Our Coordination Paradox
Everyone agrees we must get better at coordination, so why aren't we doing anything about it?
What would you consider to be the major challenges facing humanity?
The list is surprisingly short. It’s surviving climate change. It’s safely ushering in artificial intelligence. It’s making sure that we can survive the next major pandemic, or other existential risks. It’s about factors that will ensure that each of earth’s inhabitants can flourish, like political stability, inequality, and education.
These are certainly all major challenges, and worthy of our planetary attention.
Now ask yourself, what would it take to solve these challenges?
The problems they represent may seem insurmountable, given their inclusion in our list. But that’s not actually the case. Every challenge has viable approaches and possible solutions. And there’s often strong consensus on which of these would work best, or at least would be worth trying.
And yet there’s one thing that each of these solutions share: a fatalistic pessimism that they would not stand a chance of actually working.
The reason why isn’t because these solutions are blocked by technology or resources. In fact, there is often sufficient technology and resources available to help solve the challenge.
The reason these solutions are blocked is because of us—humans. More specifically, the problem is our capacity to coordinate as a human collective. Or more precisely, the problem is our incapacity to coordinate over almost anything of importance at all.
Our capacity to coordinate is defined by many things, but it often boils down to how skilled we are as a species at being selfish. We’re very good at caring for ourselves, yet we’re very bad at extending that care to anyone beyond our immediate circles of interest. This is often attributed to “tribalism”, due to our evolutionary ancestors spending the majority of their time in small tribes. Our survival depended on making clear distinctions between “us” and “them”. Our instincts evolved to distrust outsiders, to discount their claims, and to assume the worst about their intentions.
Fast forward a few dozen millennia, and these are the biases that make coordinating beyond the tribe so difficult today. They create obstacles so monumental that they may as well be a brick wall, one that halts the progress of every proposed solution that slams into it. Every attempt to get through, over, or around the brick wall seems to end in the same result: failure.
Not convinced? Ask an expert tackling one of the above challenges to describe their most pressing obstacles. They might begin with challenges specific to their own solution, but soon they will converge on the brick wall of coordination they keep smashing into. And they’ll have the battle scars to prove it. They’ve tried everything they can think of yet nothing has worked.
Keep listening and the stories will all start to sound the same.
Sometimes it’s a promising inquiry that simply concedes to the reality of the brick wall: “We could never even agree on a human reward function, much less one that could guide AI alignment efforts.”
At other times an entire technology is stifled by coordination failures: “Of course nuclear power isn’t cheap enough. We let irrational regulation prevent an entire generation of innovation from happening.”
The worst is when experts bash their heads straight into the wall so technology can’t even get a seat at the table: “Forget any sane geo-engineering experiments. The senior decision-makers at this year’s COP couldn’t even agree on what to order for lunch.”
Often it's doing whatever is necessary to avoid the brick wall altogether: “We can’t depend on sacrifice. Our only chance to avert climate disaster is to make green energy cheap enough to be in everyone’s selfish interest to use it.”
The most tragic scenario is when the resources exist yet still are not enough: “It’s not even about technology or resources. We could end world hunger today if we could just agree on using what we already have.”
This brick wall should make one thing perfectly clear: we’re going to have to get good at coordination if we want any chance at solving our greatest challenges.
Everyone agrees that improving our coordination is necessary. In fact, it seems so obvious that it has become a truism, where we no longer seem to appreciate how important this point is.
Perhaps rephrasing it will help us look at it with fresh eyes. What we’re saying is that for every major challenge facing humanity, every viable solution is downstream of our ability to coordinate.
Take a moment to let that sink in. It means that improving our capacity to coordinate would go further than any other human endeavor to help address these challenges.
Wouldn’t this suggest that improving this capacity is the most important thing we should be doing? If nothing is more important than improving our capacity to coordinate, then why aren’t we radically reorienting our societies to do so? Why is no one working on this?
One look at the state of our societies would indicate that improving coordination is very far down on the priority list. That may not seem the case, but it’s easy to clarify this point by imagining a different society, one that did prioritize improving coordination.
What would such a society look like? It would have a field of “coordination studies” on par with economics and psychology. Academics could pursue doctorates in coordination. Ambitious youth interested in increasing our capacity to coordinate would be incentivized by obvious career paths and financial rewards. Developing collective human intelligence would receive as much investment as developing artificial intelligence. Awards and prizes would incentivize breakthroughs in coordination. It would raise up role models and historical examples through education, myths, and holidays celebrating key moments in human coordination history. It’d have super heroes of coordination.
