Infinite Content: Chapter 11
Scenario Analysis: Giving Uncertainty Some Shape
This is the draft eleventh chapter of my book, Infinite Content: AI, The Next Great Disruption of Media, and How to Navigate What’s Coming, due to be published by The MIT Press in 2026. The introductory chapter is available for free here. Subsequent draft chapters will be serialized for paid subscribers to The Mediator and can be found here.
What do you do when confronted with a lot of Rumsfeld’s known unknowns?
In the early 1970s, Shell (then called Royal Dutch Shell) faced the same problem as every other big oil company: an increasingly uncertain global energy environment.
Throughout the 1950s-late 1960s, the West enjoyed, and relied on, abundant cheap Middle Eastern oil. Western oil companies owned concessions in many Middle Eastern countries: long-term, exclusive contracts to control exploration and production, with relatively small royalties paid back to the host nation. OPEC (Organization of the Petroleum Exporting Countries) was uncoordinated and lacked much leverage. But by the late 1960s-early 1970s, nationalism was rising in many oil-producing countries, who were starting to nationalize oil production, and OPEC was gaining traction.
In The Art of the Long View, Peter Schwartz chronicles how Shell used scenario planning to break from conventional, mostly linear forecasting. The Shell team developed multiple plausible scenarios, each built around different combinations of political, economic, and social drivers. One of the key scenarios envisioned a sharp cut in oil supply driven by geopolitical events—a scenario that occurred in 1973, when Arab oil producers imposed an embargo in response to Western support for Israel during the Yom Kippur War.
Other oil companies were caught flat-footed by the crisis, but Shell was better prepared. It had already started planning around the possibility of constrained oil supply. It was able to make faster operational and investment decisions, reallocate resources, and adjust its global supply chain more nimbly than its competitors.
Maybe even more important, according to Schwarz, was that Shell’s executives started to think about uncertainty differently. They started to prepare for a range of possible outcomes.
Confronted with so many uncertainties about the evolution of GenAI—all of which will collectively determine just how disruptive it will be—let’s use the same discipline in this chapter.
We Don’t Know a Lot
To recap last chapter, there is a lot we don’t yet know about how GenAI will evolve:
How realistic will GenAI get?
How much will it reduce the need for human labor and, therefore, costs?
Will it enable enough creative control for the most demanding professionals?
Will consumers accept it—and for which uses?
Will unresolved legal issues slow progress and adoption?
What new forms of media will GenAI enable and will they matter?
As I wrote in the last chapter, that’s a lot of not knowin’. But we shouldn’t just throw up our hands. One of the most useful tools for framing potential outcomes in an uncertain environment is a scenario analysis matrix.
How Scenarios Work
The concept behind a scenario analysis is straightforward. It entails identifying the two most important variables (or, to stay Rumsfeldian, known unknowns); determining the polar extreme outcomes for these variables over a set timeframe (five years out, ten years out, whatever); and constructing a 2 x 2 matrix that produces four potential future-state scenarios.
To be clear, our goal in using scenarios is not to predict which of these futures will occur. We don’t and can’t know. Scenario planning is valuable not because it tells us which future is likely, but because it helps us reason across a wide range of outcomes. The most illuminating part is writing a narrative describing each scenario. Think of these narratives as news articles from alternate futures, explaining how we got to that (possible) future state.
The scenarios are intentionally extreme, so reality will probably fall somewhere between them. But the exercise helps define the bounds of what will probably unfold; the signposts that would indicate we are heading in one direction or another; and the potential implications of different outcomes. It also helps make abstract problems feel a bit more concrete, especially when the scenarios are specific. The more specific, the better.
