What this has given me is food for thought on how to think about the future. One approach is to project forward from today, including thinking about the possibility of disruptive events. The other is to look at the future you want to create and reverse engineer your way back to what needs to happen to get to that future. These approaches are probably unfairly opposed. I associate them with Eric and Hildy respectively. Not because Eric or Hildy is a strong exemplar of either (though they may be), but simply because I know of the approaches through them.
Today I want to think about the approach represented by Eric. Actually, I want to set the groundwork for some further thinking to be done in a subsequent post.
Eric's book Future, Inc sets out a methodology for thinking about the future. It is systematic, first starting with:
The first step is to understand the system, its key components. Knowing the system, you can then look at trends (qualitative) that might affect system components, and attempt to put numbers around them to form forecasts (quantitative). This is important stuff, and seeks to narrow questions (forecasts) down to a sufficiently narrow base such that they can be answered.
Having the data, it then needs to be explored for implications. Two approaches are suggested:
- A "futures wheel", where a single change can be fanned out into primary, secondary and tertiary implications. The example from Eric's book is below:
|Futures wheel, page 93 of Future, Inc by Eric Garland.|
- A "cross-impact analysis". This is a pretty straightfoward analysis that enables separate trends to be systematically analysed to see how they might interact upon combination.
So all of the above gathers information that is likely to be interesting but boring. And thus useless. Sure, it will capture some compelling factors to consider, but it will be abstract. This is overcome by developing scenarios.
Scenarios are stories we tell about a point of time in the future, preferably at least 5-10 years in the future to escape the orbit of here and now. The stories are not to be likely (as Herman Kahn put it, "the most likely future isn't"), but instead are to be plausible enough to think through what they mean. The stories are also close to home, even prosaic. So it might be "how does a typical family have breakfast in 2020", or "what is the headline of your local paper's 'Year in Review' edition for 2020". By using a story and making it local, you cut through the fog of abstraction and come down to something that people can relate to.
Eric suggests more than one, and preferably four scenarios, for presentation and discussion. Having only two leads to people polarising to one or the other, three permits people to select for a "middle ground" between two apparent extremes, and more than four starts to become too confusing. Four is just right.
The interesting bit comes from the scenarios. They are used (and here I shift in focus from Eric Garland to Nicholas Nassim Taleb) to understand where you are vulnerable to black swans, or low probability, high impact events. You work out how to position yourself, your business, such that these events do not kill you, and ideally you are strengthened through them. That is NOT the same is predicting the events. That would be impossible.
Applying this to the future of waste
For today I won't track right through this process. Waste is complex. Or at least, it seems complex to me having spent a fair bit of time within the industry. So to start, I've done a dump of components of the waste system into Freemind. It's copied below as a picture, and you can get the Freemind version here. I welcome you to look and play.
|Systems map for waste. Find the Freemind version here|
From here, I want to explore this system through to understanding trends, forecasts and implications such that I can prepare four (4) scenarios. And then explore how to become antifragile. Of course, my analysis won't be right. It will need significant improvement. But the point is to do the exercise and see what comes of it. To learn through trying.