Of course our societies have none of these things. Instead, anyone with any ambition to improve coordination faces an uphill battle. They would be fighting a culture that seems intent on believing that improving coordination isn’t a viable pursuit.
You could imagine some genius—some Einstein of coordination—emerging only to wallow in obscurity because no one would take his ideas seriously. Even if he could prove the effectiveness of some brilliant coordination idea with mathematical certainty, who would listen? How would it get resourced? This Einstein would more likely get mocked and ridiculed than taken seriously.
And that’s the coordination paradox. Even when we can agree that solving the greatest challenges we face will depend on improving our capacity to coordinate, we act as if this capacity is not something we can or should improve at all.
It turns out that we can’t even coordinate on the need to improve our ability to coordinate.
But why is this? Why have we become so fatalistic about our capacity to coordinate?
There is an underlying assumption that everyone seems to be making: that whenever we are talking about coordination, we are talking about something inherent in human nature, something forged from millions of years of evolution to become a genetic inevitability.
This means that whatever our current capacity for coordination is, that capacity is something we are just going to have to deal with, because to change that capacity would require changing human nature itself, and changing human nature is obviously not possible.
Based on this assumption, suggesting that we should try to improve our coordination is viewed as preposterous. It would be like suggesting that we should evolve wings to reduce the impact from our air travel on climate change. It just doesn’t work that way. Our own human nature is the one thing we cannot overcome.
This assumption is not entirely incorrect—it is certainly based on plenty of real evidence. You can read about Dunbar’s number, the upper limit of how many people we can keep track of and coordinate with. You can read about all sorts of red/blue psychology tests that show how we’ll use almost any difference to divide ourselves into “us” versus “them”. You can read about culturally robust personality traits that manifest as persistent political differences. You can read about selfishness and free-riders and even “Moloch”, the monster that eats coordination for breakfast.
The anecdotal evidence may be even more persuasive: who doesn’t have their own story of hilarious failure that resulted from trying to coordinate other humans? If this does not describe you, then you clearly have never tried to get more than a few humans to agree on anything of significance. I recommend participating in a local parent-teacher fundraiser, a planning committee of any sort, or even a fantasy football draft to catch up.
And of course the historical evidence of humanity trying to solve our own planetary challenges is significant and mounting. It’s as if each challenge is trying to outdo each other in creating the most persuasive case study on how not to coordinate. Our response to Covid may have taken the current lead, but likely not for long.
But this assumption is not entirely correct either. There is a big difference between being a bias and being an indelible part of our nature. Human nature is in fact not immutable. We’re not only uniquely good at coordination, but we have a long history of getting better at it. And this history includes leveraging technologies to further augment our capacities to coordinate.
It’s worth digging into these arguments so we can understand how inaccurate this assumption is.
By any measure, humans are very malleable. Our nature is defined both by our genetic make-up and the cultures we create for ourselves. Genetically speaking, humans are still evolving. In fact, our DNA has changed more in the past thousand years than it did in the previous 100,000. And we’re only beginning to understand how epigenetics and other factors impact evolutionary dynamics more than we thought. Evolution is not done with us.
Meanwhile, the pace of cultural change is accelerating much faster than genetic evolution can keep up with. This is dramatically altering how we think of ourselves and each other. This can be seen everywhere: in the changes in our identities, in our gender relations, and even in our desire to have children. These are fundamental aspects of the human condition that are dramatically changing at speeds so fast we can observe them in decades, not millenia.
Technology has amplified the scope and pace of these cultural changes—in a sense the Internet, the mobile phone, and social media are all grand experiments on human nature that we’re only just starting to get the test results back from.
So human nature is not set in stone; it’s more like clay, capable of being reshaped by the hands of genetics and culture that are constantly learning new techniques. Our nature has changed tremendously just in the last century, and it will change even faster in the next. There’s no reason to think that our capacity to coordinate will not be included with these changes.
The second reason that improvement isn’t hopeless is because coordination is what defines the very success of our species.
