An organizational intelligence environment should...

The principles below guide our work. Ambitious as they are, we have already built a prototype — an intelligence-software environment that demonstrates how they integrate and function together. It will evolve as we engage with the people and organizations interested in exploring it with us.

Our Approach

FIVE: Prompt people to calculate or estimate the magnitude of each factor and its importance relative to the others. This requires that they function as both conceptual and computational visualizations.

SIX: Encompass both the challenge-and-opportunity environment we are addressing and the organizational strategies and tactics we use, and will use, to address it. This lets us map our response to a situation directly onto the situation itself.

SEVEN: Provide robust data storage, access, and engineering — ingestion, preparation, filtering, and querying, including mathematical operations and machine-based learning and intelligence — with the capacity for real-time responsiveness during "what-if" scenarios.

EIGHT: Modularize machine-based intelligence (MBI) operations for both training and tasks. Broad, open-ended prompts cede human oversight and produce "black box" outputs that are difficult to verify before implementation; modularizing MBI by factor, or by groups of factors, equips and protects human control.

NINE: Be a world we want to inhabit. Systems thinking is demanding; the environment's aesthetics and use of space should catalyze focus and invite exploration. Like a fine studio, writer's cabin, or retreat, it should be made inviting — an aim served here by nature-inspired design, or biomimicry.

TEN: Support synchronous and asynchronous collaboration alongside persistent, ubiquitous communication. Those closest to the issues a model represents should be able to contribute their thinking directly into the same model, in the same environment, as the leaders who normally steer organizational intelligence. This optimizes the value of an organization's entire workforce, supports buy-in, and strengthens alignment.

Taken together, those first four principles address our core observations and rest on bedrock ideas in visualization. Chief among them is Shneiderman's Mantra — "Overview first, zoom and filter, then details on demand" — named for the visualization pioneer Ben Shneiderman, along with the integrated-annotation principles of Edward Tufte, a foundational thinker in the field.

ONE: Engage the brain's most powerful faculties for perception, interpretation, comprehension, and memory: its visuo-spatial systems. Our minds evolved to learn and remember spatialized information so that we could navigate our surroundings and meet our needs. It is why memory competitions are still won using the method of loci, or "memory palace."

TWO: Prioritize the visual modeling of our thinking: the understanding we draw from data, from intuition cultivated through experience, and from the voices of stakeholders. It is what we ultimately act on. Visuali:zed data must be available in an instant to support and illuminate every factor, but it should not take center stage, as it does in current systems.

THREE: Show us everything at once — all the factors, every single one we can imagine— from those that are quantified and those that aren’t. By seeing it all, we can discover both negative and promising synergies — the spark of creativity. The most ephemeral and ambiguous factors often derail the best-laid plans. But it must be structured to avoid cognitive overload. It is the opposite of reductive thinking that has dominated credible thought for centuries.

FOUR: Let us examine any single factor in the greatest detail without losing sight of the whole — moving through and around a comprehensive model as easily and as gracefully as a hummingbird in a garden.