The Case for Plan B

Richard Westall, Sword of Damocles. Wikimedia Commons.

In my previous essay, I argued that rising to the challenge of the permanent problem is daunting but manageable: you can get there from here. Working within the existing institutional structures of liberal democratic welfare-state capitalism, we can see lines of incremental progress and reform that lead to a substantially more dynamic and inclusive society – one in which capacities and opportunities for stable, secure, and fulfilling lives are steadily expanding.

It would be exceedingly odd if I didn’t believe all that. After all, I’m a vice president at the Niskanen Center, recently touted as “the most interesting think tank in American politics” for its willingness to break through ideological logjams in search of solution’s to the country’s biggest problems. In a white paper from five years ago that outlines Niskanen’s distinctive policy vision, my co-authors and I wrote:

We believe that this vision offers America its best chance to show itself and the world that the liberal democratic capitalist welfare state — the best model of governance ever devised, producing the richest, healthiest, best-educated societies that ever existed — can still work to improve ordinary people’s lives, expand their opportunities, and engage them with meaningful, respected roles in the great common endeavors of society.

I believed that then, and I still believe it today. What’s more, I don’t just think that the policy reforms we advocate can take us to a much better place; I also think that enough of this agenda can actually get enacted and implemented to make a decisive difference.

But the fact that I’m not winding down this Substack shows that I believe some other things as well — in particular, that developing and enacting incremental reforms within existing institutional structures does not exhaust the set of constructive responses to the permanent problem. Here let me describe anything that might fit into this category of “other” responses as Plan B: possibilities for new institutional forms and structures to organize social cooperation, possibilities for the evolution of existing forms and structures in ultimately radical new directions, possibilities for major cultural shifts that either spur or adapt to institutional change.

I’ve got three reasons why, even if you’re fully committed to making the best of liberal democratic welfare-state capitalism as I am, it makes sense to broaden one’s reformist gaze and consider a wider range of Plan B options. First, given the pace and unpredictability of change in contemporary society, appropriate planning for the future needs to take place on multiple time horizons. Second, although muddling through is possible and worth working toward, it’s not guaranteed. If some of our core institutions buckle under the strain, things will go much better if we’ve already begun cultivating the seedlings of new institutions better adapted to prevailing conditions. And third, muddling through means getting to a place considerably better than we are now — but even so, we may discover there that we are stuck on a relatively low peak of societal fitness, and that moving to greater heights will require radically new ways of doing things.

As with virtually all projects involving humans, pondering, planning, and working on how tomorrow’s society might be better than today’s can benefit from a division of labor — specifically, with regard to time horizons. At the Niskanen Center, our focus is appropriately on the here and now: the major national problems that either press or loom today, and how best to deal with them within the constraints of contemporary reality. We understand that our policy ideas could take years or even decades to come to fruition, but our ideas always need to be responsive to what’s going on now and to shift as conditions shift.

But when we step back from the here and now and take a wider view, we see that our present is situated in an era of incessant, ongoing, transformational change. Just to put things in personal perspective, one of my grandfathers had two grandfathers who fought in the Civil War, and one of them had a great-grandfather who fought in the American Revolution. That’s the whole history of our country in just eight generations. Or in other words, my two great-great-grandfathers were fighting on the eve of the Second Industrial Revolution, the marriage of science and technology that created the modern growth explosion; and the great-grandfather of one of them was fighting just as the First Industrial Revolution was getting underway in England.

On Twitter and his Substack, Brad DeLong has been noodling through exactly how many “modes of production” we have been through over the past millennium. Starting with feudalism in 1000 C.E., it took about 650 years by his reckoning to get to maturity for the next mode, which he calls gunpowder empire commercial societies. That’s where we were when the United States was founded and my ancestor, Joel Stokes, was helping out at the Battle of Guilford Court House. It took another 220 years or so to get to maturity for the next mode, the steam power societies of the First Industrial Revolution — just after the Civil War had bled America dry. Since then, according to DeLong, we’ve changed modes every 40 or 50 years: the Second Industrial Revolution economy of around 1910, followed by the mass production economy of around 1960, followed by the global value-chain economy of around 2000, and culminating now in the contemporary information-biotech revolution.

