Unpredictable Patterns #78: Time machines

Clocks, symbolic time, organizational rhythms, energy and complexity

Dear reader!

Thank you for the comments on the last piece on capabilities - and the thoughtful questions. I would like, in particular, to highlight ForBetterOrgs.coms work here, and this link. I found it very interesting reading. In this note we will think about time machines and how they work. I hope you enjoy it and that the summer is coming along nicely for everyone!

Time Machines

In science fiction the time machine is a wonderful tool - it allows you to travel back and forth in time to explore history (with a mighty set of paradoxes to choose from, such as what happens if you murder your ancestors?). The image of time is like a parallel landscape we can traverse and get lost in. But that is not the only way to think about time machines. We can also think about machines producing time.

This line of thinking starts with recognizing the importance of the clock - measuring time and allowing us to divide our day up in different ways, but also creating an accountability for how we use time time allotted to us. A clock does not only measure time - it creates a certain kind of time that has uses and costs associated with it. We can think of it as the different between cloth and the clothes we make of them. Cloth is a neutral material, but with clothes/clocks come cultural rules and expectations (do not run around naked/ do not waste time loitering about).

Clocks produce symbolic forms of time.

The Heideggerian analysis here is simple: technology makes time available as resource and so, in some sense, also diminishes it. But that is not quite right: the clock does make time available as something - but is it just as a raw material or resource? No, that is too simplistic - what technology does here is that it makes time into a symbolic form, a cultural, psychological building block.

Rather than succumb to the Heideggerian reductionism when we look at technology and clocks, we should listen to his great rival Ernst Cassirer. His concern with symbolic forms allows us to think differently about the clock and see it as something else - a symbol engine or a semiosis device.

Clocks generate symbols that can produce meaning and so grow into culture and community. That is why they have such a profound effect on history, and why they matter so much when we think about time.

A clock, then, is a kind of time machine that produces symbolic time. This is far from the innocent measuring of time, of course - and this is important. The kind of time machines that we will be interested in are all producing different kinds of time, and they do so at some kind of cost.

The cost of clocks.

So what is the cost of clocks? It is easy to succumb here to the simplistic story of the clock as regimenting our lives, forcing us to work and robbing us of our leisure - to simply say that the cost of clock is freedom. But that is clearly at best a simplification, and more likely a falsehood.

The cost of clocks is complexity.

This is true of all time machines. They create complexity, and that complexity then needs to managed in different ways - and so leads to the demand for even more ingenious time machines of different kinds. The reason behind this is simple: when we produce symbolic time we also create possible combinations of time that can be used and utilised in different ways - those new patterns of use now have both more individual components (moments as measured by clock) and greater connection between us.

This latter point is worth thinking about. Clocks produce symbolic time at a higher resolution than before - seconds, minutes, hours - but they also create another weird quality that is surprisingly powerful: synchronicity. Suddenly we can do things at the same time or in relation to someone else on a schedule.

Symbolic time can be shared time and so connects us in a great network of moments.

That, the fact that we do not only create distinct moments but that these are connected between all of us, creates a complex network of time at a higher resolution than we have ever seen before. We can think about the clock as a lens through which we suddenly see time in a new way.

The thing is this: a clock does not create absolute symbolic time - but it creates time for you. It unlocks granular moments of time for you, and as you realize others have clocks as well you realize that you exist in this vast network of moments, interconnected and interrelated with others who can do things with you, after you, before you. This network of time is exploding with a multitude of possible rhythmic patterns that interact - and so now we see it clearly: the clock produces immense complexity.

And note that we say complexity, and not chaos. We can still navigate this new time - but we need to adapt and think about how we do it, and so we change, our culture changes and time changes too.

Clocks are useful ways to think about organizations as well. One way to assess the competence of an organization is to look at the clocks it has built. The most obvious clock that a modern, multi-national organization will build - especially if it is manufacturing something - is its logistical supply chain. We talk about "chains" but what we should really talk about is supply clocks.

A good supply clock produces symbolic time for the company that it can use to plan, act and react to the environment it is in. If it is well constructed and part of a network of clocks it also creates predictability and trust.

(It is worth spending a second on that notion, by the way: time and trust are related. Sometimes trust is said to be born out of consistently good behavior, but that is only half a truth. Real trust is more related to understanding an actor’s clocks - the way they interact with us and the world, and their own symbolic production of time. If I know to what clock someone is keeping, I can also trust them.)

