Like some of you, I have recently spent a greater than average amount of time dwelling on the possible and proximate onset of a nuclear dark age. But, I've been interested in how we’d rebuild from such a catastrophe for a while. 

Ethos: I’m a researcher in the global health and wellbeing space and have practically no knowledge of this area except for a few books, podcasts, and a paranoid sister. I’m out of the longtermist loop, so call me out if I’m ignoring some classic forum post or showing similar ignorance.

I think of the cause area of existential risk as “prevention and recovery from capital "c" Catastrophes”. Prevention gets more attention than recovery. I see why, but I wonder if there is more room to lighten the next dark age[1] and save our potential to pave the galaxy in hedonium. It seems like there’s more room for organisations to work on recovery because it seems, on brief reflection, that there are some tractable interventions to work on and just about no one (???) working on them. 

Below I list some rough ideas about what we could do to shorten the next dark age. Some of these may be new—others I’ve seen and am reposting for list-making purposes. A few of these were briefly mentioned in a recent forum post

  • Identify promising areas to reboot civilisation from (AKA Foundations). What are the countries or parts of countries where it would be cost-effective to prepare the local government and economy to survive an apocalypse? I hear New Zealand commonly mentioned as relatively robust to catastrophes, but it seems plausible that other places have better characteristics. What about New Caledonia, Tasmania, the Azores, Puerto Rico, or Patagonia in Argentina?
  • Distribute books like The Knowledge to libraries in places that are best placed to weather doomsday.
  • Collect “The Knowledge +”. One book is insufficient to store enough information to rebuild civilisation. What are the other books that are essential to read? Which Wikipedia articles could act as a deep digital appendix to The Knowledge?
    • Perhaps we could also push the values of the post-collapse world towards EA alignment by adding generally wise and entertaining fiction and nonfiction.
  • Create a device that’d store "The Knowledge +".  I imagine an e-reader fitted with a solar-powered charger and the most reasonably durable battery. These could potentially be sold commercially to recoup some costs.
  • Design a post-apocalyptic video game that teaches the basics of rebuilding civilisation. An easier route would be to modify existing games like Don’t Starve to add more technical details like the Haber-bosch process. This also has potential commercial viability.
  • Allfed is an EA organisation that advances technology to feed the world if there is a catastrophic shortfall in food production. Allfed is the only organisation I know of explicitly focusing on catastrophic risk's “recovery” part. It seems like it could be a good idea to move part of Allfed to a location more suitable for restarting modern civilisation than Alaska.
  • Facilitate geographic EA community diversification. I think EA ideas should survive a catastrophe. We could make that likelier by having EAs themselves survive the calamity. Taking the long view, this could be especially valuable in the locations that will be highly influential in the next age. If we identify clear areas that will most likely withstand nuclear winter or a super-pandemic, we should be interested in having strong and diverse EA communities there.
  • Build a broader community around effective prepping? This would involve first building a community of “preppers” whose preparations are more calibrated to the likeliest catastrophes (e.g., nuclear war is more probable than a leftist coup d’etat). Better preppers would presumably lead to more survivors in case of a calamity. The second part of this effort would involve imparting better values to the likeliest survivors of a catastrophe.
    • Found more self-sufficient proto-utopian communities? I’m sceptical that founding experimental communities will change the present world dramatically. It’s unclear that they had a considerable positive impact in the past. Having communities that have already performed a dry run reboot of civilisation could be a valuable source of knowledge about rebuilding when Ragnarök arrives. I see this as a much more expensive version of previous points.
  • Cause prioritisation question: How bad would a dark age be, anyway? If the average person's wellbeing goes from -5 to 5, what would it have been in the 1300s (a bad century IMO), 1600s or heck, now? This question was inspired by Holden Karnofsky's posts on a long history.

Below I repost existing ideas in the Future Funds Idea List

  • (5) Infrastructure to recover after catastrophes: We want to ensure that humanity is in a position to recover from worst-case catastrophes. For example, we’d like to ensure that humanity has reliable access to the tools, resources, skills and knowledge necessary to rebuild industrial civilization if there were a global nuclear war or a worst-case global pandemic. We’d be especially keen to see “civilizational recovery drills”: attempts to rebuild key industrial technology with only the tools and knowledge available to survivors.
  • (11) Keeping coal in the ground: We’d be excited to see a project that enables donors (governments and philanthropists) to easily buy coal mines and retire them. Other approaches to keeping fossil fuels in the ground interest us as well—re-industrialization after a civilizational collapse could be much harder if fossil fuels are scarce.
  • (12) Demonstrate the ability to rapidly scale food production in the case of nuclear winter: In addition to quickly killing hundreds of millions of people, a nuclear war could cause nuclear winter and stunt agricultural production due to blocking sunlight for years. We’re interested in funding demonstration projects that are part of an end-to-end operational plan for scaling backup food production and feed the world in the event of such a catastrophe. Thanks to Dave Denkenberger and ALLFED for inspiring this idea.

