Why Predicting the Future is So Hard and a Hack that Could Help.

Even people who have worked as futurists and forecasters for years if not decades find it super difficult to imagine future events that are not continuations of current trend lines. This dynamic was illustrated again this month when Politico published an article “The Incredible, World-Altering ‘Black Swan’ Events That Could Upend Life in 2025.”

Politico asked a broad group of prognosticators to identify some of the black swan events that could puncture 2025. Now a black swan event, as conceptualized by Nassim Taleb, is inherently unpredictable. If you can frame out a low probability, high impact event, then it is by definition not a black swan. People also speak about grey rhinos —events that we know are going to happen at some point, such as a catastrophic earthquake in the Pacific Northwest, but that we ignore and/or don’t adequately plan for, primarily because short-term thinking can’t justify the long-term expense.

At best some of the events imagined in the Politico article are grey rhinos. For example, I would put the largest cyberattack in history, a new pandemic, and a market crash in that category. Almost all the predictions basically extend current trend lines out to their worst-case scenarios, although a couple of the projections actually imagine an unanticipated, favorable turn of events. Hooray! A persistent bias among forecasters, and science fiction writers, is to imagine an ever-bleaker future for humanity.

It's not that efforts to think with some specificity and imagination about the future aren’t useful and important. They are quite productive and we—the intelligence community, policymakers, the media, YouTubers, academia, everyone—don’t do it enough. A proof point is the similar article Politico published in January 2024. There are some stunning predictions that rhymed with or resembled actual events of the past year: death at a Trump rally, a violent attack on a candidate, and a hurricane meets Jan. 6. We can anticipate the future; we just can’t anticipate black swans.

Is despair then our only option? I would argue not quite. There are some techniques we could use—and I’ve played with—that can help us shine a weak flashlight into the dark corners from which black swans emerge.

The future is not entirely linear and it’s also not completely random. It is a maddening combination of both, peppered with a bunch of other stuff that we don’t often consider when thinking strategically about the future, such as human emotions, pop culture, family dynamics, religions, so forth and so on. Toward the end of my career in government, I devised a futures game that I would play with analysts. I created decks of hundreds and hundreds of cards covering every category I could think of that could affect the future. Many of the categories are obvious: countries, economics, technologies, military developments. But I tried hard to capture the full range of human experience, so you could draw the Death card, or the Humor card, or the Fashion card, or the Facebook card, or the NFL card, or the Club Culture card, ad infinitum.

The cards were organized into about a dozen categories and, depending upon how much time we had and how complex we wanted to make the activity, teams of a few people would draw one card each from some or all the categories. They were given a short amount of time, never more than an hour, to come up with scenarios about the future that incorporated the cards they drew. (We found the exercise worked best when teams were asked to work quickly. Don’t overthink it. It’s also good to diversify from only words. Drawings are nice. Building with clay, legos can lead to insights.)

I’ll never forget the scenario that a small team developed more than ten years ago based on the three cards they had drawn: 1. China; 2. UAV’s; 3. Religion They came up with the following: At some point in the future, China develops a constellation of satellites and UAV’s that allows it to monitor in real time events all over the world to an alarming amount of specificity. The capability is so impressive that the Chinese system is able to make predictions about the future, in a sense replicating omniscience. Billions of people across the world develop a type of religious faith in the Chinese system.

I found the scenario feasible and alarming. I think about it today in terms of how people may eventually react to ubiquitous and more nuanced Artificial Intelligence. I also think it’s useful to feed AI engines random prompts and ask them to generate future scenarios. I remember asking ChatGPT to generate a scenario involving Elon Musk, sub-Saharan Africa, and dancing. Here’s one of the scenarios it produced:

In a bold move to combine his space exploration goals with cultural diplomacy, Elon Musk invites dancers from across sub-Saharan Africa to participate in a dance-off competition themed around interplanetary colonization. The top performers win a spot on the "Mars Cultural Mission," a project to bring art and culture to the first Martian colony.

The grand finale, held in Johannesburg, features Elon Musk attempting to freestyle alongside performers. His clumsy but spirited effort sparks memes and trends worldwide. The initiative boosts the African arts scene and Musk’s public image, showcasing a blend of technology and humanity as key to space exploration.

Anyway, that’s the method I came up with to distance people from their current concerns so they can think more randomly and creatively about possible future events. This technique also provides an opportunity for intuition to contribute to future forecasting. Considering and connecting random, seemingly unrelated variables is an intuition sweet spot.

I think others have come up with similar approaches. But I don’t know of anyone who has engaged in this activity methodically and repetitively. Doing it once in a while can be interesting, but not particularly productive. Keeping track of the scenarios and analyzing them for patterns, similarities and differences could yield interesting results. Revisiting them at intervals to see which now seem more or less likely might also be useful.

Are there black swans that we can now recognize in hindsight? I’ve heard 911 described as a black swan, but I’m not persuaded; we knew the US was the number one enemy of Al Qaeda; we just failed to imagine fully how the US could be attacked for maximum impact. One black swan candidate that does come to my mind are the 2nd and 3rd and 15th order impacts of the internet. By the mid-1990s I was convinced the internet would change organizations and media and commerce. But I at least did not imagine how the internet could undermine the legitimacy of Western democracies. But it did.

Any guesses as to what will happen next?