But individual teams also have their own clocks. Synchronising them is key to a greater organizational unity, and the careful design of team clocks is a difficult thing. It is not just about meetings, but about understanding how a team relates to the production of symbolic time overall. In what time chunks do we think? That matters not least because we want to be able to understand if we are making progress, and making progress is only possible in time.

How a team chunks time then also becomes the basis for how it can improve its own performance.

The use of clocks on dashboards and as an instrument is a way to make time into a tool.

One of the fathers of modern management - Frederick Taylor - understood this, but we know can see that he only got it partly right: the idea that standardizing the clocks and then using them to measure the work done focused not so much on the clocks as the effects of the clock. Taylor took time as a given, not as an element of design. In contrast, we know now that we can construct different clocks for different forms of work: the sprint, the long term project and the day to day need different clocks, and versatility in shifting between different clocks is key to an organizations ability to adapt.

At the limit we can think of designing an organization as the task of designing a complex, interlocking set of clocks that produce symbolic time - the basic language of work - and then we can think of everything we can formulate in that language as the overall capability of the organization as a whole.

This is also a clue to how we can understand a large category of failures: they are simply examples of organizational arrythmia and a collapse of symbolic time.

Technological progress and time

There are other kinds of time machines as well. In fact, it can be argued that we are time machines ourselves. We are taught to think that we act in time, and that we exist in time - but the reality is that the brain produces time and so in some sense is a clock itself.

Brains are already quite complex, so they can be seen as a response to the evolutionary complexity produced by earlier clocks and the connection in time. One theory about human brain evolution is that we evolved big brains in order to deal with social complexity - and that explanation would fit well with our overall model. Clocks are produced to deal with complexity, but produce new complexity in turn.

The clocks we produce represent our ability to harness and think about energy as well — think about the hand-powered early clocks.

Complexity and clocks are locked in a spiral that resembles a red queen problem: it takes all of the running you can do to remain in the exact same place.

But let's look more closely at how brains produce time. What is it that is happening here? And I do not mean at a neurological level (that is worth coming back to), but am more interested in a coarse-grained model. If we zoom out there is one interesting thing that is quite obvious: brains produce time and consume energy.

The human brain runs at 12 watts, or a fifth of a 60w light bulb. A large part of the energy it consumes is engaged in producing different kinds of biological time as well as symbolic time. The brain is a series of clocks running in a very energy efficient way.

This is intriguing, since it suggests that there is a fundamental connection here between clocks and energy, or time and energy, that we could explore.

One perspective is efficiency: how much energy does it take to produce the organizational clocks and keep them running? The more energy efficient an organization's production of time is, the better off they will be.

Another perspective allows us to think about technology and time more abstractly. Many technologies produce or shift time in ways that are useful to us. A car will take us faster to a destination than if we walk, but at the cost of more energy (and complexity). Cars, thus, are really time machines - and they allow us to navigate the network of moments in new ways.

The same holds for, say, the Internet - a vast time machine that allows us to tap in to a global network of interconnected moments in ways that are hard to understand if we take a moment to reflect on the wonders that have become commonplace for us.

All of this at the cost of energy, and complexity.

Now we can start to ask the question of what the exchange rate between time and energy is, and think about that as a key element in understanding technologies.

Assume we make a car that can take us from A to B in 20 minutes, and that it would take us 60 minutes to walk there. Now assume that this car consumes X joules and that your walking would consume Y joules -- well, then we can look at the exchange rate of time and energy here.

And what holds for transportation holds for other things too - including things like internet search. In one well-known study the participants were divided up in groups where one group used the internet and the other a library. The average time for offline searches - where they succeeded - was 19-22 minutes, but for online searches that was 7-9 minutes. And this was for queries where the offline library could find the answer - so the Internet clearly is a time-machine in the sense that it has produced a much more information-rich kind of time for us (“time savings” does not quite capture what is happening here, I feel).

Sundials as an energy efficient clock.

The follow-up question then becomes at what cost in energy and complexity this time is produced, and how energy effiicient the Internet is as a time machine. If we look at the sum total of tasks we can accomplish with the internet - and accompanying kinds of rich, symbolic time it produces - my guess is that the Internet is producing time at a very efficient rate and may in fact be one of the most energy efficient time machines that we have every invented as humanity. The complexity costs, however, are harder to assess - but this is worth coming back to.

