16 Recent Publications on Existential Risk (Nov & Dec 2019 update)

Link post for: https://​​www.cser.ac.uk/​​news/​​re­cent-pub­li­ca­tions-de­cem­ber-2019/​​ and https://​​www.cser.ac.uk/​​news/​​8-re­cent-pub­li­ca­tions-ex­is­ten­tial-risk-novem­ber-20/​​

Each month, The Ex­is­ten­tial Risk Re­search Assess­ment (TERRA) uses a unique ma­chine-learn­ing model to pre­dict those pub­li­ca­tions most rele­vant to ex­is­ten­tial risk or global catas­trophic risk. The fol­low­ing are a se­lec­tion of those pa­pers iden­ti­fied for the last two months.

Please note that we provide these cita­tions and ab­stracts as a ser­vice to aid other re­searchers in pa­per dis­cov­ery and that in­clu­sion does not rep­re­sent any kind of en­dorse­ment of this re­search by the Cen­tre for the Study of Ex­is­ten­tial Risk or our re­searchers.

De­cem­ber 2019 update

1. Risk and Re­sponse to Biolog­i­cal Catas­tro­phe in Lower In­come Countries

Nat­u­ral and in­ten­tional biolog­i­cal risks threaten hu­man civ­i­liza­tion, both through di­rect hu­man fatal­ity as well as fol­low-on effects from a col­lapse of the just-in-time de­liv­ery sys­tem that pro­vides food, en­ergy and crit­i­cal sup­plies to com­mu­ni­ties globally. Hu­man be­ings have mul­ti­ple in­nate cog­ni­tive bi­ases that sys­tem­at­i­cally im­pair care­ful con­sid­er­a­tion of these risks. Res­i­dents of low-in­come coun­tries, es­pe­cially those who live in ru­ral ar­eas and are less de­pen­dent upon global trade, may be the most re­silient com­mu­ni­ties to catas­trophic risks, but low-in­come coun­tries also pre­sent a height­ened risk for biolog­i­cal catas­tro­phe. Hotspots for the emer­gence of new zoonotic dis­eases are pre­dom­i­nantly lo­cated in low-in­come coun­tries. Crowded, poorly sup­plied health­care fa­cil­ities in low-in­come coun­tries provide an op­ti­mal en­vi­ron­ment for new pathogens to trans­mit to a next host and adapt for more effi­cient per­son-to-per­son trans­mis­sion. Strate­gies to ad­dress these risks in­clude over­com­ing our nat­u­ral bi­ases and rec­og­niz­ing the im­por­tance of these risks, avoid­ing an over-re­li­ance on de­vel­op­ing spe­cific biolog­i­cal coun­ter­mea­sures, de­vel­op­ing gen­er­al­ized so­cial and be­hav­ioral re­sponses and in­vest­ing in re­silience.

2. The role of ex­perts in the pub­lic per­cep­tion of risk of ar­tifi­cial intelligence

The goal of this pa­per is to de­scribe the mechanism of the pub­lic per­cep­tion of risk of ar­tifi­cial in­tel­li­gence. For that we ap­ply the so­cial am­plifi­ca­tion of risk frame­work to the pub­lic per­cep­tion of ar­tifi­cial in­tel­li­gence us­ing data col­lected from Twit­ter from 2007 to 2018. We an­a­lyzed when and how there ap­peared a sig­nifi­cant rep­re­sen­ta­tion of the as­so­ci­a­tion be­tween risk and ar­tifi­cial in­tel­li­gence in the pub­lic aware­ness of ar­tifi­cial in­tel­li­gence. A sig­nifi­cant find­ing is that the image of the risk of AI is mostly as­so­ci­ated with ex­is­ten­tial risks that be­came pop­u­lar af­ter the fourth quar­ter of 2014. The source of that was the pub­lic po­si­tion­ing of ex­perts who hap­pen to be the real movers of the risk per­cep­tion of AI so far in­stead of ac­tual dis­asters. We an­a­lyze here how this kind of risk was am­plified, its sec­ondary effects, what are the va­ri­eties of risk un­re­lated to ex­is­ten­tial risk, and what is the dy­nam­ics of the ex­perts in ad­dress­ing their con­cerns to the au­di­ence of lay peo­ple.

