The Map of Global Warming Prevention

TL;DR: Small prob­a­bil­ity of run­away global warm­ing re­quires prepa­ra­tion of ur­gent un­con­ven­tional mea­sures of its pre­ven­tion that is sun­light dim­ming.


The most ex­pected ver­sion of limited global warm­ing of sev­eral de­grees C in 21 cen­tury will not re­sult in hu­man ex­tinc­tion, as even the thaw­ing af­ter Ice Age in the past didn’t have such an im­pact.

The main ques­tion of global warm­ing is the pos­si­bil­ity of run­away global warm­ing and the con­di­tions in which it could hap­pen. Ru­n­away warm­ing means warm­ing of 30 C or more, which will make the Earth un­in­hab­it­able. It is un­likely event but it could re­sult in hu­man ex­tinc­tion.

Global warm­ing could also cre­ate some con­text risks, which will change the prob­a­bil­ity of other global risks.

I will not go here in all de­tails about na­ture of global warm­ing and es­tab­lished ideas about its pre­ven­tion as it has ex­ten­sive cov­er­age in Wikipe­dia (https://​​en.wikipe­​​wiki/​​Global_warm­ing and https://​​en.wikipe­​​wiki/​​Cli­mate_change_miti­ga­tion).

In­stead I will con­cen­trate on heavy tails risks and less con­ven­tional meth­ods of global warm­ing pre­ven­tion.

The map pro­vides sum­mary of all known meth­ods of GW pre­ven­tion and also of ideas about scale of GW and con­se­quences of each level of warm­ing.

The map also shows how pre­ven­tion plans de­pends of cur­rent level of tech­nolo­gies. In short, the map has three vari­ables: level of tech, level of ur­gency in GW pre­ven­tion and scale of the warm­ing.

The fol­low­ing post con­sists of text wall and the map, which are com­pli­men­tary: the text pro­vides in depths de­tails about some ideas and the map gives gen­eral overview of the pre­ven­tion plans.

The map: http://​​im­mor­tal­​​warm­ing3.pdf


The main fea­ture of cli­mate the­ory is its in­trin­sic un­cer­tainty. This un­cer­tainty is not about cli­mate change de­nial; we are al­most sure that an­thro­pogenic cli­mate change is real. The un­cer­tainty is about its ex­act scale and timing, and es­pe­cially about low prob­a­bil­ity tails with high con­se­quences. In the case of risk anal­y­sis we can’t ig­nore these tails as they bear the most risk. So I will fo­cus mainly on the tails, but this in turn re­quires a fo­cus on more marginal, con­tested or un­proved the­o­ries.

Th­ese un­cer­tain­ties are es­pe­cially large if we make pro­jec­tions for 50-100 years from now; they are con­nected with the com­plex­ity of the cli­mate, the un­pre­dictabil­ity of fu­ture emis­sions and the chaotic na­ture of the cli­mate.

Clathrate methane gun

An un­con­ven­tional but pos­si­ble global catas­tro­phe ac­cepted by sev­eral re­searchers is a green­house catas­tro­phe named the “run­away green­house effect”. The idea is well cov­ered in wikipe­dia https://​​en.wikipe­​​wiki/​​Clathrate_gun_hypothesis

Cur­rently large amounts of methane clathrate are pre­sent in the Arc­tic and since this area is warm­ing quickly than other re­gions, the gasses could be re­leased into the at­mo­sphere. https://​​en.wikipe­​​wiki/​​Arc­tic_methane_emissions

Pre­dic­tions re­lat­ing to the speed and con­se­quences of this pro­cess differ. Main­stream sci­ence sees the methane cy­cle as dan­ger­ous but slow pro­cess which could re­sult even­tu­ally in a 6 C rise in global tem­per­a­ture, which seems bad but it is sur­viv­able. It will also take thou­sands of years.

It has hap­pened once be­fore dur­ing Late-Pa­le­ocene, known as the Pa­le­ocene-Eocene ther­mal max­i­mum, https://​​en.wikipe­​​wiki/​​Pa­le­ocene%E2%80%93Eocene_Ther­mal_Max­i­mum (PETM), when the tem­per­a­ture jumped by about 6 C, prob­a­bly be­cause of methane. Methane-driven global warm­ing is just 1 of 10 hy­pothe­ses ex­plain­ing PETM. But dur­ing PETM global methane clathrate de­posits were around 10 times smaller than they are at pre­sent be­cause the ocean was warmer. This means that if the clathrate gun fires again it could re­sult in much more se­vere con­se­quences.

