Narrative Slipstream Effects

This entry is part 13 of 14 in the series Psychohistory

Drafting is a behavior in bird-flock-like systems where one agent rides the slipstream of another in a way that delivers a collectivizable benefit, usually net energy savings. The instantaneous savings rates from drafting can be very non-trivial, ranging from 5-50%, depending on the agent geometry, formation topology, physics of the situation, and other conditions. Birds, bicyclists, race-car drivers, long-distance runners, and truckers on highways do it. It is possible to do it with airplanes, though the technology hasn’t been commercially deployed yet, as far as I know. Autopilots capable of maintaining the precise wingtip-to-wingtip formations required, for long periods (which human pilots can’t do), were developed in the early 90s. It is possible to do it with driverless cars. The reason only race cars do it today is that it requires precision bumper-to-bumper driving in platoons (linear formations of several cars) at high speeds, which ordinary drivers can’t do. A project back in the 90s, the Berkeley PATH project, demonstrated this with specially kitted-out Buicks on specially modified “smart” highway segments. Teslas today have the hardware and software capability to do it. The main blocker is not technological, but legal: who will be held liable if a platoon crashes?

Drafting offers a very fertile metaphor and mental model for social systems comprising individuals who “travel together” in a conceptual space with political, cultural, and economic dimensions. Something like Hofstede’s cultural dimensions model might be a suitable metaphoric space for thinking about societal formation flight. The equivalent of the shared travel path is the shared grand narrative, and we can think in terms of narrative slipstream effects delivering the benefits of drafting.

In some ways, narrative slipstream effects are the evil twin of network effects. You draw power from a shared goal rather than a shared capability. Network effects drive behavioral divergence and variety; narrative slipstream effects drive behavioral convergence and homogeneity. I’ll illustrate this with examples later.

The thing about drafting is that you need long stretches of aligned, unchanging orientation, at relatively high speeds, to get the benefits. So trucks on a long stretch of highway can draft, and bank fuel savings, but trucks navigating busy city interiors can’t. One of the problems I worked on for my PhD was figuring out, for some simple cases, whether or not it was worth it for 2 agents with different origins and destinations to meet up, and get into a drafting formation for a shared leg of their journeys. As you might expect, it’s mostly worth it when you can construct a long shared leg, such as with two airliners on a transatlantic flight. Say a NYC-London flight hooking up with a Boston-Paris flight. You’d rendezvous close to the US East Coast, and split up close to Europe. In between, for a couple of thousand miles, you’d enjoy the benefits of drafting.

The idea struck me as a nice metaphor for what grand narratives actually are. Grand narratives can be considered large-scale formation flight paths through narrative space, modeled perhaps as a Hofstede space crossed with time. When thousands to millions of people are living out versions of aligned, small stories within a larger one, some stories will represent “drafting” off other stories. Stable societies can be understood as drafting formations.

There is a basic systems engineering problem though: formation travel is a fundamentally unequal regime of cooperation. In V-shaped geese formations, the apex bird is expending more energy than the rest. This problem fixes itself with geese: the lead bird tires out faster, and starts falling behind, and an adjacent bird takes the lead. The V gets more asymmetrical as the formation rotates, and can end up rebalancing itself over time in messy ways. Birds very far back along a leg of the V, which get fewer benefits, might reposition, or fork off new V’s. As these complicated second-order effects play out, you get those long, straggling fractal skeins, with many V’s, that you’ve probably seen during migration season.

With trucks, bicyclists, or airplanes, you’d have to work out a fair scheme to rotate the lead position (or beneficiary vehicles could perhaps compensate lead vehicles financially — this would be an ideal blockchain application). Ridesharing (with actual sharing) is a weaker version of formation flight, where many riders share a car, instead of many cars drafting.

Public transit is of course an extreme case, based on aggregating travel on long-term stable shared travel routes. The physics and economics details differ from situation to situation, but the broad phenomenology of varieties of formation travel is the same.

In societies, there is an obvious mapping to how hard people in different societal roles work, though the variance in input factors is higher than with the flight abilities of geese.