Brad’s point in working through this periodization is to note the inability of social arrangements and institutions — what Marx called the “relations of production” and the crowning social/cultural/political “superstructure” – to adapt to the underlying technical-economic reality before that reality has changed qualitatively once again. Once upon a time, society had centuries to work out some kind of viable, stable equilibrium; now, though, the rug is pulled out from under us every few decades, long before the adjustment process has had time to sort things out. We scramble madly but we can’t keep up, like Lucy in the candy factory.

Accordingly, although confronting and grappling with contemporary problems as they present themselves is clearly our absorbing priority, at least some of us should be looking down the road toward the next technological dispensation and thinking through the kinds of institutional adaptations that will be desirable or necessary. And we may well come to the conclusion that dramatic changes in technological endowments need to be matched with substantial alterations to our institutional environment.

As we do our best to plan for an uncertain future, we have to take seriously the possibility that our efforts to muddle through will prove inadequate. For one thing, they may be upended by some major shock — some catastrophe, whether natural or human-created, that results in massive disruptions to economic, social, and political order. We’ve just struggled through years of a global pandemic with moderate to high transmissibility but limited lethality. Our luck was bad but could have been far worse, and even so the global death count exceeded 20 million and the combined economic damage amounted to trillions of dollars. And there are good reasons to expect that future pandemics await: if you tried purposefully to design ideal conditions for the zoonotic transmission of new pathogens and their rapid, worldwide spread, the combination of climate change, habitat loss, factory farming, and dense global travel networks would be hard to improve upon. And with really rotten luck, one of those pathogens may be as deadly as it is easy to contract.

Other natural perils suspend their swords above us as well. Asteroid strikes, supervolcano eruptions, massive solar storms that wipe out our electrical grid — it could be thousands of years before one of these hits, or it could be next week.

There’s also the possibility that we do ourselves in. Humans have been able to create hell on earth at the local level since early days, but it’s only been since that first Trinity test at Alamogordo that we have had the capacity to wipe out civilization across the planet. Since we fired a couple of shots in anger, that proverbial rifle has been hanging on the wall for 77 years now — and in what we may presume to have been the first act of the play, the period of U.S.-Soviet superpower rivalry, it was nearly taken down and used on several occasions. What act are we in now, and how long until someone’s trigger finger can’t resist the itch?

The risk of nuclear catastrophe is limited by the scale and complexity of the effort needed to threaten one: nuclear warfare is a business limited to nation-states, and starting one requires the successful navigation — or circumvention — of elaborate procedures. But perhaps the single most unnerving tendency associated with technological progress is the one that bends toward putting ever greater powers in the hands of ever smaller groups of people. The recent covid-19 experience has revealed just how far this tendency has now advanced. Although it is generally believed that the novel coronavirus originated naturally, it’s possible that it leaked accidentally from the Wuhan Institute of Virology. Security at such labs turns out to be far more lax than might be expected, and accidental releases are not exactly uncommon. Moreover, we’ve learned that scientists around the world are engaging in poorly supervised “gain of function” research that could produce genetically engineered pathogens of nightmarish ferocity. To my mind, the creation and release of such a super pathogen, whether intentionally or by accident, looms as the single most plausible existential risk we face.

The one getting all the attention these days, though, is artificial general intelligence. I confess that I have no idea how to evaluate the danger here. Let’s contrast the risks of efforts to develop AGI with what seems to me to be the much clearer case of gain-of-function research. The latter promises modest upsides (the possibility of understanding pathogens better and developing more effective vaccines and treatments) and apocalyptic downsides with clearly understandable mechanisms for causing harm (germs that can kill most or all of us). With AGI research on the other hand, the potential upsides and downsides both seem virtually unlimited, but the mechanisms for how the downsides might play out are all completely conjectural (at least for now — it’s easier to picture how humans might lose control once AGI systems and robots to do their bidding in the world are both ubiquitous). What I do know is that many if not most of the people most deeply involved with AGI research believe the risks to be considerable. Here is one particularly measured and thoughtful statement of the case.

I will say that the “alignment” problem — making sure AI is in sync with humanitarian values — is much broader than uncontrollable autonomous superintelligence. The most straightforward, and to my mind much more immediately concerning, kind of misalignment comes from nonautonomous AI in the hands of bad people. My colleague Sam Hammond, whose writings about AI have been consistently illuminating and thought-provoking, muses about some dark possibilities here. There is also the prospect of AI in the hands of hostile regimes — namely, China. Such national security concerns militate against any AI research pause that doesn’t include the Chinese.

