Your submission is now in Draft mode.

Once it's ready, please submit your draft for review by our team of Community Moderators. Thank you!

Your essay is now in Draft mode

Once you submit your essay, it will be available to judges for review and you can no longer edit it. Please make sure to review eligibility criteria before submitting. Thank you!

Submit Essay

Once you submit your essay, you can no longer edit it.


This content now needs to be approved by community moderators.


This essay was submitted and is waiting for review by judges.

Nuclear proliferation, force sizes, & yields

{{qctrl.question.commentStr()}} {{qctrl.question.estimateReadingTime()}} min read
Flourishing Futures Nuclear Risk Tournament

This is a quickly written, rough post intended to provide context for several questions in the Nuclear Risk Forecasting Tournament, mostly via slamming together chunks from various other things I’ve been writing. Comments are very welcome, and could inform what I write in some associated research outputs!

Relevant forecasting questions from this tournament

Relevant questions that aren’t part of this tournament

Proliferation and stockpile sizes

Some people seem to see substantial nuclear proliferation and/or substantial increases in arsenal sizes as more likely than seems reasonable to me. Other people seem to entirely ignore these possibilities and to thus underestimate nuclear risk[1] or fail to notice or appreciate the benefits of some options for intervening to reduce nuclear risk.[2]

Overall, I think it’s unlikely that in the coming decades there’ll be proliferation to substantially more states (e.g., to more than 2 new states) or substantial increases in stockpile sizes (e.g., any state doubling its arsenals). But those possibilities do seem worth thinking about, since they seem plausible - more than 1% chance, maybe more than 10% - and could substantially increase risk from nuclear weapons.

I find it useful to bear the following points in mind:

  • There’s no fundamental reason why there couldn’t be nuclear proliferation to many more states over the coming decades, or why states can’t multiply the total number of warheads or yield of their stockpiles.
  • There’s also no fundamental reason why those outcomes have to occur, or why there couldn’t be de_proliferation or _decreases in stockpile sizes.
  • It seems a wide range of future trends in proliferation and stockpile sizes are plausible, in light of history[3], the rising number of nuclear-latent powers[4], and potential future technological, economic, or geopolitical changes.[5]
  • History also apparently suggests people tend to overestimate the likelihood or pace of nuclear proliferation. (Disclaimer: I haven’t thoroughly read that source.)

Numbers and yields of detonations

The higher the yield of a nuclear weapon, the more immediate fatalities and smoke (from burning buildings and so forth) the offensive use of that weapon can be expected to result in. As such, expected harm - and existential risk specifically - from nuclear conflict is probably higher the larger the expected total yield that would be used if nuclear conflict occurs.[6] In fact, given how “hard” it may be for nuclear conflict to result in an existential catastrophe, it seems plausible that a substantial portion of total existential risk from nuclear conflict comes from low-probability scenarios where total arsenal sizes and yields become several times larger and then a large portion of those arsenals are actually used.

But how likely is such a scenario? Getting at least an order of magnitude estimate of that could help inform decisions about how much various actors should prioritize nuclear risk reduction and, if they do, to what extent they should pursue potential “intermediate goals” such as reducing total yields in state arsenals or differentially reducing the likelihood that (especially) high-yield weapons are used.


[1] E.g., they may focus entirely on how likely extinction or societal collapse would be given _current _arsenal sizes, rather than also accounting for the low-probability, high-stakes scenario in which arsenal sizes rise.

[2] In particular, I think reducing the chance of major increases in arsenal sizes may be a relatively promising and neglected intervention option. E.g., one could aim to reduce the chance of a nuclear arms race between India and Pakistan or between the US and China.

[3] Consider that, in the decades after nuclear warheads were developed, the number of countries that possess nuclear weapons rose to 9 and global stockpiles rose to 70,000 warheads. On the other hand, in the decades since then, few countries have made serious efforts to develop nuclear weapons, and global stockpiles have declined to ~13,000 warheads. See also Our World in Data's graphs.

[4] Nuclear latency is “the possession of many or all of the technologies, facilities, materials, expertise (including tacit knowledge), resources and other capabilities necessary for the development of nuclear weapons, without full operational weaponization” (Pilat, 2014). The number of nuclear-latent states has risen over time and seems likely to continue to rise in future (Pilat, 2014). I think this is probably mostly evidence that it’s plausible there’ll be substantial proliferation, but it could perhaps also be seen as evidence that it’s plausible there’ll be no or very little proliferation, since recent history suggests there can be an increase in nuclear-latency without an increase in the number of nuclear-armed states.

[5] For example, some plausible technological developments may reduce the cost or technological barriers of proliferation or of increasing arsenals sizes, or may threaten to undermine deterrence at current arsenal sizes and thus incentivise states to increase arsenal sizes.

[6] I say “probably” because this can be complicated by things like the nuclear conflict probably being somewhat less likely the more harmful it’s expected to be, since that probably somewhat enhances deterrence.

Metaculus help: Predicting

Predictions are the heart of Metaculus. Predicting is how you contribute to the wisdom of the crowd, and how you earn points and build up your personal Metaculus track record.

The basics of predicting are very simple: move the slider to best match the likelihood of the outcome, and click predict. You can predict as often as you want, and you're encouraged to change your mind when new information becomes available.

The displayed score is split into current points and total points. Current points show how much your prediction is worth now, whereas total points show the combined worth of all of your predictions over the lifetime of the question. The scoring details are available on the FAQ.

Thanks for predicting!

Your prediction has been recorded anonymously.

Want to track your predictions, earn points, and hone your forecasting skills? Create an account today!

Track your predictions
Continue exploring the site