by Sam Culper III
I’ve been reading a great book about startups in which the author talks about validated learning. (As an aside, each of your intelligence preparations should be treated as a startup company. Instead of input costs you have study time, and instead of revenue you have information. Is your study time contributing to the flow of intelligence information into your company? Let’s work together to solve that problem.) The concept of validated learning is that ideas need to be tested and the results measured in order to replicate (identify then re-achieve) the most effective courses of action. It’s a very structured approach that isn’t just about being able to learn something, but about being able to identify what you were able to learn, how much time it took you to learn it, the resources required to learn it — what we’d call the unit economics of learning. You’ve identified all the time and resource inputs of learning, so you can project what future time and resource inputs will be required to learn something else (or to teach it) in the future.
I’ve been a big proponent of a good scholarship of history and war (and intelligence). This is an extension of validated learning. In a scholarly approach to warfighting, the input and results were measured, and a) warfare progressed because doctrine was improved, b) the doctrine worked as expected because it was based on previous validated learning, or c) the doctrine failed because the enemy adapted to that doctrine. We don’t just select a course of action because it sounds good, but we make judgements and estimations based off what’s been previously measured and deemed successful, or based off known adversary vulnerabilities. Finding, knowing, and never losing the enemy is a chief task of intelligence so accomplishing that task with the greatest efficacy should be a top priority.