One way to describe human evolution is by telling the story of our increasing coordination. Our populations exploded when we began orchestrating group hunts to take down much larger prey. Our intelligence co-evolved with the demands of our complexifying social circles. We increased the complexity of our social organizations to take advantage of greater and greater coordination. The scale of our coordination expanded from tribes to kingdoms to empires, all the way to the global markets and supply chains of today. It took the coordination of entire civilizations to create humanity’s greatest achievements, from the Great Pyramid to the Great Wall to the great Apollo Project.
The final reason that improving our coordination isn’t hopeless is because of technology. We use technology to radically augment virtually every human skill we possess, including our strength, our speed, our memory, our cognition, and our endurance. Technology is just as capable of increasing our capacity to coordinate.
It is the nature of technology that the same innovations that have led to some of our worst coordination failures have also enabled our greatest coordination breakthroughs. It was the invention of writing that enabled the recording of laws, transactions, and historical events that made it possible for complex societies to coordinate. It was the invention of the printing press that led to the democratization of knowledge and the explosion in collective intelligence. It was the invention of the Internet that unlocked coordination beyond the boundaries of geographic borders and social barriers.
Today we are seeing more intentional efforts to tackle specific coordination failures. Emerging technologies like blockchain are exploring decentralized and transparent systems for governance and financial transactions. Machine learning is enabling software to understand the values of large groups of people in ways that can break through political polarization. AI algorithms are capable of making complex datasets more legible for collective decision-making. These are the technologies that will help unlock new scales of coordination.
One reason why we seem to be perennially bad at coordination is because the scale of our problems have always increased with the scale of our capacity. The planetary scale of our current challenges, in which every living creature on earth has a stake in the outcome, is an entirely new caliber of scale. Of course we’re going to be bad at it! We’ve had less than 100 years to practice and improve. Why would we all of a sudden be skilled planetary collaborators without the necessary training, education, and technological augmentation? A mere century is not enough time for trial and error to work its magic.
None of this is to suggest that our biases against coordination aren’t real and significant. But these arguments should convince us that improving our capacity to coordinate is not an absurd fantasy. It’s time that we start operating on different assumptions.
What happens when we no longer make these fatal assumptions?
What happens when we agree that our capacity to coordinate is not static, but that it can improve or degrade like any other skill we possess? And what if we can agree that these improvements are not only possible but worth devoting resources to?
Changing these assumptions changes everything. It transforms the coordination paradox into a coordination revelation.
What is revealed is that our capacity to coordinate represents a giant untapped solution space, one that has been largely ignored and neglected. Once freed from the assumptions of genetic fatalism it becomes ripe for exploration. It suggests that any serious investments in improving our capacity will have great potential for outsized returns. It means that every breakthrough in psychology, technology, and governance has corresponding opportunities for breakthroughs in coordination.
The first reaction to this revelation should be sheer optimism. It’s as if the world’s biggest secret has now been made known: We have this latent collective superpower that can help us solve everything, and it’s a skill that is just sitting there, waiting to be studied and developed and improved like any other.
And why should we doubt this? What in our history would lead us to believe that we have exhausted every attempt to improve our capacity to coordinate? How many of humanity’s greatest minds have seriously endeavored to improve coordination? How many of our best ideas for improvement have been sufficiently resourced? How many experiments have we tried?
The problem now is to overcome all of the friction that has encrusted onto our societies based on centuries of fatalistic assumptions. There are countless talented people everywhere who do think improvement is possible, and who are setting out to prove it with a range of ingenious ideas. What they need are catalysts that can help jumpstart their ideas, easier pathways that can put those ideas into practice, and movements that can unite them all together.
Overcoming this initial coordination barrier is the first problem to solve. Some minimum coordination will be necessary to jumpstart the process. But this jumpstart could be the very beginning of a coordination explosion, one that triggers a positive feedback loop where initial gains of improved coordination can be pumped right back into the process of further increasing our capacities to coordinate.
This is how our vicious cycle turns virtuous. This is how we can transform our capacity to coordinate into a planetary force capable of solving problems in ways we never thought possible. However this new feedback loop begins, it could be an event that goes down in history as significant as any of the technological breakthroughs leading to artificial intelligence.
However it may begin, it will start by recognizing that this challenge is not hopeless. We are not genetically predetermined to wallow in the woeful state of our current capacity to coordinate. Improvement is possible. Somewhere a catalyst will emerge to unite the efforts that so many of us see as worthy of pursuit, so we can jumpstart a process of self-improving capacity.
All we need is a little coordination.