There’s no need to sort any of this out for present purposes, however. What’s clear is that the massive uncertainties associated with continuing AI progress and the possible emergence of AGI offer very good reasons for Plan B thinking. Indeed, the need for Plan B thinking is even more obvious if we take the lofty upsides of AGI seriously: it’s hard to imagine how social institutions wouldn’t change radically in a world guided by artificial superintelligence.

Finally, our self-destructive capacities now extend to changing the planet’s climate through massive emissions of carbon and other greenhouse gases. At present the risks do not appear existential: the best estimates are that the ultimate temperature change will be well short of apocalyptic, and the harshest effects will sadly be visited on the world’s poorer regions rather than its most advanced. These estimates are just estimates, though, and even so they point to a future facing serious disruptions and difficulties. These stresses, along with everything else that may be going on, could be enough to cause calamitous breakdowns in social order.

All of these disaster scenarios point toward the need for a specific kind of Plan B thinking: namely, ideas for new institutions and arrangements that can make our technology- and organization-intensive way of life more resilient. If we do face serious risks of catastrophe over the coming decades and centuries, how can we better distribute key capacities so that a major shock in one place doesn’t bring the whole edifice down?

Civilization-disrupting shocks are one way that muddling through could fail. But such failure, if it comes, is more likely to arrive with a whimper than a bang. This blog up to now has provided a lengthy compendium of the economic, social, and cultural forces that hinder the progress of liberal democratic welfare-state capitalism toward wider human flourishing and indeed push in the opposite direction. If those dark forces cannot be successfully resisted and beaten back, where will they end up carrying us?

My own sense of the likeliest dystopian outcome is one in which things gradually devolve into an ever more repellent caricature of where we are today: a cossetted and insular elite that maintains plebeian acquiescence in its domination through a combination of intensive surveillance and digital bread and circuses; the general abandonment of efforts to improve the material conditions of ordinary people in favor of facilitating their escape to frictionless virtual fantasy worlds; the ongoing substitution of stimulation, spectacle, and distraction for relationships with flesh-and-blood people and engagement with the physical world; the continued triumph of safetyism and the therapeutic ethic, and the consequent marginalization of value systems that prioritize anything above healthy bodies and calm, pleasant affect; the progressive breakdown of epistemological and ethical universalism and retreat into identitarian tribalism; deepening estrangement from the idea that history has anything valuable to teach us other than a catalog of unenlightened horrors; and thus, in sum, the dismal flattening and narrowing of existence that is the inevitable result of disconnection from all the primary sources of human greatness and fulfillment.

Notwithstanding all the foregoing, accepting the need for Plan B thinking does not require a pessimistic frame of mind, or even a dutiful, prudential commitment to consider the full panoply of downside risks. We can assume that muddling through will succeed — that, working within the existing institutional framework, we can build a much better world. All that is needed is mindfulness of those longer time horizons and a bit of imaginative curiosity: a recognition that our existing institutions, supremely useful in the past and still serviceable in the present, are unlikely to be the unchanging end point of social development.

To make this a bit more concrete, let me refer back to a passage from this blog’s kickoff essay:

When we shift our gaze to more distant time horizons and contemplate the longer-term future, it seems likely to me that the social institutions and ways of life we inherited from the past will need to be reimagined and in some cases transformed to meet the challenges of the permanent problem. While I don’t see mass technological unemployment as imminent, I am still dubious that mass employment for wages at workplaces outside the home is the final word on how the division of labor is to be organized. I am dubious that consumerism is the final word on society’s primary motivation for productive work. I’m quite sure that commercial mass media are driven ineluctably toward toxic divisiveness and sensationalism in their coverage of public affairs and that some alternative is desperately needed. And I’m increasingly of the opinion that elections are not always the best way of choosing our political representatives.

I should hasten to add that I don’t see these bedrock institutions of contemporary society — wage employment, consumerism, commercial mass media, electoral democracy — disappearing from the scene. That is, I can’t imagine happy circumstances under which these institutions have been supplanted altogether. But my sense is that all of them are overloaded: they’re bearing more weight than they ought to, being pushed to function beyond their appropriate limits and consequently malfunctioning. They don’t need to be replaced; instead, they need to be relieved of some of the excessive responsibilities we currently task them with and supplemented with other institutions that complement and balance them.

This idea that our current social order is overextended — that we have pushed key social institutions well past the point of sharply diminishing returns — is central to my thinking about how society needs to develop to rise to the challenge of the permanent problem. I will explore that idea further in my next essay.