Arguably, entrepreneurs are at the apex of the formation, working harder than average to produce new wealth from a wealth-production technological frontier, using invested old wealth. In corporatist states (which is basically all developed states to different degrees) where the state takes the lead on innovation, this role might be played by taxpayer-funded scientists, with entrepreneurs lined up right behind them, doing their first-order drafting by jumping on NSF or DARPA-funded research findings. This wealth trickles down the legs of the V formation, in rough order of class. Working-class individuals get the smallest benefit because they’re so far back, the potential energy of the slipstream is mostly gone. The sweet spot is probably some sort of middle-class position, with relatively secure and easy kind of work. Like say being a government employee in a sleepy department, or a tenured professor. You’re not so far ahead that you have to work harder than average to generate the collective slipstream potential, but you’re not so far back that all the surplus (aerodynamic potential) is used up by the time the slipstream hits you.

In societies, unlike in goose or vehicle formations, there’s no natural or designed rotation in the higher-energy-cost leader position. Instead, you have dissipative generational wealth-transfer mechanisms, and slow social churn and vertical mobility “rotating” the formation. New entrepreneurs will tend to emerge unpredictably, but not entirely randomly (they are more likely to emerge from inherited wealth classes that were recently benefitting the most), and take up leadership positions in new wealth creation.

It’s kinda messy, but you can actually model the economics of this formally in the case of societies as well. In grad school a couple of decades ago, I did a short independent study project (with Robert Axelrod, he of Prisoner’s Dilemma fame) applying the formation flight models from my main PhD work to economies, using a framework known as turnpike theory. The idea is that economies have a path of optimal long-term growth that is mathematically similar to the “formation flight” leg of agents traveling together:

…if we are planning long-run growth, no matter where we start, and where we desire to end up, it will pay in the intermediate stages to get into a growth phase of this kind. It is exactly like a turnpike paralleled by a network of minor roads. There is a fastest route between any two points; and if the origin and destination are close together and far from the turnpike, the best route may not touch the turnpike. But if the origin and destination are far enough apart, it will always pay to get on to the turnpike and cover distance at the best rate of travel, even if this means adding a little mileage at either end. The best intermediate capital configuration is one which will grow most rapidly, even if it is not the desired one, it is temporarily optimal.

From Linear Programming and Economic Analysis (1958) by Dorfman, Samuelson, and Solow

In the case of the turnpike model I studied, the “turnpike” is the formation regime rather than a feature of the landscape. For the simplest illustration, think of 2 agents moving in 1 dimension; with positions x1 and x2. The turnpike in that case is the 45 degree x1=x2 line on the x1x2 plane, which we called the coordination plane. My late advisor, Pierre Kabamba, came up with the basic idea of coordination space, which I then ran with for analyzing formation flight rendezvous and breakup. An economist friend pointed out the similarity to turnpike theory, which is how I ended up doing the fun little economics side project.

My basic mapping was that if you treat x1 and x2 as the cumulative progress enjoyed by 2 classes in society (measured by something like income share), say rich and poor, getting in formation on the x1=x2 line (or one with a different slope) is a way to model the hypothesis that equitable progress is in some sense optimal and efficient. You still have to come up with arguments for why the turnpike is efficient in this case, analogous to the drafting physics in literal formation flight. For example, you could argue that off the equitable growth turnpike, either the poor turn to crime, or the rich turn to rent-seeking, or both. Many such arguments are possible, but the basic thought is that a “formation flying” society enjoys surplus collectivizable benefits that a fragmented society doesn’t. If that surplus is high enough, people form and sustain societies. If it doesn’t they strike out on their own. The specific argument depends on the specifics of how wealth is being created and shared in the society.

There’s a cute connection to prisoner’s dilemma game theory: getting on the “turnpike” is equivalent to hitting the cooperate-cooperate evolutionary stable state (for example, via two players using tit-for-tat strategies). You can then milk the C-C payoffs indefinitely until one player defects. I didn’t work on the project long enough to connect the game theory interpretation to the formation flight metaphor mathematically, but I suspect it’s possible, because continuous game theory is already used to model pursuit-evasion laws for missiles and targets (which can be considered an adversarial regime of formation flight).

Here’s the thing though: “societal formation flight” only works while societies enjoy long periods of stable orientation and aligned heading. Decades of aligned economic interests is the equivalent of long, shared flight segments. If rich, middle class, and working class face a wealth-production frontier that can be expected to last for decades, it’s worth it to “get into formation” on the turnpike by forming a stable societal order. It’s not necessarily going to be equitable, but being part of the formation will beat not being part of it. There will be some net benefit.

