Don’t Predict. Imagine.

“It’s tough to make predictions. Especially about the future.”

It was true when the great Yogi Berra said it*. It’s true today. Even with all of our advanced data collection and analytics technologies, human beings still struggle to predict everything from the direction of the economy, to election outcomes, to whether the New York Giants will return to Super Bowl glory. 

So how is it that exactly one year ago –– as inflation was peaking, the stock market was about to start plummeting, the war in Ukraine was grinding, and the Dobbs decision had poured gasoline on America’s already deeply polarized politics –– we were able to write a detailed description of how over the coming year the economy could steer into a “soft landing” and the political system could deliver significant bipartisan accomplishments?  

And these descriptions weren’t generic. We specifically talked about how technology could unlock a productivity and investment surge, driven in large part by “mass-market AI tools.” We anticipated the possibility of bipartisan deals “to further strengthen the nation’s core with a focus on investments in R&D, critical infrastructure, national defense, and education.” [If you don’t believe us, you can read it all here]

Remember, this was all written well before the launch of ChatGPT, the taming of inflation or the passage of the CHIPS Act, infrastructure bill, and Inflation Reduction Act.

Are we some sort of super forecasters? Not at all. 

The context in which we described these possibilities was no prediction. It was one of four scenarios we laid out for how the US political-economy could evolve over 2022-24. 

Much of what we described in that particular scenario, which we termed “The Roaring 20s After All,” has, in fact, come to pass. It really looks increasingly probable that the U.S. will check inflation without triggering a recession, and it is also clear that a significant driver of that result is the advent of a slew of productivity and investment enhancing technologies, including AI, machine learning, robotics, spatial computing, and more. It is also clear that while our politics are as ugly and as personal as ever, Washington is finding ways to pass significant legislation on a bipartisan basis, while avoiding the worst outcomes like a default on U.S. debt. 

Of course, the fact that one of our four scenarios has turned out to be an effective description of what has transpired means that the other three scenarios were largely off-base. 

So what was the value of an exercise in which one scenario yields accurate insights about the future and the others don’t?  

Three things:

  • First, scenarios help us resist the natural tendency to assume that the future will be like the present only more so. Back in August 2022 all the arrows were pointing in the most depressing directions. Inflation was surging. The war in Ukraine was entering dangerous new territory. Leading publications were openly speculating about a new American civil war. It was easy to extrapolate those trendlines and confidently predict that everything going badly would just get worse. Scenarios force us to fight that urge and imagine futures in which whatever the current dynamics happen to be, they change direction. 
  • Second, scenarios give us the chance to “live” in plausible and distinctly different futures and to think deeply about how those futures might come to be. For our Roaring 20s scenario, we started with a framework of a soft-landing and a less polarized politics. That then required us to investigate how such a world could come to pass, leading to the insight that it probably would require things like advances in general purpose technologies and a shift in politics from public grandstanding to backroom dealmaking. By forcing yourself to mentally inhabit a future, even if you don’t think that future is likely, you find yourself discovering interesting dynamics that may change your sense of probabilities and possibilities.
  • Finally, scenarios help us sort the signals from the noise. Having written these scenarios, we became far more sensitive to information about new technologies that might be significant productivity enhancers, while focusing our attention less on the public noise in our politics and more on the back-channel deals that were being cut. Our scenarios had indicated that these might be strong indicators of an overall direction, so we paid a lot of attention to those signposts, which made us more attuned to the probability of a soft-landing + less polarization sooner than we would have been otherwise. This heightened sensitivity allowed us to prepare ourselves and our clients for outcomes that might otherwise have come as a shock.

As we now peer further into the future, we will continue to fight the desire to make predictions we know are likely to be wrong. Instead, we will construct scenarios that push us to imagine possibilities that may feel remote or counterintuitive, because doing so is the best, perhaps only, way to position oneself to navigate the enormous uncertainty of the world.

We hope Yogi would approve.