Unpredictable Patterns #74: Endings
Bangs, whimpers, tech-lash, the hype curve, the chasm and the value of endgame studies in strategy
Dear reader,
Summer is officially here. I am spending time on the island and sorting through a few things. I captured this photo of a steamboat at our landing the other day, and it is a solid indication that summer is indeed here.
My work has made me think about the ways things end - and why endings are predictive of the overall logic of a system or a set of events. So this week’s notes are about endings! And hence, naturally, about beginnings.
How it ends
How things end matters. The slow decay of a house into a ruin, or the catastrophic collapse of a building into rubble, result in roughly the same outcome - but the process itself can have radically different impact on the person in the house.
Studying endings is important, because it also allows us to understand when we are in new circumstances. Some ends are creeping and slow, others are quick and decisive - but they can be equally hard to detect - and paying attention to how things end help us understand when it is time to start anew.
Let's take a very timely example: the pandemic. Has it ended? Is it over? These, it turns out, are very different questions. It has ended in the sense that many countries have now lifted their restrictions and the world is slowly returning to some kind of pre-pandemic state - but it is far from over if we look at the prevalence and spread of the virus that caused the pandemic.
The pandemic required a political end before it was epidemiologically over.
It is easy to think that this is just due to political expediency, but there is more to learn from the end of the pandemic than this. Things that end - in some sense - before they are over are dependent on a balance of considerations, and at some point the core cause of the phenomenon becomes less important than the effects of the response to the event.
This is a nested game - one in which we are playing against the virus, but at the same time also engaging in political games nested in economic games and in the overall game of human attention. A crisis - like the pandemic - requires sustained attention to continue, and at some point our attention falters, and distraction erodes the foundation of the crisis itself.
And that is a central insight, I think: a crisis ends in distraction.
This is not just an observation, but equally a recipe for crisis response. If you can distract from the crisis at hand, the crisis will end - even if the core reasons or causes for the crisis subsists.
And it is also a clue to why crisis sometimes persist longer than you would expect them to: it is because they are used to focus a wandering public attention that could otherwise turn to other things that those trying to exercise power would rather we do not pay attention to.
We could go so far as to say that it is only in the ending that we really understand a thing. The way it ends is the way it was - and the end is the logical conclusion of the dynamic at play.
This is well known to us - and we judge a book or a movie not so much by its cover as by its end. The beginning is important, yes, but a good end will remain with us long after we have forgotten the plot. I am sure you can come up with several really good endings - my own favorites include The Usual Suspects, The Sixth Sense and Fargo, with Thelma and Lousie as a close runner-up - and in books this is even more the case - endings can make books classics in a few sentences.
Good endings are not answers, but rather like questions that the movie or book really wants you to ask yourself. A good ending hands something over to you, asks you a question and resolves nothing. Endings that leave nothing are thoroughly unsatisfying and often excludes the reader or viewer, leaving them as observers, rather than participants in art.
Endings rarely resolve anything at all.
The end, here, is a distant cousin of death, and works as a sort of existential punctuation mark. It brackets what came before and asks you to understand it as a whole. How do we know that the hollow men in T.S. Eliots poem were hollow? Because of the way the world ends in the poem - not with a bang but with a whimper.
Images of endings
The logic of ends can be described in simple diagrams. Aristotle's poetics suggests that the shape of a dramatic arc is built through increasing conflict to a key moment in which a great realization is paired with a turning point, and then followed by the denouement. In a picture:
This is the foundational dramatic arc, and what we often expect to find when we look for stories in the world. But there are alternatives to this, and one of the most interesting is actually found in business literature: the Gartner hype-curve.
It is a relative of Aristotle's dramatic arc, but now so overused that it is sometimes hard to get people to take it seriously. The hype curve attempts to describe how a technology is introduced and valued by a market, and it captures the fact that we have a hard time valuing technologies correctly. We overestimate, as the saying goes, the short term impacts of technology and underestimate the long term impact.
Most people focus on the first parts of the hype curve: the peak of inflated expectations and the trough of disillusionment. These two elements describe how we first get excited by a technology and then disappointed that it did not provide a silver bullet for all of our challenges, and it is really a valuable reminder for us to think about as we see new technologies introduced.
The hype curve, when combined with business thinker Geoffrey Moore's idea of a "chasm" that every technology needs to cross between early adopters and mature users, is also a useful tool to think through which technologies that are still at risk for failure, and which technologies have entered the ecosystem as more or less constant fixtures.