What’s a “shared” heading for societies in this metaphor? The cliche answers are well known: a space race, an external adversary (like the Soviet Union and the US for each other during the Cold War), religion (the Crusades), and so on. These are actually bad answers though, since they represent zero-sum growth contexts, where your growth comes at the expense of someone else’s decline. Better answers appeal to a technological wealth frontier, such as fossil-fuel based technologies, or Moore’s Law.

It doesn’t really matter what the frontier is, but it has to satisfy the following criteria to deliver the benefits of societal “formation flight”:

  1. The receding target horizon should be maintainable a few generations out in time
  2. Those who benefit least (V-tail geese) should still do better than they would solo
  3. The renewable wealth frontier should be partially legible, with correlated opportunities

The first condition can be interpreted as follows: two trucks in a drafting formation will stay in formation for a long stretch of highway, but as you approach the set of exits where one needs to get off, there will be an optimal “defection” point at which that truck should break away and exit. If you are wealthy, and you think broad-based prosperity will collapse soon, you will “exit” and try to engineer defensible rents for your useless progeny to enjoy for as long as possible, in an expected future of declining (negative-sum) prosperity. If you think growth and prosperity will continue indefinitely, you’re more likely to leave more of your wealth to society at large in some form, because your useless progeny will benefit more from rising tides floating all boats.

Arguably, the shortest answer to the ever-popular question of wtf happened in 1971 is that modern American societal formation flight ended because the incentives for the wealthy to stay in formation collapsed.

The middle class and poor too, have exit options. To the extent it is real, the “Great Resignation” (and the similar “laying flat” trend in China) is something like an exit from formation flight.

The second condition is the obvious minimum to prevent class revolt, though stronger classes can manipulate the perception of value by weaker classes in various ways through violence or false-consciousness programming. Being in the formation need not actually be better than going off solo. It must only seem that way. Weaker classes must view exit as a worse option, whether or not it is actually worse (the ongoing effort to get the middle class to return to work from offices feels like this kind of programming). You don’t need TINA (There Is No Alternative), you only need belief in TINA.

The last condition is the most interesting to me. For societies, you get this effect when the economy is being driven by a growth thesis, correlating sources of wealth over a clear time horizon, with an uncertain but exciting future beyond. Usually, this is a technological age with a receding horizon of recognized, correlated opportunities that you slowly catch up with. When you are close enough, and the horizon is receding slowly enough, you enter some sort of wealth-production endgame, where it seems like the renewable wealth-production source is drying up.

Peak Oil is an example of one such endgame narrative milestone, one which we thought was close for a while, but turned out not to be.

Another is Moore’s Law potentially coming to an end around the 1-nanometer mark, which is theoretically a couple of doublings (or about a decade) away. But as with Peak Oil, we may have more time than we think. Perhaps we can get further down to about 500 picometers, at which point we’re talking chip features tens of atoms wide, and quantum shit will definitely get in the way.

Note that these narrative milestones that govern cooperation/defection behaviors don’t need to be real, or based on true models of reality. The participants in the economic order just have to believe in them. If you’re on a “fossil fuels” wealth production/growth turnpike or a “Moore’s law” turnpike. You just have to believe the highway is ending, and that you have to break away and find a new source of wealth. Which may or may not favor coordination with your current formation partners. The story ends when enough people stop believing in it, not when the opportunity space actually dries up.

The correlation requirement is interesting, and the catalyst for cooperation. It is not sufficient to have a wealth-production frontier of random opportunities. People in a casino full of uncorrelated slot machines have no particular reason to talk to each other, and mostly don’t.

The opportunity space has to be correlated in ways that foster confidence in indefinitely extended cooperation. When wealth streams from such frontiers begin to falter and fail, the incentives to defect from the grand narrative go up. You then get a more atomized exploratory culture that my buddy Carlos Bueno dubbed an antiflock. Instead of a gold rush to one spot with a rich vein, you have people dispersing over a larger area.

I suspect we may be on the verge of an antiflock era of history now, at all levels.

There is a case that could be made here that history is in fact a series of convergences and divergences driven by the rise and fall of correlated opportunities on wealth frontiers. That the only reason we ever coordinate enough to form societies is to go after such frontiers, and that otherwise we’re not actually a very social species. I suspect this is what many libertarians believe.