3. Global Catas­trophic Threats from the Fun­gal King­dom: Fun­gal Catas­trophic Threats

The fun­gal king­dom poses ma­jor catas­trophic threats to hu­man­ity but these are of­ten un­ap­pre­ci­ated and min­i­mized, in biolog­i­cal threat as­sess­ments. The causes for this blind spot are com­plex and in­clude the re­mark­able nat­u­ral re­sis­tance of hu­mans to pathogenic fungi, the lack of con­ta­gious­ness of hu­man fun­gal dis­eases, and the in­di­rect­ness of fun­gal threats, which are more likely to me­di­ate their de­struc­tive effects on crops and ecosys­tems. A re­view of his­tor­i­cal events re­veals that the fun­gal king­dom in­cludes ma­jor threats to hu­man­ity through their effects on hu­man health, agri­cul­ture, and de­struc­tion of ma­teriel. A ma­jor con­cern go­ing for­ward is the like­li­hood that phys­iolog­i­cal adap­ta­tions by fun­gal species to global warm­ing will bring new fun­gal threats. Fun­gal threats pose sig­nifi­cant challenges spe­cific to this group of or­ganisms in­clud­ing the po­ten­tial for in­ter­con­ti­nen­tal spread by air cur­rents, ca­pac­ity for rapid evolu­tion, a paucity of effec­tive drugs, the ab­sence of vac­cines, and in­creas­ing drug re­sis­tance. Pre­pared­ness against bio-catas­trophic risks must in­clude con­sid­er­a­tion of the threats posed by fungi, which in turn re­quires a greater in­vest­ment in my­col­ogy-re­lated re­search.

4. The Pan­glos­sian poli­tics of the geoclique

So­lar ra­di­a­tion man­age­ment (SRM)–a form of geo­eng­ineer­ing–cre­ates a risk of ‘ter­mi­na­tion shock’. If SRM was to be stopped abruptly then tem­per­a­tures could rise very rapidly with catas­trophic im­pacts. Two promi­nent geo­eng­ineer­ing re­searchers have re­cently ar­gued that the risk of ter­mi­na­tion shock could be min­imised through the adop­tion of ‘rel­a­tively sim­ple’ poli­cies. This pa­per shows their ar­gu­ments to be premised on hero­ically op­ti­mistic as­sump­tions about the prospects for global co­op­er­a­tion and sus­tained trust in an SRM de­ploy­ment sce­nario. The pa­per ar­gues that worst-case sce­nar­ios are the right place to start in think­ing about the gov­er­nance of SRM.

5. Char­ac­ter­is­tics of Microbes Most Likely to Cause Pan­demics and Global Catastrophes

Pre­dict­ing which pathogen will con­fer the high­est global catas­trophic biolog­i­cal risk (GCBR) of a pan­demic is a difficult task. Many ap­proaches are ret­ro­spec­tive and premised on prior pan­demics; how­ever, such an ap­proach may fail to ap­pre­ci­ate novel threats that do not have ex­act his­tor­i­cal prece­dent. In this pa­per, based on a study and pro­ject we un­der­took, a new paradigm for pan­demic pre­pared­ness is pre­sented. This paradigm seeks to root pan­demic risk in ac­tual at­tributes pos­sessed by spe­cific classes of micro­bial or­ganisms and leads to spe­cific recom­men­da­tions to aug­ment pre­pared­ness ac­tivi­ties.

6. De­moc­ra­tiz­ing cog­ni­tive tech­nol­ogy: a proac­tive approach

Cog­ni­tive tech­nol­ogy is an um­brella term some­times used to des­ig­nate the realm of tech­nolo­gies that as­sist, aug­ment or simu­late cog­ni­tive pro­cesses or that can be used for the achieve­ment of cog­ni­tive aims. This tech­nolog­i­cal macro-do­main en­com­passes both de­vices that di­rectly in­ter­face the hu­man brain as well as ex­ter­nal sys­tems that use ar­tifi­cial in­tel­li­gence to simu­late or as­sist (as­pects of) hu­man cog­ni­tion. As they hold the promise of as­sist­ing and aug­ment­ing hu­man cog­ni­tive ca­pa­bil­ities both in­di­vi­d­u­ally and col­lec­tively, cog­ni­tive tech­nolo­gies could pro­duce, in the next decades, a sig­nifi­cant effect on hu­man cul­tural evolu­tion. At the same time, due to their dual-use po­ten­tial, they are vuln­er­a­ble to be­ing coopted by State and non-State ac­tors for non-be­nign pur­poses (e.g. cy­bert­er­ror­ism, cy­ber­war­fare and mass surveillance) or in man­ners that vi­o­late demo­cratic val­ues and prin­ci­ples. There­fore, it is the re­spon­si­bil­ity of tech­nol­ogy gov­er­nance bod­ies to al­ign the fu­ture of cog­ni­tive tech­nol­ogy with demo­cratic prin­ci­ples such as in­di­vi­d­ual free­dom, avoidance of cen­tral­ized, equal­ity of op­por­tu­nity and open de­vel­op­ment. This pa­per pro­vides a pre­limi­nary de­scrip­tion of an ap­proach to the de­moc­ra­ti­za­tion of cog­ni­tive tech­nolo­gies based on six nor­ma­tive eth­i­cal prin­ci­ples: avoidance of cen­tral­ized con­trol, open­ness, trans­parency, in­clu­sive­ness, user-cen­tered­ness and con­ver­gence. This ap­proach is de­signed to uni­ver­sal­ize and evenly dis­tribute the po­ten­tial benefits of cog­ni­tive tech­nol­ogy and miti­gate the risk that such emerg­ing tech­nolog­i­cal trend could be coopted by State or non-State ac­tors in ways that are in­con­sis­tent with the prin­ci­ples of liberal democ­racy or detri­men­tal to in­di­vi­d­u­als and groups.