The perhaps more interesting parts of the hype curve are the latter two parts: the slope of enlightenment and the plateau of productivity. The slope of enlightenment is the place where we really understand a technology's true potential - it is the learning phase of introducing a new technology, and it requires a lot of failure. We scale the slope of enlightenment with great sacrifice: this is where most companies die and where we find business ideas strewn on the mountainside. The slope of enlightenment could equally be called the "steel bath of the real" or something gruesome like that. This phase is where we meet the dis-benefits of a technology and negotiate the net value of the technology to society.
Critics of a technology are crucial to this part - because they force us to peel away both disappointment and overly inflated hopes, and they demand that we ask the ultimate question: is the net impact of this technology beneficial or not?
For the first wave of internet technologies - consumer tech like social media and services like search - the slope of enlightenment coincides with the so-called "tech-lash".
This is important: it is easy to think that the tech-lash implies that we are stuck in the trough of disillusionment, but are very much not - we are scaling the slope of enlightenment, and if anything it would probably be beneficial to accelerate the ascent by forcing the criticism to the core: is there a net benefit here or not?
I am in no doubt about the answer: it seems obvious to me that the net benefit of an advertising-based consumer internet is far greater than its eventual cost - but let's hash that debate out in the open. And let's do it by thinking through what happens if we remove the technologies we criticize, and then look at what would replace them. And, yes, the answer is not going to be binary - it is not going to be banning social media or allowing it, but thinking through what a stable equilibrium state looks like and how we can achieve such a state through a reasonable allocation of responsibility across industry, regulators and, yes, users.
And that brings us to the fourth stage - the plateau of productivity. This is how the tech-lash ends - in a realization that there is a stable, productive state to be achieved here. What this tells us is that the tech-lash will not end with a bang, but with a whimper - with a slow transition into a stable state, and I think we are at the beginning of that end already.
In fact, when we write the history of the tech-lash, it is likely that the period we describe will be the decade between 2012 and 2022, and after that we enter a new phase - the plateau of compliance.
We can, then, use the hype curve to map out the regulatory stances along the development of a technology.
The peak of inflated expectations becomes the cyber libertarian phase, the trough of disillusionment the prohibition phase, the slope of enlightenment is equal to evidence-based regulation (and tech-lash), and the plateau of productivity is essentially the rise of a new regulated industry with clear compliance requirements.
This is what we have seen in many other industries over time - exuberant belief in technology as a liberating force, disappointment and then slow and measured look at the evidence, in order to build a stable market. The pattern repeats itself across new technologies.
The perhaps most obvious example just now is how crypto, web3 and blockchain are falling into the trough of prohibition. Proponents and advocates will need to think carefully about how to best scale the slope of evidence based regulation next, and how to maneuver between criticisms.
Scaling this slope of regulation is not easy - the wealth of regulatory initiatives can be dazzling and it is easy to become overwhelmed. But navigating this phase is key to figuring out how to move into the next stage - the plateau of compliance.
The plateau of compliance then translates to compliance costs that slowly increase over the years - with the corresponding barriers to entry, until a new use of technology unsettles the equilibrium again.
Not with a bang, but with slowly rising compliance costs.
Endings are beginnings
An end also symbolises something else - it is not just the conclusion of something, but it is the point at which we can no longer rely on our experience in a clear-cut way. To be at your wits' end is to have reached a point where you are facing something truly new. This is why a popular figure of thought is to declare the end of something and then look at what comes after. It is a useful, if somewhat overused, method - but there is value in charting what currently is ending.
Back in 1999 I was working as an analyst for the Swedish Office of Science and Technology, and we were asked to write a future piece about privacy, and one of the things we observed was that there was a lot of talk of the end of privacy. What we observing was a broad trend - and we were in the midst of it - that has accelerated since 1999.
Declaring the end of privacy is now a fairly regular starting move for both scholars and pundits, but few have asked the question of how privacy really ends and what comes next. Is it with a bang or a whimper? Is it slow or fast? And is the post-privacy world one in which we simply add identities and so hide in the noise? The end of privacy may also be the end of the idea of a single identity that can be revealed - and the rise of a pseudonym model.
Other ends that are interesting to track include the final finality: the end of the world. Here we see a somewhat worrying pattern in the n-gram evidence:
If anything our fascination with the apocalypse has increased in the last couple of decades, then. And we see this in both fiction and non-fiction writings, in music and art. We are increasingly curious about what it would mean for the world to end, and how it could happen: asteroids, nuclear war, cAItastrophes and other horrors are put on display as we try to understand the existential risks we face.
Is this a good sign? I believe it is - to some degree. A focus on the fact that the world could end, and a focus on the question of how we can collectively avoid that, may allow us to take charge of our own fate.