But some sci­en­tists think that it may hap­pen quickly and with stronger effects, which would re­sult in run­away global warm­ing, be­cause of sev­eral pos­i­tive feed­back loops. See, for ex­am­ple the blog http://​​arc­​​

There are sev­eral pos­si­ble pos­i­tive feed­back loops which could make methane-driven warm­ing stronger:

1) The Sun is now brighter than be­fore be­cause of stel­lar evolu­tion. The in­crease in the Sun’s lu­minos­ity will even­tu­ally re­sult in run­away global warm­ing in a pe­riod 100 mil­lion to 1 billion years from now. The Sun will be­come thou­sand of times more lu­mi­nous when it be­comes a red gi­ant. See more here: https://​​en.wikipe­​​wiki/​​Fu­ture_of_the_Earth#Loss_of_oceans

2) After a long pe­riod of a cold cli­mate (ice ages), a large amount of methane clathrate ac­cu­mu­lated in the Arc­tic.

3) Methane is short liv­ing at­mo­spheric gas (seven years). So the same amount of methane would re­sult in much more in­tense warm­ing if it is re­leased quickly, com­pared with a sce­nario in which it is scat­tered over cen­turies. The speed of methane re­lease de­pends on the speed global warm­ing. An­thro­pogenic CO2 in­creases very quickly and could be fol­lowed by a quick re­lease of the methane.

4) Water va­por is the strongest green house gas and more warm­ing re­sults in more wa­ter va­por in the at­mo­sphere.

5) Coal burn­ing re­sulted in large global dim­ming https://​​en.wikipe­​​wiki/​​Global_dim­ming And the cur­rent switch to cleaner tech­nolo­gies could stop the mask­ing of the global warm­ing.

6) The ocean’s abil­ity to solve CO2 falls with a rise in tem­per­a­ture.

7) The Arc­tic has the biggest tem­per­a­ture in­crease due to global warm­ing, with a pro­jected growth of 5-10 C, and as re­sult it will lose its ice shield and that would re­duce the Earth’s albedo which would re­sult in higher tem­per­a­tures. The same is true for per­mafrost and snow cover.

8) Warmer Sibe­rian rivers bring their wa­ter into the Arc­tic ocean.

9) The Gulfstream will bring warmer wa­ter from the Mex­i­can Gulf to the Arc­tic ocean.

10) The cur­rent pe­riod of a calm, spotless Sun would end and re­sult in fur­ther warm­ing.

An­thropic bias

One un­con­ven­tional rea­son for global warm­ing to be more dan­ger­ous than we used to think is an­thropic bias.

1. We tend to think that we are safe be­cause not run­away global warm­ing events have ever hap­pened in the past. But we could ob­serve only a planet where this never hap­pened. Milan Cirn­cov­ich and Bostrom wrote about it. So the real rate of run­away warm­ing could be much higher. See here: http://​​www.nick­​​pa­pers/​​an­throp­ic­shadow.pdf

2. Also we, hu­mans tend to find our­selves in a pe­riod when cli­mate changes are very strong be­cause of cli­mate in­sta­bil­ity. This is be­cause hu­man in­tel­li­gence as a uni­ver­sal adap­ta­tion mechanism was more effec­tive in the pe­riod of in­sta­bil­ity. So cli­mate in­sta­bil­ity helps to breed in­tel­li­gent be­ings. (This is my idea and may need ad­di­tional proof).

3. But if run­away global warm­ing is long over­due this would mean that our en­vi­ron­ment is more sen­si­tive even to smaller hu­man ac­tions (com­pare it with an over-pres­sured bal­loon and small nee­dle). In this case the amount of CO2 we cur­rently re­lease could be such an ac­tion. So we could un­der­es­ti­mate the frag­ility of our en­vi­ron­ment be­cause of an­thropic bias. (This is my idea and I wrote about here: http://​​www.slide­​​av­turchin/​​why-an­thropic-prin­ci­ple-stopped-to-defend-us-ob­ser­va­tion-se­lec­tion-and-frag­ility-of-our-en­vi­ron­ment)

The timeline of pos­si­ble run­away global warming

We could name the run­away global warm­ing a Venu­sian sce­nario be­cause thanks to a green­house effect on the sur­face of Venus its tem­per­a­ture is over 400 C, de­spite that, ow­ing to a high albedo (0.75, caused by white clouds) it re­ceives less so­lar en­ergy than the Earth (albedo 0.3).

