When do you need Augmented Decision Intelligence?

A primer on difficult decisions

Adam M Ross
The Tradespace
5 min readOct 10, 2023

--

We all make decisions. Countless times a day in fact.

Should I get out of bed now?
What should I eat for breakfast?
Which route should I take to work?

Most of these decisions are made quickly and without much fanfare or regret. Even so, the world we live in is complex, filled with more information than a human could possibly comprehend. For everyday decisions, not thinking about every aspect of the decision is usually fine.

But sometimes we need to make decisions that have consequences. I mean real consequences, like making a large purchase
ever buy a car? … a house? … a boat?

or choosing a direction for your life
ever pick a life partner? … a college to attend? … a city to live in?

Beyond the personal, many of us work in organizations which make decisions that commit large sums of money:
… portfolio investments?
… acquiring a company?
… a new product line?

large amounts of effort:
… designing a new aircraft?
… spacecraft?
… aircraft carrier?

and even lives:
… troops for peacekeeping?
… disaster relief?

How are these consequential decisions made in a complex world? They are made by abstracting the world and focusing our attention on part of the problem in order to simplify the decision. Sometimes people rely on intuition or “gut feeling” to do this. Unfortunately, limits of our minds can result in substantial neglect if we are in a complex decision situation. These limits are a primary motivator for people to use decision aids, which are intended to prevent neglect and provide traceability for decisions (both very important for organizations!).

So how do we know when we might need decision aids? Here are five factors that scream “help me please!” if they show up in a decision you are facing: nonlinearity, dimensionality, interactions, delay, and circumstances.

Most of us are better at linear extrapolation in contrast to nonlinear relationships which abound

Nonlinearity. Humans are great at extrapolating from limited information. Most of us default to linear extrapolation, which is often a pretty good approximation. Except when it isn’t. Linear extrapolation means future results can predicted using a straight line from previous results. For example, if I made $5 yesterday and $6 today, I expect $7 tomorrow and $8 the day after and so forth. Our savings accounts do not work this way, however, where interest is “compounded.” This means that future interest is dependent on our current balance. As interest is added to the balance, future interest grows faster than linear since the balance is continually growing. This “miracle of compounding interest” is great for savings accounts and terrible for loans. If your problem has nonlinearity, you will benefit from aid since we are terrible at nonlinear extrapolation.

If there are many factors, you may only be able to think about the top 3, potentially neglecting important info

Dimensionality. Research claims that humans can only think about approximately 5 chunks of information at a time (maybe 7 or 8 if you really practice, but many are closer to 3). If you have more than that number of factors in your problem, you are likely only thinking about a subset at a time. If the neglected information include factors that you care about, you may be missing a large piece of the problem and subjecting yourself to future regret.

Sometimes having separate components is NOT the same as having them together

Interactions. It is often difficult for humans to predict outcomes when factors interact with one another. Is eating a deconstructed peanut butter and jelly sandwich the same thing as eating a constructed one? What about any recipe for that matter? The adage “the whole is greater than the sum of the parts” specifically applies to this concept. When interaction occurs, it can be difficult to guess the outcome. There is a reason a chef ought to taste the meal as it is prepared.

When effect doesn’t immediately follow cause, the delay may prevent understanding what has happened

Delays. Human conception of cause and effect is based on getting feedback. Someone throws a baseball at a glass window and the window breaks. What if the window didn’t break immediately? What if the window suddenly broke the next day instead? The next week? The next year? Would we know that the baseball had caused it? When there are delays, it prevents us from understanding cause and effect, and therefore our ability to predict outcomes.

If situations change, our expectations from prior experiences may no longer hold

Circumstances. As humans gain experience in their personal and professional lives, they build wisdom. This wisdom often takes the form of heuristics, or rules of thumb that we’ve adopted to predict the future. These heuristics work because they tend to be true. But they are only true if the circumstances in which they are true still hold. It is hot in Florida in the summer. Of course! It is cold in Maine in winter. Yes! It rains 8 inches of rain in Manhattan in fall. It does? It did this year. As climate change happens, some of our weather heuristics might need to be updated. So too do our heuristics more generally. If circumstances are atypical, heuristics may not hold, and outcomes may seem less predictable.

Given one or more of these five factors could throw a wrench in your decision making, what does aid look like? Augmented Decision Intelligence (ADI), which underlies Diakronos’ tools and techniques, is a way of overcoming human limitations so you can handle the five factors above. ADI leverages the Core Concepts of Tradespace Exploration to organize and manage complex decision situations. Using ADI through computation and interactive exploration, you can have greater confidence that you are not neglecting essential information for your consequential decisions. Augment your decision intelligence by checking out our growing suite of ADI-enabled software and professional services, including EpochShift.

--

--