7. You Can’t Handicraft the Apoca­lypse: The In­vidious Con­se­quences of “Opt­ing Out”

For sub­jects of ne­oliberal au­thor­i­tar­i­anism, the pre­car­i­ous­ness of ev­ery­day life is am­plified in the face of catas­trophic cli­mate change. Rather than build net­works of soli­dar­ity to shape a new world, au­thor­i­tar­ian ne­oliber­al­ism en­courages an­ti­so­cial in­di­vi­d­u­al­is­tic schemes to weather the storm by val­oriz­ing in­di­vi­d­u­als who can pre­pare them­selves for the worst. This es­say ex­tends Thorstein Ve­blen’s cri­tique of the Handicraft Move­ment of the early twen­tieth cen­tury to ex­plain the ap­peal of prep­ping, as well as its in­ad­e­quacy in the face of catas­tro­phe. Ve­blen shows how the Handicraft Move­ment was merely an­other way to con­spicu­ously con­sume. This es­say echoes that cri­tique and re­casts prep­ping as hand­icraft­ing the apoc­a­lypse, con­spicu­ously con­sum­ing even at the end of the world. It shows the in­ad­e­quacy in the face of an ex­is­ten­tial threat and con­cludes with a di­alogue be­tween Ve­blen and Bog­danov to the­o­rize con­sciously di­rect­ing in­dus­trial pro­duc­tion to­ward demo­cratic ends.

8. Does Biotech­nol­ogy Pose New Catas­trophic Risks?

Ad­vances in biotech­nol­ogy in the twenty-first cen­tury, fueled in large part by the field of syn­thetic biol­ogy, have greatly ac­cel­er­ated ca­pa­bil­ities to ma­nipu­late and re-pro­gram bac­te­ria, viruses, and other or­ganisms. Th­ese ge­netic en­g­ineer­ing ca­pa­bil­ities are driv­ing in­no­va­tion and progress in drug man­u­fac­tur­ing, biore­me­di­a­tion, and tis­sue en­g­ineer­ing, as well as biose­cu­rity pre­pared­ness. How­ever, biotech­nol­ogy is largely dual use, hold­ing the po­ten­tial of mi­suse for de­liber­ate harm along with pos­i­tive ap­pli­ca­tions; defenses against those threats need to be an­ti­ci­pated and pre­pared. This chap­ter de­scribes the challenges of man­ag­ing dual-use ca­pa­bil­ities en­abled by mod­ern biotech­nol­ogy and syn­thetic biol­ogy and high­lights a frame­work tool de­vel­oped by a Na­tional Academies com­mit­tee to aid anal­y­sis of the se­cu­rity effects of new sci­en­tific dis­cov­er­ies and pri­ori­ti­za­tion of con­cerns. The pos­i­tive as­pects of syn­thetic biol­ogy in pre­pared­ness are also de­tailed, and policy di­rec­tions are high­lighted for tak­ing ad­van­tage of the pos­i­tive as­pects of these emerg­ing tech­nolo­gies while min­i­miz­ing risks.

