All day, every day we make decisions. Apparently, 35,000 of them! If the first decision you make everyday is to hit snooze on your alarm and the last decision you make is to put down your phone and sleep, that’s a lot of decisions in 16 waking hours. In fact, in the 7 minutes it takes to make a pot of coffee you make 255 decisions. It's no surprise that making decisions can be exhausting.

Obviously, not all decisions are created equal. In the scheme of things, does it really matter if you watch Netflix or go to the movies?

But big decisions matter. The kinds of decisions we only make occasionally but have consequences for years to come.

Bad decisions bring regret. Regrets tend to be worse when poor decisions have social consequences, bring lost opportunities, or lead you further away from the person you aspire to be. Some of the worst regrets come from doing something that breaks up a relationship, spending time on things that turn out not to matter, and failing to change destructive patterns of behavior.

But small decisions matter too because each one accumulates into bigger influences in our lives. Many of these small decisions are better thought of as one of a class of decisions. For example, the cupcake you want to eat instead of a carrot may not be the straw that breaks the camel’s back, but if you make that choice every day, you’ll fail to meet your health goals.

In between the big life decisions (should you get married, should you take that new job) and the little life decisions (which product should you buy) are the decisions that matter in the context of a workday, or a project, or an important endeavor. They can be complex and multifaceted, and their consequences can be highly uncertain.

Then along comes data, promising to show us the way and make decisions easier. Data is tangible, objective, solid, fact, incontrovertible. Data is truth, right?

You may think that data is the best representation of truth. After all, data is measurable and quantifiable. But quantifiable isn’t the same as accuracy and certainty. The premise of the analytics movement is this: there is only one optimizable answer for any given problem and humans won’t know (or be able to work out) that answer because our judgment is fundamentally flawed.

This, quite frankly, is nonsense.

Big data and artificially intelligent systems predict the future from patterns in past data. To believe that data has the answer to everything is to believe the world won’t change. But the world will change. In fact, it’s changing faster and is far more complex than either artificial intelligence or human intelligence can grasp.

You may be thinking, Wait, what? I don’t make decisions with data or artificial intelligence! We would posit that you do.

Data is simply what we choose to pay attention to. Data could be house sales in your neighborhood, a like on your Instagram story, or a massive database of numbers that you encounter at work. Data is the person you swiped right on Tinder last night.

All day, every day our decisions are influenced by AI. Algorithms guide small decisions, like what to watch on Netflix, and large decisions, such as comparing house prices online.

We would go so far as to say that all decisions are data-driven simply because, at some point in your past, you’ve absorbed information from data. Your intuitions on how to get to work are influenced as much by your experience of Google Maps as by the physical experience of sitting in a trac jam or of choosing to take the train instead. Data changes our brains. Our intuitions are data-driven, and, increasingly, AI-enabled.

We humans are a remarkable feat of evolution. We evolved to make complex decisions with the limited computation capabilities of our brains. But the world is far too complex for any individual to understand completely. So we evolved shortcuts in the form of intuition and biases that allow us to make good enough decisions in complex situations.

A human who needed thousands of examples of a lion hiding in grass was a dead human. So we evolved to learn from a small number of examples. Learning from a small number of examples means that we have to generalize to reach a conclusion. These limits on data introduce bias—shortcuts in our reasoning which draw from many disparate sources.

We also evolved cultural mechanisms to allow us to collaborate and make decisions with each other. We can only live inside our own minds, so we have language, institutions, and customs to extend beyond our individual life spans and computation capacities.

Our modern world of big data and AI has allowed for immense progress and opportunity. But it has also abstracted us from a culture of working together and challenged our human agency to make choices and decisions. Human decision-making is a remarkable achievement, but making accurate decisions is still very hard.

Here’s the thing...our minds don’t necessarily do well with data.

For humans, reality is defined by space. Our sense of what is entwines inseparably with our sense of space. We inherited some mathematical skills from evolution—we can distinguish basic geometrical shapes and know that numbers can be ordered along a line—but without formal education in mathematical structures, our intuitions do not match reality.

Very young children and unschooled adults do not understand that all consecutive numbers are equidistant. They think large numbers are closer together than small numbers. One and two are further apart than eleven and twelve. Even after correcting this intuition, we do not crunch numbers like a machine. Instead, we use motion along a number line. It takes longer to calculate 9 - 6 than 9 - 2 because our brain’s algorithm moves along a line. It takes longer to take 6 steps than 2.

Abstract quantities such as numbers and data violate the principles of our spatial reasoning, which is why they can be so hard to think about. But think about data we must. Which means we need strategies for making decisions when we have our own intuitive knowledge, knowledge from others, and knowledge from data. Because our decision skills evolved in the physical world where decisions were immediate, tangible, and guided by our senses, we need techniques that help us in the modern world of data and concepts.

Machines can keep virtually unlimited things in memory at a time. Machines can process data across virtually unlimited dimensions. And these calculations are what form the everyday data we use to make decisions all day. The product recommendation from an e-commerce site or the chatbot response on a help site are driven by data so complex we can’t understand it. The data is, in fact, only readable by the machine and we can’t ask the machine to explain its reasoning like we can ask a friend or colleague.

You may have already studied problem-solving and decision-making processes. But the problem with problems is that they never stick to the script! A decision that started out relatively straightforward often becomes more tangled and opaque the more you learn about it.

There is no easy checklist for making decisions. It takes practice to be able to recognize and sidestep biases and human errors of judgment. And it takes practice to build the skills to find truth in data.

‍