Next are ideas taken from a recent list of proposals for new ideas to the FTX Future Fund

Greg Colbourn suggested: Should the worst happen, and a global catastrophe happens, we want to be able to help survivors rebuild civilisation as quickly and efficiently as possible. To this end, burying caches of machinery that can be used to bootstrap development is a useful part of a civilisation recovery toolkit. Such a cache could be in the form of a shipping container filled with heavy machines of open source design, such as a wind turbine, an engine, a tractor with backhoe, an oven, basic computers and CNC fabricators, etc. Written instructions would also be included of course! Along with a selection of useful books. First we aim to put together a prototype of such a cache and test it in various locations with people of various skill levels, to see how well they fair at "rebuilding" in simulated catastrophe scenarios. Learning from this, we will iterate the design until at least 10% of simulations are successful (to a what is judged to be a reasonable level). We ultimately aim to bury 10,000 such caches at strategic locations around the world. Some will be in well known locations (for the case of sudden catastrophe); some hidden with their location to be automatically broadcast should a catastrophe be imminent (to protect them from vandals and malevolent actors); and some hidden with some level of "treasure hunt" required to find them (to provide longer term viability should first attempts to rebuild fail).

Edit: Greg changed the text of his comment to indicate a variety caching strategies beyond concealment, so what I say next makes less sense now. I think this is a fine addition to the list, but having to “treasure hunt” for them seems like it do more harm than good. I expect that we will stumble into most catastrophes. If that’s the case, we should make these caches easy to find. 

PhilC proposed: In the event of GCRs, conflicts or disasters, communication systems are key to sensemake and coordinate effectively. They prevent chaos and further escalation of conflicts. Today, there are many threats to the global communication infrastructure including EMPs, widespread cyber attacks,  and solar flares.

To elaborate on this, it seems like having a backup network of HAM radios with independent energy equipment to power them could be a very cheap backup. 

Feel free to contribute to this list or to add any comments or criticisms. If you're anything like me you should try and savor the fact that things are really good, for us, for now. And try not to freak out. 


  1. ^

    A civilisational collapse is bad because fewer people will exist with lower levels of wellbeing. I probably weigh the latter more than the former. I assume large swatches of the population have net negative wellbeing in a dark age. A catastrophe is also dire because it’ll probably knock civilisation dramatically off course. I don’t see “building back better” as remotely likely. I expect catastrophe survivors will have dramatically worse values, and whatever they build will depend on those ideas. To be clear, I expect people to collaborate and sacrifice during short term disasters. However, if the global food supply dramatically shrinks, I don’t expect those who remain to be the nicest.

Sorted by Click to highlight new comments since:

Thanks for the post.

It seems like it could be a good idea to move part of Allfed to a location more suitable for restarting modern civilisation than Alaska.

Only two members of ALLFED are in Alaska-the rest are other places in the US, UK, Canada, Spain, Germany, Australia, India, and more countries if you count board members and volunteers. But it is still NATO city majority, so we are looking at methods of reducing risk to the team.

Great post. I find your idea of a post-apocalyptic game particularly interesting. I think this may be not only promising in increasing catastrophe resiliency, but also perhaps in simply being a fun way for people to gain useful knowledge even if a catastrophe never happens (potentially by gaining knowledge of basic science etc). I've been thinking about this latter proposition recently anyway—I think there's a lot of scope for improving education with educational games that are actually fun to play—but the idea of hitting two birds with one stone seems like a great prospect.

I've just been writing a post about this very idea when I stumbled on this post, so I'd love to continue the discussion!