A green­house catas­tro­phe can con­sist of three stages:

1. Warm­ing of 1-2 de­grees due to an­thro­pogenic C02 in the at­mo­sphere, pas­sage of «a trig­ger point». We don’t where the tip­ping point is, we may have passed it already, con­versely we may be un­der­es­ti­mat­ing nat­u­ral self-reg­u­lat­ing mechanisms.

2. Warm­ing of 10-20 de­grees be­cause of methane from gas hy­drates and the Sibe­rian bogs as well as the re­lease of CO2 cur­rently dis­solved in the oceans. The speed of this self-am­plify­ing pro­cess is limited by the ther­mal in­er­tia of the ocean, so it will prob­a­bly take about 10-100 years. This pro­cess can be ar­rested only by sharp hi-tech in­ter­ven­tions, like an ar­tifi­cial nu­clear win­ter and-or erup­tions of mul­ti­ple vol­ca­noes. But the more warm­ing oc­curs, the lesser the abil­ity of civ­i­liza­tion to stop it be­comes, as its tech­nolo­gies will be dam­aged. But the later that global warm­ing hap­pens, the higher the tech will be that can be used to stop it.

3.Moist green­house. Steam is a ma­jor con­trib­u­tor to a green­house effect, which re­sults in an even stronger and quicker pos­i­tive feed­back loop. A moist green­house will start if the av­er­age tem­per­a­ture of the earth is 47 C (cur­rently 15 C) and it will re­sult in a run­away evap­o­ra­tion of the oceans, re­sult­ing in 900 C sur­face tem­per­a­tures. (https://​​en.wikipe­​​wiki/​​Fu­ture_of_the_Earth#Loss_of_oceans ). All the wa­ter on the planet will boil, re­sult­ing in a dense wa­ter va­por at­mo­sphere. See also here: https://​​en.wikipe­​​wiki/​​Ru­n­away_green­house_effect


If we sur­vive un­til pos­i­tive Sin­gu­lar­ity, global warm­ing will be not an is­sue. But if strong AI and other su­per techs don’t ar­rive un­til the end of the 21st cen­tury, we wll need to in­vest a lot in its pre­ven­tion, as the civ­i­liza­tion could col­lapse be­fore the cre­ation of strong AI, which means that we will never be able to use all of its benefits.

I have a map, which sum­ma­rizes the known ideas for global warm­ing pre­ven­tion and adds some new ones for ur­gent risk man­age­ment. http://​​im­mor­tal­​​warm­ing2.pdf

The map has two main vari­ables: our level of tech progress and size of the warm­ing which we want to pre­vent. But its main vari­able is the abil­ity of hu­man­ity to unite and act proac­tively. In short, the plans are:

No plan – do noth­ing, and just adapt to warming

Plan A – cut­ting emis­sions and re­mov­ing green­house gases from the at­mo­sphere. Re­quires a lot of in­vest­ment and co­op­er­a­tion. Long term ac­tion and re­mote re­sults.

Plan B – geo-en­g­ineer­ing aimed at block­ing sun­light. Not much in­vest­ment and unilat­eral ac­tion are pos­si­ble. Quicker ac­tion and quicker re­sults, but in­volves risks in the case of switch­ing off.

Plan C – emer­gency ac­tions for Sun dim­ming, like ar­tifi­cial vol­canic win­ter.

Plan D – mov­ing to other planets.

All plans could be ex­e­cuted us­ing cur­rent tech lev­els and also at a high tech level through the use of nan­otech and so on.

I think that cli­mate change de­mands that we go di­rectly to plan B. Plan A is cut­ting emis­sions, and it’s not work­ing, be­cause it is very ex­pen­sive and re­quires co­op­er­a­tion from all sides. Even then it will not achieve im­me­di­ate re­sults and the tem­per­a­ture will still con­tinue to rise for many other rea­sons.

Plan B is chang­ing the opac­ity of the Earth’s at­mo­sphere. It could be a sur­pris­ingly low cost ex­er­cise and could be op­er­ated lo­cally made. There are sug­ges­tions to re­lease some­thing as sim­ple as sulfuric acid into the up­per at­mo­sphere to raise its re­flec­tion abil­ities.