Living with the end in mind as a civilization can foster a collective memento mori that will allow us to focus on the kind of inventiveness and curiosity that will allow us to at least delay the end, and perhaps craft a new beginning.
Imagining endings is imagining new beginnings - and acknowledging an impending end is a good way to break free from the sense of fatalism that complexity can infect us with. Understanding the end is understanding the conditions for a new start, and new strategies, drawing from what we have learned, become more clearly visible.
Then, of course, there are things that we should end. Designing endings for things that are just consuming energy and attention is hard, but necessary - and often we don't even ask the question of how we should put an end to something that clearly is not getting us the outcome we need.
A first step can be to just ask the question: how does this end, and then explore if, indeed, it should.
That question - how does this end? - also is helpful in looking at complex processes and systems. We may not be able to predict how the system evolves - but by fast forwarding to how we believe it ends we may unlock some of the logic behind the system.
How does civilization end? How does the geopolitical game end? The answers are not always cataclysmic, and often it pays to think more about whimpers than bangs. That is not to say that things are not scary: whimpers are often much more scary scenarios than bangs!
A world that ends in a whimper may never even see what did not hit it.
Endgames
Bobby Fischer - a brilliant chess player and complicated person with some vile views - wrote an introduction to chess that started exclusively from endgames. He then expanded those backward, toward the more complex middle games, looking at how endgames could be used to understand the game backwards. His insight and genius was to realize that close analysis of the endgames actually encapsulated much of the basic principles of the game over all.
His method was a special case of an approach that sometimes is called ”starting from the end”. In this mental model we start from the outcome we want and then back our way to the present state of the system, looking closely at the moves and changes needed to achieve our goals.
This is a great model, but it is often simplified to such an extent that it no longer works: if we just start from where we want to end up and then find a single path to that outcome we have not done even half of what this method requires of us. The trick with starting from the end is that you need to spend time understanding the different ways you can get to your desired outcome.
Starting from the end means understanding all the different paths that lead there, and then figuring out the commonalities between the different endgames you have studied. Just like Fischer could teach chess from end games, you can learn strategy from endgames - but not from a single endgame alone.
The great advantage of endgames is simplicity. The way a chess game ends is usually characterized by the use of fewer pieces, and more decisive positions. Solving the game is easier from the end. The beginning - the opening - of a game is infinitely more complex because of the multiple pieces still in play.
Here we find another reason to study endings: they accentuate the strategic strengths of individual pieces. If you are in an end game with two knights against a king, you will need to understand the strategic logic of the knight closely before you play. Is there a path to check mate? What is required given the dynamic at play? (No, but you can solve for stalemate).
Endings allow you to look closely at the strategic strengths and weaknesses of specific plays, and pieces. The end articulates the particularity of individual moves and players.
This is true for general strategy as well. Looking at, say, legislative processes and political games we would do well to study not just the outcomes, but the last few moves in a game. We will find surprising regularities: the logic becomes clearer at the end.
In chess, players often study compilations of endgames. It is equally interesting for strategic thinkers in other pursuits - like politics - to compile their own collections of endgames. How was this or that law passed? How did someone survive a scandal or succumb to it? How did a political party lose power? What about individual political careers?
The study of endgames, across domains, reveals the deeper logic at play in life.
So what?
The study of how things end is often obscured by the impact of the end as such.
That something ends makes us think that the matter is closed - and we turn to the next new thing and forget the value to be found in how something ends. Understanding how companies end, issues in politics dissolve and how individual careers end provides a deep source of insight not just into what to do to avoid those same ends, but also to understand the overall logic of the games or conflicts we participate in.
So what can you do? Here are a few things that I recommend.
Post mortems are great - but focus on the end and then imagine the alternative endings. How could this have ended and what drove the logic of the end?
Distinguish between whimpers and bangs - what kinds of end do you predict for the issue you are involve in? Where do you expect a bang and why? Where a whimper? The fast and dramatic ending is very different from the slow and silent!
Endings are sometimes very slow and gradual - and recognizing them is hard. The end of the Roman Empire is just one example of an ending that echoed for a long time - another is the end of religion in Western societies. These slow endings are formative for societies, organizations and individuals, but often not recognized as endings. Which slow endings are you caught up in and why do they matter?
Endings are beginnings. Where do you think an ending has matured into a new beginning? Think about the end of privacy - it is hardly a settled issue, but it is opening up new beginnings!
Think about how things end. How will the Internet end? What about capitalism? China? The US? Mapping endings may seem a morbid thing - but it is surprisingly interesting!
As always, thank you for reading.
Nicklas