My tentative intention would be translating the information found in The Knowledge, or some subset of it, into a video game. In order for it to be cost-effective, the video game would need to be:

  • Actually a good game. I suspect many games designed as edutainment (or having social impact, a political message etc.) fail to achieve their goals, while fun games which just “happen” to teach something can be quite effective. Programming basics (Human Resource Machine, a good few other games), physics (Kerbal Space Programme), history (Crusader Kings and a multitude of other games), the exact steps of performing an appendectomy (Life & Death - I admit it's quite niche, but I actually still remember how to do it ;) ), etc.
  • Easy to implement with a small team and on a small budget. Modding probably is the way to go, though it might interfere with making the game available for free, which is my next point. As much as I love Don’t Starve, modding This War of Mine,  Rimworld or NEO Scavenger might be a better idea, since those games have more realistic aesthetics, which might enable us to reuse more assets. One Hour One Life might also be a good one to look at since it’s already open source.
  • Free and cross-platform (or at least playable on mobile). We would want to reach as many people as possible, and checking off these 2 boxes would probably boost it by one or two orders of magnitude compared to a paid PC game.

Some reasons why creating this video game (or an EA-aligned video game) might be a good idea:

  • I conjecture that many more people would engage with a video game presenting a particular topic in a fun way than with a non-fiction book on the same subject. A relevant comparison (cursory research suggests that for both books and video games reviews correspond to roughly 2% of sales):
    • The Knowledge: How to Rebuild our World from Scratch by Lewis Dartnell has roughly 2,100 Amazon reviews (UK and US storefronts combined). 
    • NEO Scavenger by Blue Bottle Games has 3,700 Steam reviews (9,000 if you combine PC and mobile sales), and I think NEO Scavenger is as niche as it gets. This War of Mine is a more successful (but still relatively low-budget) indie game in the same genre, tackling a similar topic, and has sold 4.5 million copies and has 63K reviews. Note: both of these are paid-for games; intuitively, they would probably have had at least 10x more downloads if they were free.
  • There’s probably quite a significant overlap between EAs and people who dabble in coding video games as a hobby. We could capitalise on that; if the game is a spare time passion project for a bunch of us, that could theoretically lead to infinite cost effectiveness - assuming we don’t accidentally create the most addictive video game ever and end up causing more harm than good ;) (For a more serious analysis of cost-effectiveness see below).
  • The broader EA community would hopefully help promote the game, and give us feedback in the design process.

Things to consider: 

  • Cost effectiveness. Realistically, I think the game would need some development budget; this would lower the risk of the team getting burnt out or distracted from the game by other priorities. We’d need to estimate how many people we hope to reach, how much they would learn, and put a dollar value on teaching a random gamer the Haber–Bosch process.  
  • This interesting article on profitability of indie games isn’t exactly heartening – even if we assume that making the game free would broaden our reach, we’d need to sanity-check that we have a greater chance of success than an average game developer.

I have some more thoughts and ideas I’d love to discuss with anyone interested in getting involved in the project.

P.S. As a tangent, I was also thinking about another idea for an EA-aligned video game; something like Reigns or Democracy  3 where you’re a multi-billionaire starting a charitable foundation aiming to solve all the world’s problems. The player would have to assign donations to various causes (possibly including longermism and animal welfare), while balancing PR, fundraising, talent pool etc. In order to not put people off, I think it would need to be very low-key about its EA message, perhaps not even mentioning the term – it’s just that donating to malaria net distribution would “happen” to bring you closer to victory than sponsoring a new museum.

This seems like a cool initial exploration of a potentially important area. Thank you + well done!

The device idea seems intuitively promising.

In a talk at EAGx Oxford, someone (I forget who -- maybe Anders Sandberg) mentioned the idea that the internet archive is a tiny org with not much investment. If that whole infrastructure could be backed up somewhere and stored in a way that would survive loss of electricity/other major systems above ground and also this location publicly known / easily accessible somehow, would that achieve the same purpose?

(I have no technical knowledge about these things and for all I know, this already exists. I'm just spitballing.)

This article series on the Age of Malthusian Industrialism may provide some insight on what the next dark age might realistically look like. One possible way an upcoming dark age could be averted is through radical IQ augmentation via gene editing/embryo selection.

Great post! I've edited the end of my FTX Future Fund project idea to say "We ultimately aim to bury 10,000 such caches at strategic locations around the world. Some will be in well known locations (for the case of sudden catastrophe); some hidden with their location to be automatically broadcast should a catastrophe be imminent (to protect them from vandals and malevolent actors); and some hidden with some level of "treasure hunt" required to find them (to provide longer term viability should first attempts to rebuild fail). "

Thank you! I edited the post to reflect your updated text. 

More from Jpmos
Curated and popular this week
Relevant opportunities