“Ac­cord­ing to Keith’s calcu­la­tions, if op­er­a­tions were be­gun in 2020, it would take 25,000 met­ric tons of sulfuric acid to cut global warm­ing in half af­ter one year. Once un­der way, the in­jec­tion of sulfuric acid would pro­ceed con­tin­u­ously. By 2040, 11 or so jets de­liv­er­ing roughly 250,000 met­ric tons of it each year, at an an­nual cost of $700 mil­lion, would be re­quired to com­pen­sate for the in­creased warm­ing caused by ris­ing lev­els of car­bon diox­ide. By 2070, he es­ti­mates, the pro­gram would need to be in­ject­ing a bit more than a mil­lion tons per year us­ing a fleet of a hun­dred air­craft.” https://​​­nol­o­gyre­​​s/​​511016/​​a-cheap-and-easy-plan-to-stop-global-warm­ing/​​

There are also ideas to re­cap­ture CO2 us­ing ge­net­i­cally mod­ified or­ganisms, iron seed­ing in the oceans and by dis­pers­ing the car­bon cap­tur­ing min­eral oliv­ine.

The prob­lem with that ap­proach is that it can’t be stopped. As Seth Baum wrote, a smaller catas­tro­phe could re­sult in the dis­rup­tion of such en­g­ineer­ing and the con­se­quent im­me­di­ate re­turn of global warm­ing with a vengeance. http://​​seth­​​ac/​​2013_Dou­bleCatas­tro­phe.html

There are other ways pf pre­vent­ing global warm­ing. Plan C is cre­at­ing an ar­tifi­cial nu­clear win­ter through a vol­canic ex­plo­sion or by start­ing large scale for­est fires with nukes. This idea is even more con­tro­ver­sial and untested than geo-en­g­ineer­ing.

A re­gional nu­clear war I ca­pa­ble of putting 5 mln tons of black car­bon into the up­per ath­mo­sphere, “av­er­age global tem­per­a­tures would drop by 2.25 de­grees F (1.25 de­grees C) for two to three years af­ter­ward, the mod­els sug­gest.”

http://​​­tion­al­geo­​​news/​​2011/​​02/​​110223-nu­clear-war-win­ter-global-warm­ing-en­vi­ron­ment-sci­ence-cli­mate-change/​​ Nu­clear ex­plo­sions in deep forests may have the same effect as at­tacks on cities in term of soot pro­duc­tion.

Fight­ing be­tween Plan A and Plan B

So we are not even close to be­ing doomed by global warm­ing but we may have to change the way we re­act to it.

While cut­ting emis­sions is im­por­tant it will prob­a­bly not work within a 10-20 year pe­riod, quicker act­ing mea­sures should be de­vised.

The main risk is abrupt run­away global warm­ing. It is low prob­a­bil­ity event with the high­est con­se­quences. To fight it we should pre­pare rapid re­sponse mea­sures.

Such prepa­ra­tion should be done in ad­vance, which re­quires ex­pen­sive sci­en­tific ex­per­i­ments. The main prob­lem here is (as always) fund­ing, and reg­u­la­tors’ ap­proval. The im­pact of sulfur aerosols should be tested. Com­pli­cated math mod­els should be eval­u­ated.

Con­tra-ar­gu­ments are the fol­low­ing: “Openly em­brac­ing cli­mate en­g­ineer­ing would prob­a­bly also cause emis­sions to soar, as peo­ple would think that there’s no need to even try to lower emis­sions any more. So, if for some rea­son the de­liv­ery of that sulfuric acid into the at­mo­sphere or what­ever was dis­rupted, we’d be in trou­ble. And do we know enough of such mea­sures to say that they are safe? Of course, if we be­lieve that his­tory will end any­ways within decades or cen­turies be­cause of sin­gu­lar­ity, long-term effects of such mea­sures may not mat­ter so much… Another big is­sue with chang­ing in­so­la­tion is that it doesn’t solve ocean acid­ifi­ca­tion. No state ac­tor should be al­lowed to start geo-en­g­ineer­ing un­til they at least take sim­ple mea­sures to re­duce their emis­sions.” (com­ments from Less­wrong dis­cus­sion about GW).

Cur­rently it all looks like a poli­ti­cal fight be­tween Plan A (cut­ting emis­sions) and Plan B (geo-en­g­ineer­ing), where plan A’s ap­proval is win­ning. It has been sug­gested not to im­ple­ment Plan B as an in­crease in the warm­ing would demon­strate a real need to im­ple­ment Plan A (cut­ting emis­sions). Reg­u­la­tors didn’t ap­prove even the small­est ex­per­i­ments with sulfur shield­ing in Bri­tain. Iron ocean seed­ing also has reg­u­la­tory prob­lems.