Novem­ber 2019 update

1. The apoc­a­lypse: It’s not the end of the world

Hu­man­ity is fac­ing mul­ti­ple pos­si­ble apoc­a­lypses, with nar­ra­tives that of­ten miss an im­por­tant point: The apoc­a­lypse prob­a­bly won’t be quick or fi­nal. It will be an en­vi­ron­ment, not an event or an end point for hu­man­ity. The apoc­a­lypse is more likely to bring mis­ery than cathar­sis or sal­va­tion. Although worst-case sce­nar­ios the­o­ret­i­cally make it eas­ier to pre­vent dire out­comes, in the case of slow-mov­ing apoc­a­lypses such as cli­mate change, it’s difficult for hu­mans to en­vi­sion the scale of the prob­lem and to imag­ine how we will ac­tu­ally ex­pe­rience it.

2. Re­vis­it­ing the cli­mate col­lapse: The view from Nuuk in the year 2070

Plane­tary warm­ing is one of sev­eral ex­is­ten­tial threats to hu­man civ­i­liza­tion. We are now in the cli­mate end-game, fac­ing a choice be­tween dra­matic ac­tion or a world plunged into out­right chaos. The con­se­quences of a failure to re­spond ap­pro­pri­ately to the risks are ex­plored in a sce­nario that illus­trates the im­pacts of poorly-miti­gated fos­sil fuel use over the next 50 years, in­clud­ing mas­sive dis­rup­tion of hu­man so­cieties, and iden­ti­fies the main causes of the epochal failure of gov­ern­ments to pro­tect the peo­ple and their fu­ture.

3. The ex­is­ten­tial threat of an­timicro­bial re­sis­tance

This ar­ti­cle pre­sents a sce­nario por­tray­ing the eco­nomic and hu­man costs that an­timicro­bial re­sis­tance could im­pose on so­ciety 30 years from now, if it is not ad­dressed soon.

4. How an In­dia-Pak­istan nu­clear war could start—and have global consequences

This ar­ti­cle de­scribes how an In­dia-Pak­istan nu­clear war might come to pass, and what the lo­cal and global effects of such a war might be. The di­rect effects of this nu­clear ex­change would be hor­rible; the au­thors es­ti­mate that 50 to 125 mil­lion peo­ple would die, de­pend­ing on whether the weapons used had yields of 15, 50, or 100 kilo­tons. The ram­ifi­ca­tions for In­dian and Pak­istani so­ciety would be ma­jor and long last­ing, with many ma­jor cities largely de­stroyed and un­in­hab­it­able, mil­lions of in­jured peo­ple need­ing care, and power, trans­porta­tion, and fi­nan­cial in­fras­truc­ture in ru­ins. But the cli­matic effects of the smoke pro­duced by an In­dia-Pak­istan nu­clear war would not be con­fined to the sub­con­ti­nent, or even to Asia. Those effects would be enor­mous and global in scope.

5. The haz­ard from frag­ment­ing comets

Comet dis­in­te­gra­tion pro­ceeds through both sub­li­ma­tion and dis­crete split­ting events. The cross-sec­tional area of ma­te­rial ejected by a comet may, within days, be­come many times greater than that of the Earth, mak­ing en­coun­ters with such de­bris much more likely than col­li­sions with the nu­cleus it­self. The hi­er­ar­chic frag­men­ta­tion and sub­li­ma­tion of a large comet in a short-pe­riod or­bit may yield many hun­dreds of such short-lived clusters. We model this evolu­tion with a view to as­sess­ing the prob­a­bil­ity of an en­counter that might have sig­nifi­cant ter­res­trial effects, through at­mo­spheric dust­ing or mul­ti­ple im­pacts. Such an en­counter may have con­tributed to the large an­i­mal ex­tinc­tions and sud­den cli­matic cool­ing of 12 900 yr ago, and the near-si­mul­ta­neous col­lapse of civil­i­sa­tions around 2350 BC.

6. Change isn’t always good

An­thro­pogenic cli­mate change is one of the most ex­is­ten­tial threats hu­man­ity has faced. Its even­tual out­comes, if we do not in­ter­vene now, will be catas­trophic. Land­scape ar­chi­tects are in a unique po­si­tion to be­come some of the most in­fluen­tial voices in iden­ti­fy­ing solu­tions. There is no silver bul­let that will solve the crisis. Mul­ti­ple strate­gies re­quire a sys­tems-think­ing ap­proach on a va­ri­ety of fronts and at a di­ver­sity of scales. De­sign­ing a more re­silient fu­ture must span the spec­trum of ecolog­i­cal, eco­nomic, and com­mu­nity-based ap­proaches to tackle big top­ics like re­gain­ing the planet’s bio­di­ver­sity, re­think­ing the im­pacts of our cur­rent agri­cul­tural prac­tices, and en­gag­ing poli­ti­cal lead­ers and or­di­nary cit­i­zens to sup­port strate­gic in­vest­ments that will re­duce risk.

