poindexter’s name remained in circulation, and Hicks & Associates, Booz-Allen & Hamilton, and SAIC, all contractors involved with TIA implementation, remained customers.
:: CIM ::
Before the funding cut, we were given a demo of an app by a small company that has since bought out by General Dynamics’ information systems division for a tidy $1.5 billion.
The software was called the “CIM” – Critical Intent Modeler.
Let that name sink in for a bit…
.
.
.
The CIM’s purpose in the TIA context is described as follows:
“Understanding Intent represents the capability for analysts and decision-makers to develop an understanding of the intent of an individual or set of individuals with regards to a future action based on current understanding.”
Here’s a screenshot of the UI, based on a scenario of a suspect who might choose to weaponize a biological agent – it was their standard public demo:
It’s not secret software, you can find other references online, mostly via people’s resumés and consulting company portfolio pages.
The software was originally designed to “reverse-engineer” outcomes as a means of determining their likelihood and to identify those who might play a role in making them happen: start with an outcome (such as a terrorist releasing a weaponized biological agent), then list the criteria necessary to achieve that outcome, then feed in all sorts of data to find out which, if any, criteria have been met and by whom. If a person or group of people meets enough criteria, those people become “persons of interest.”
However, in the way TIA seemed to be planning to use it, it would probably do the opposite – start with a demographic (such as all readers of a certain book), then for each person in that demographic, gather additional circumstantial information to “prove” whether or not the person could eventually pose some sort of threat (such as other books read, the type of work they do, dietary preferences, whether or not they are a Quaker, organizations to which they belong, political affiliations, and so on). If an individual had enough “check marks” in the criteria list, it could imply that at some time in the future that individual might think of some other activity that could be considered threatening (for example, reading Molly Ivens’ book “Shrub,” being a Quaker, and talking critically of the president at a peace gathering could be a gateway to taking some kind of action against the President in the future – maybe). Or as Orwell might have described it, the system would identify those who may be in a stage of “pre-thought” regarding a “pre-crime.” In other words: You haven’t done anything, you haven’t even thought of doing anything, but you’ve got to be monitored, because you could think of something someday.
:: RKF ::
An educated guess on my part is that this is the current moniker for TIA. Again, I have no actual proof, other than my own conjecture based on what I knew of the companies working on the project and knowing about the existence of CIM. So, take it with a grain of salt the size of Utah.
So, what is RKF? It’s “Rapid Knowledge Formation.”
Rapid Knowledge Formation is a program run by DARPA (recognize that acronym?). Its goal is to create an automated system that uses templates to filter through massive amounts of data and determine whether a particular person’s data meets enough criteria to be worthy of scrutiny.
You can see the slides, presentations and more, on the RKF project in all sorts of places on the web:
http://www.iet.webfactional.com/…
(There are tons more, google Rapid Knowledge Formation to see a plethora.)
One of the key elements in RKF is … CIM.
In one of the earliest presentations (link to the IET RKF Kickoff slides), critical intent models are created and updated using Veridian’s CIM. The models are used as inputs for a knowledge base based. Knowledge workers (also called subject matter experts), use this knowledge base for their analysis.
Another place you’ll see CIM is in this document, the TIA Systems Description, on page 17.
In TIA, it’s used for building and updating models – coincidentally, the same thing it’s used for in RKF. Hmmm…
As with CIM, the RKF project’s public presentations focus on biological agent weaponization.
In other presentations, the model templates are referred to as “domain templates.” Biological weapons is one of the “domains” for which templates have been created. Others, exist.
Structured Argumentation represents the capability for analysts to use models, which capture expert knowledge, to reason about potential threats, intent, and the like.
CIMStructuredArgumentationImage