But the same logic works in the op­po­site di­rec­tion. China and the coal com­pa­nies will not cut emis­sions, be­cause they want to press poli­cy­mak­ers to im­ple­ment plan B. It looks like a pris­oner’s dilemma of two plans.

The differ­ence be­tween the two plans is that plan A will re­turn ev­ery­thing to its nat­u­ral state and plan B is aimed on cre­at­ing in­stru­ments to reg­u­late the planet’s cli­mate and weather.

In the cur­rent global poli­ti­cal situ­a­tion, cut­ting emis­sions is difficult to im­ple­ment be­cause it re­quires col­lab­o­ra­tion be­tween many ri­val com­pa­nies and coun­tries. If sev­eral of them defect (most likely China, Rus­sia and In­dia, who have heavy use of coal and other fos­sil fuels), it will not work, even if all of Europe were so­lar pow­ered.

Tran­si­tion to zero-emis­sion econ­omy could hap­pen nat­u­rally in 20 years af­ter elec­tric trans­porta­tion will be­come wide­spread as well as so­lar en­ergy.

Plan C should be im­ple­mented if the situ­a­tion sud­denly changes for the worse, with the tem­per­a­ture jump­ing 3-5 C in one year. In this case the only op­tion we have is to bomb Pi­natubo vol­cano to make it erupt again, or prob­a­bly even sev­eral vol­canos. A vol­canic win­ter will give us time to adopt other geo-en­g­ineer­ing mea­sures.

I would also ad­vo­cate for a mix­ture of both plans, be­cause they work on differ­ent timescale. Cut­ting emis­sions and re­mov­ing CO2 us­ing the cur­rent level of tech­nolo­gies would take decades to have an im­pact on the cli­mate. But geo-en­g­ineer­ing has a re­ac­tion time of around one year so we could use it to cover the bumps in the road.

Espe­cially im­por­tant is the fact that if we com­pletely stop emis­sions, we could also stop global dim­ming from coal burn­ing which would re­sult in a 3 C global tem­per­a­ture jump. So stop­ping emis­sions may re­sult in a tem­per­a­ture jump, and we need a pro­tec­tion sys­tem in this case.

In all cases we need to sur­vive un­til stronger tech­nolo­gies de­velop. Us­ing nan­otech or ge­netic en­g­ineer­ing we could solve the warm­ing prob­lem with less effort. But we have to sur­vive un­til this time.

It seems to me that the idea of cut­ting emis­sions is over­hyped and so­lar man­age­ment is “un­der­hyped” in terms of pub­lic opinion and fund­ing. By chang­ing that mis­bal­ance we could achieve more com­mon good.

An un­pre­dictable cli­mate needs a quicker reg­u­la­tion system

The man­age­ment of cli­mate risks de­pends on their pre­dictabil­ity and it seems that this is not very high. The cli­mate is a very com­plex and chaotic sys­tem.

It may re­act un­ex­pect­edly in re­sponse to our own ac­tions. This means that long-term ac­tions are less fa­vor­able. The situ­a­tion could change many times dur­ing their im­ple­men­ta­tion.

The quick ac­tions like so­lar shield­ing are bet­ter for man­age­ment of poor pre­dictable pro­cesses, as we can see the re­sults of our ac­tions and quickly can­cel them or make them stronger if we don’t like the re­sults.

Con­text risks – in­fluenc­ing the prob­a­bil­ity of other global risks

Global warm­ing has some con­text risks: it could slow tech progress, it could raise the chances of war (prob­a­bly already hap­pened in Syria be­cause of draught http://​​fu­ture­​​2016/​​07/​​22/​​cli­mate-change-is-the-most-ur­gent-ex­is­ten­tial-risk/​​), it could ex­ac­er­bate con­flicts be­tween states about how to share re­courses (food, wa­ter etc.) and about the re­spon­si­bil­ity for risk miti­ga­tion. All such con­text risks could lead to a larger global catas­tro­phe.

Another con­text risk is that global warm­ing is cap­tures al­most all the available pub­lic at­ten­tion for global risks miti­ga­tion, and other more ur­gent risks may get less at­ten­tion.