The antithesis is that collectivist coordination is its own reward; that we “formation fly” because we simply like mutuality and interdependence more than isolation and independence, and any collectivizable rewards for doing so are simply gravy.

I think, perhaps, both are true. Human nature is varied enough that there are probably genuine (genetic) strong individualists and mutualists among us, with narrower or wider ranges of preferences, around varied set points. So at societal level, a very large range of convergence and divergence regimes are possible. Depending on where wealth and prosperity are coming from, you could get anything from Asimov’s Spacer type worlds where individuals live on vast private roboticized estates with little to no mutual contact, all the way to Gaia/Borg-like extreme coordination conditions.

I do suspect the extremes are unstable though, because they are too legible. Smart people will be able to calculate and optimize selfish strategies. Hence partial legibility in the last condition. The opportunity space must be somewhat legible, and seen to be correlated enough to inspire cooperation, but if you can see it too clearly, and especially see its limits too clearly, cooperation will break down. At the edges of visibility the opportunity space must dissolve into myth.

But to actually work to drive convergence, a grand narrative built around a wealth-production turnpike must produce an actual slipstream with drafting opportunities. You can’t simply make up a random narrative and expect this to happen. At least not in 2022, when the bar has been set so high by technological modernity for actual prosperity. This is something authoritarians and theocrats fail to see.

When the narrative is forced top-down, as in the case of the original space race, all the promised benefits are in the future, constituting a sort of mythologized afterlife that drives coordination in ways similar to religion. In the case of the space race, at least for a while, everybody was excited about an afterlife of boldly exploring and settling the galaxy, or at least the solar system, in spaceships. Though the space program had immediate benefits too, they were much narrower, and do not explain the societal formation flight that the space race drove.

Such top-down narratives are fragile. The present remains zero-sum or even negative-sum: money for the space program has to be taken from (say) social services, often via inefficient state redistribution, until the afterlife heaven is actually achieved. What saved the space race from the fury of disappointed citizens was that the LEO economy (communications, weather, remote sensing, GPS), though less exciting than space colonies, proved valuable enough to eventually pay for it all many times over. Not the heaven of galactic life, but a positive return here on Earth at least.

Fossil fuels and Moore’s Law are much more robust narratives with very strong real slipstreams. Both started top-down, via capital-intensive opening-up of frontiers, but quickly transformed into broad frontiers accessible to almost everybody.

Both produce real benefits in the present, which exceed the immediate costs of externalities, but also catalyze dimly glimpsed afterlife regimes of mythic proportions. In the former case, the “afterlife” was initially seen as a mid-century-modernist heaven (flying car cities, GM’s original 1939 Futurama vision) and later as hell (pollution, ozone depletion, climate change). In the latter case, computers produce real benefits in the present, have comparatively fewer externalities, and hold out a different promise of afterlife abundance (software running on solar-powered computers eats everything, AIs and robots do everything, we solarpunks just sit back and enjoy ourselves).

These examples suggest that a grand narrative produces a slipstream capable of sustaining cohesion in large societies if it satisfies the following conditions:

  1. It produces benefits today
  2. It promises an afterlife where heaven >> hell
  3. There is a consolation prize even if we don’t get to heaven

These are quite stringent criteria that are not easy to satisfy. For example, the new space race is right on the edge of eligible. The low-hanging fruit of the LEO economy is already won, so that’s not enough. If Elon Musk (or someone else) hits the target of $10/kg to LEO (two orders of magnitude better than today), the prize is not communications satellites, space telescopes, remote sensing, or Starlink. The prize is cheaper access to regimes beyond LEO to reopen the abandoned 1970s vision of the space frontier. Like the moon, Mars, asteroids, and industrial production in space (the Jeff Bezos vision).

That’s a pretty sweet heaven >> hell, but it is not clear that the first and third criteria are satisfied. And the “hell” in this case is darker than what was perceived in the 60s. In the 60s, the negative externalities of the space race were primarily conceived in terms of more dangerous nuclear weapons. Today, the primary negative externality we worry about with space is GHG emissions (high-altitude emissions are especially bad).