A wealth of complexity

That time is at the heart of the modern information society is not surprising, and not a new insight either. In these notes we have often referred to Herbert Simon's paper “Designing organizations for an information-rich world”, and his seminal insight that with a wealth of information comes a poverty of attention and a need to allocate it efficiently - and Simon pointed out that it was probably through new technologies (such as AI) that would need to do this.

Simon also defines attention as approximately the time someone spends on something - and this resonates well with what we have said so far. We can then read Simon's insight as telling us that with a wealth of complexity comes a poverty of time and a need to produce it efficiently.

Magic produced many clocks with high complexity yields.

We now have put forward the hypothesis that technology indeed does allow us to produce different kinds of rich, symbolic time that we can use in novel ways - and that the efficiency of time production is a key factor in understanding and evaluating technology. That also gives us the opportunity to ask other kinds of questions that could be interesting to think about and where the answer is not immediately obvious to me:

1. Assume we build a system that perform certain tasks t(1) to t(n) and that it takes 10 years to train a human being to do this well and so that consumes X joules, but the machine takes Y joules - for which tasks is it then worth it to automate if we just look at energy consumption? This will of course depend on energy prices to some degree - but we can start to make educated guesses.

2. A machine beats a human when it comes to a certain task X. The human's performance is based on X hours personally invested and Y hours that others have historically invested in order to produce the knowledge the individual human used to start their own training. The sum total energy consumed is X+Y hours over the time span of human civilization to produce a certain competence at the task - at what energy exchange rates do we judge the machine's performance reasonable? How do we think about the energy efficiency between human generations and over a civilization in terms of producing symbolic time?

3. Are there technologies that are only possible under certain energy conditions? We can think about the famous idea of Kardashev-type civilizations - the first is one that is able to harness all the energy of the sun that reaches the planet, the second harnesses all the energy radiated by the star and the third all the energy produced at the scale of the galaxy - and then we can ask: which kinds of time could a Kardashev III civilization produce? Or do you need to be a Kardashev II to produce certain kinds of technology? Are the adverse effects to produce certain classes of technology in a Kardashev I civilization (too early in some sense)?

These are frivolous questions, but there is some value in thinking about them - not least because energy still is scarce. I do not find that this is an argument against technology at all (some may think that using energy to produce rich symbolic forms of time is irresponsible under the threat of climate change, I think it is the only way we deal with the complexity we have chosen to produce until now - not technological determinism as much as a pre-committed course of action dictated by the last centuries. We may change that at some point and choose to live differently, but in order to do so we need to create a gap between the kinds of time we can produce and the kinds of complexity that we deal with and in that gap make the choice).

So what?

So this was a lot of speculation, right? Absolutely. But the mental model of the time machine and the clock are immediately helpful in thinking about many organizational problems. We have spoken before about rhythm and time - and these are some of the most basic, and most difficult, mental models that we need to internalize if we want to learn efficiently. So what can we do?

• First, explore the clocks in your organization - where are they situated, how are they working and what kinds of complexity do they produce? Are they energy efficient? A good clock is key to good decision making as well!

• Second, list the kinds of symbolic time your organization produces and where it could be improved, connected or left behind. Budget time, supply time, OKR time, perf time, project time...who is responsible for designing your organizational time?

• Third, explore social clocks and time machines. Simple examples as political cycles of mandates or press cycles are obvious to most of us - but there are so many other clocks at play as well -- generational and seasonal (as mentioned here) - are there hidden clocks at play?

• Fourth, think about really long time clocks. The foundation of the long now is building 10 000 year clock - what are the implications of that and what kind of time would such a clock produce? Could you build a ten year or 100 year clock in your organization?

Time design is fascinating and hard - but it almost always provides a fresh perspective on things. I hope you have found this at least a bit interesting and as always let me know if you have any questions! Thank you for reading!

Nicklas

A note on distribution: I have closed down the newsletter to invite-only because my ambition was never to have a large distribution, but rather thought community - and I think we are at the right number for that. (As always it is also worth noting that everything written in here is my own views only, and not to be confused with anything else. If you do know someone who would like to be added, drop me a note.)