Many peo­ple think that run­away global warm­ing con­sti­tutes the main risk of global catas­tro­phe. Another group think it is AI, and there is no di­a­log be­tween these two groups.

The level of warm­ing which is sur­viv­able strongly de­pends of our tech level. Some com­bi­na­tions of tem­per­a­ture and mois­ture are non-sur­viv­able for hu­man be­ings with­out air con­di­tion­ing: If the tem­per­a­ture rises by 15 C, half of the pop­u­la­tion will be in a non-sur­viv­able environment

http://​​www.sci­​​re­leases/​​2010/​​05/​​100504155413.htm be­cause very hu­mid and hot air pre­vents cool­ing by per­spira­tion and feels like a much higher tem­per­a­ture. With the cur­rent level of tech we could fight it, but if hu­man­ity falls to a me­dieval level, it would be much more difficult to re­cover in such con­di­tions.

Ris­ing CO2 lev­els could also im­pair hu­man in­tel­li­gence and slow tech progress as CO2 lev­els near 1000 ppm are known to have nega­tive effects on cog­ni­tion.

Warm­ing may also re­sult in large hur­ri­canes. They can ap­pear if the sea tem­per­a­ture reaches 50 C and they have a wind speed of 800 km/​h, which is enough to de­stroy any known hu­man struc­ture. They will be also very sta­ble and live very long, thus in­fluenc­ing the at­mo­sphere and cre­at­ing strong winds all over the world. The high­est sea tem­per­a­ture cur­rently is around 30C.


In fact we should com­pare not the mag­ni­tude but speed of global warm­ing with the speed of tech progress. If the warm­ing is quicker it wins. If we have very slow warm­ing, but even slower progress, the warm­ing still wins. In gen­eral I think that progress will over­run warm­ing, and we will cre­ate strong AI be­fore we have to deal with se­ri­ous global warm­ing con­se­quences.

Differ­ent predictions

Mul­ti­ple peo­ple pre­dict ex­tinc­tion due to global warm­ing but they are mostly la­beled as “alarmists” and are ig­nored. Some no­table pre­dic­tions:

1. David Auer­bach pre­dicts that in 2100 warm­ing will be 5 C and com­bined with re­source de­ple­tion and over­crowd­ing it will re­sult in global catas­tro­phe. http://​​www.dai­ly­​​sci­encetech/​​ar­ti­cle-3131160/​​Will-child-wit­ness-end-hu­man­ity-Mankind-ex­tinct-100-years-cli­mate-change-warns-ex­pert.html

2. Sam Carana pre­dicts that warm­ing will be 10 C in the 10 years fol­low­ing 2016, and ex­tinc­tion will hap­pen in 2030. http://​​arc­​​2016/​​03/​​ten-de­grees-warmer-in-a-decade.html

3. Con­ven­tional pre­dic­tions of the IPCC give a max­i­mum warm­ing of 6.4 C at 2100 in worst case emis­sion sce­nario and worst cli­mate sen­si­tivity to them: https://​​en.wikipe­​​wiki/​​Effects_of_global_warm­ing#SRES_emis­sions_scenarios

4. The con­sen­sus of sci­en­tists is that cli­mate tip­ping point will be in 2200 http://​​­de­pen­​​news/​​sci­ence/​​sci­en­tists-ex­pect-cli­mate-tip­ping-point-by-2200-2012967.html

5. If hu­man­ity con­tinues to burn all known car­bon sources it will re­sult in a 10 C warm­ing by 2030. https://​​­scien­​​ar­ti­cle/​​mg21228392-300-hy­per­warm­ing-cli­mate-could-turn-earths-poles-green/​​ The only sce­nario in which we are still burn­ing fos­sil fuels by 2300 (but not ex­tinct and not a so­lar pow­ered su­per­civilza­tion run­ning nan­otech and AI) is a se­ries of nu­clear wars or other smaller catas­tro­phes which will per­mit the ex­is­tence of re­gional pow­ers which of­ten smash each other into ru­ins and then re­build us­ing coal en­ergy. Some­thing like global nu­clear “So­mali world”.

We should give more weight to less main­stream pre­dic­tions, be­cause they de­scribe heavy tails of pos­si­ble out­comes. I think that it will be rea­son­able to es­ti­mate the risks of ex­tinc­tion level run­away global warm­ing in the next 100-300 years at 1 per cent and act as it is the main risk from global warm­ing.