The point of this discussion isn’t to argue the specific merits or flaws of the new space race. It is to probe the question whether it is a sufficiently fertile wealth-creation frontier to induce grand narrative slipstream effects strong enough to produce societal formation flight. Much as I am all for the new space race, I doubt it will do the trick.

When you look around at all the major candidates for wealth-production frontiers today — blockchains, AI, biotech, renewables — none of them seems capable of producing enough of a grand narrative slipstream to bind today’s complex and large modern societies. So I have to conclude that we’re in for a great deal of antiflocking and divergence.

But at the grand narrative level, everybody is getting off the big turnpike, and there is not going to be a grand narrative slipstream for a while. But smaller scale narrative slipstreams — from family to city-scale — may still be viable.

Series Navigation<< A Dreaming WorldRealtechnik, Nausea and Technological Longing >>

Get Ribbonfarm in your inbox

Get new post updates by email

New post updates are sent out once a week

About Venkatesh Rao

Venkat is the founder and editor-in-chief of ribbonfarm. Follow him on Twitter

Comments

  1. interesting on turnpike theory, which posits economies of scope or information. another take is that all progress is a function of discovery in a solution space. the discovery paradoxically is a function of the scale of production itself either over time or scale space. Ie. the more we make of a thing as an economy the more of the solution space gets explored. As the solutions space (technological, informaiton, production scale etc.) gets explored progress. Interestingly it is somewhat bounded. Wrights law for progress is a single vector expressioin of this while an economy is the interconnected version, but still effecitvely enabled/limited by wrights law. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0052669 just sayin…

  2. Ravi Daithankar says

    The question then becomes, can a convergence on a “there is no grand narrative” narrative itself induce any meta-slipstream effects?

    What happens when everyone gets on the nihilistic “this turnpike goes nowhere” turnpike? What do the exits look like and what does the horizon look like? Abyss and abyss?

  3. Great post. I’m sending this to a romantic details econimics professor.

    And lucky you knowing Pierre Kabamba and finding “It”, being the Gervais Principle guy.

    …”That people saw him as the “sampled-data guy” was not a burden.

    “It was a helpful narrative entry point into the thought spaces he liked to explore, but it didn’t constrain him. And he wanted me to build a doorway like his into my own head-spaces.

    “It was advice that served me well far beyond the limits of academic control theory. Today, I’m the “Gervais Principle guy.” If I hadn’t imbibed Pierre’s way of thinking about such things, that would bother me. Instead, it delights me. 
    From;
    Remembering Pierre Kabamba
    *

    VR:- “It is a fully realized theory of management that falsifies 83.8% of the business section of the bookstore.  The theory begins with Hugh MacLeod’s well-known cartoon, Company Hierarchy(below), and its cornerstone is something I will call The Gervais Principle, which supersedes both the Peter Principle and its successor, The Dilbert Principle. Outside of the comic aisle, the only major and significant works consistent with the Gervais Principle are The Organization Man and Images of Organization.”
    https://www.ribbonfarm.com/the-gervais-principle/

  4. Rory Kaufmann says

    Good post! I liked the car analogy, clear + concrete. Overall felt like a mature take on alignment. This is one of those concepts that can be applied to everything. Easy to think about, hard to know what to do with.

    The bit about a shared gold rush reminded me of a game I’ve been playing. Milky Way Idle. It has a real (play) stock market, and after living on it for a few days, you start to see the ebbs + flows, people spreading out to cover opportunities, and how the market responds to attention. It feels strangely lonely, there are real people out there on the other side of the trades, but I only see their tracks and it’s all mediated by a system that feels more alive than they do.

    Alignment/communication is shaping up to be huge for AI in a few years. I don’t buy into the Singularity super-AI hype, but with the progress in art and playing games, it’s getting clear that there’s something there, as much as anything can be said to be anywhere.

  5. The racecar in the lead of a draft gets a noticeable benefit: a filling-in of the space at the back of the car, normally occupied by a partial vacuum. If it were not so, then the “platoon”, as you call it, could not go much faster than the singletons that it overtakes; but it does. The reason that cars need to trade around there is that following cars lose a benefit of dense, clear air — namely cooling — to a hazardous degree.

    I suspect that something similar helps a lead goose somewhat, too.

  6. Do you think DeFi creates a grand narrative slipstream?

Leave a Comment

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.