7. Ecolog­i­cal Gen­trifi­ca­tion in Re­sponse to Apoca­lyp­tic Nar­ra­tives of Cli­mate Change: The Pro­duc­tion of an Im­muno-poli­ti­cal Fantasy

Anx­ieties over the po­ten­tial im­pacts of cli­mate change, of­ten framed in apoc­a­lyp­tic lan­guage, are hav­ing a profound, but lit­tle stud­ied effect on the con­tem­po­rary Western ur­ban­scape. This ar­ti­cle ex­am­ines the ways in which cur­rent the­o­riza­tions of ‘ecolog­i­cal gen­trifi­ca­tion’ ex­press only half the pro­cess, de­scribing how green space is used for so­cial con­trol, but not how ecol­ogy is used as a jus­tifi­ca­tion regime for such pro­jects. As ur­ban­ites seek out hous­ing and liv­ing prac­tices that have a lower en­vi­ron­men­tal im­pact, ur­ban plan­ners have re­sponded by pro­vid­ing large-scale re­gen­er­a­tion of the ur­ban­scape. With the de­mand for this hous­ing in­creas­ing, ques­tions of in­equal­ity, dis­place­ment and dis­pos­ses­sion arise. I ask whether apoc­a­lyp­tic anx­iety is be­ing en­rol­led in the jus­tifi­ca­tion regimes of these pro­jects to make them hard to re­sist at the plan­ning and im­ple­men­ta­tion stages. The ar­ti­cle shows that, in cap­i­tal­iz­ing on col­lec­tive anx­iety sur­round­ing an apoc­a­lyp­tic fu­ture, these pro­jects de­poli­ti­cize sub­jects by us­ing the empty sig­nifier, ‘Sus­tain­abil­ity’, lead­ing them into an im­muno-poli­ti­cal re­la­tion­ship to the ur­ban­scape. This leaves sub­jects feel­ing pro­tected from both re­spon­si­bil­ity for, and the im­pacts of, cli­mate change. Ul­ti­mately, this has the con­se­quence of gen­trifi­ca­tion cou­pled with po­ten­tially wors­en­ing con­sump­tive prac­tices, re­bound effects and the de­poli­ti­ciza­tion of the en­vi­ron­men­tally con­scious ur­ban­ite.

8. The hu­man cost of an­thro­pogenic global warm­ing: Semi-quan­ti­ta­tive pre­dic­tion and the 1,000-tonne rule

Green­house-gas emis­sions are in­di­rectly caus­ing fu­ture deaths by mul­ti­ple mechanisms. For ex­am­ple, re­duced food and wa­ter sup­plies will ex­ac­er­bate hunger, dis­ease, vi­o­lence, and mi­gra­tion. How will an­thro­pogenic global warm­ing (AGW) af­fect global mor­tal­ity due to poverty around and be­yond 2100? Roughly, how much burned fos­sil car­bon cor­re­sponds to one fu­ture death? What are the psy­cholog­i­cal, med­i­cal, poli­ti­cal, and eco­nomic im­pli­ca­tions? Pre­dicted death tolls are cru­cial for policy for­mu­la­tion, but un­cer­tainty in­creases with tem­po­ral dis­tance from the pre­sent and es­ti­mates may be bi­ased. Order-of-mag­ni­tude es­ti­mates should re­fer to liter­a­ture from di­verse rele­vant dis­ci­plines. The car­bon bud­get for 2°C AGW (roughly 1012 tonnes car­bon) will in­di­rectly cause roughly 109 fu­ture pre­ma­ture deaths (10% of pro­jected max­i­mum global pop­u­la­tion), spread over one to two cen­turies. This ze­roth-or­der pre­dic­tion is rel­a­tive and in ad­di­tion to ex­ist­ing pre­ventable death rates. It lies be­tween likely best- and worst-case sce­nar­ios of roughly 3 × 108 and 3 × 109, cor­re­spond­ing to plus/​minus one stan­dard de­vi­a­tion on a log­a­r­ith­mic scale in a Gaus­sian prob­a­bil­ity dis­tri­bu­tion. It im­plies that one fu­ture pre­ma­ture death is caused ev­ery time roughly 1,000 (300–3,000) tonnes of car­bon are burned. There­fore, any fos­sil-fuel pro­ject that burns mil­lions of tons of car­bon is prob­a­bly in­di­rectly kil­ling thou­sands of fu­ture peo­ple. The pre­dic­tion may be con­sid­ered valid, ac­count­ing for mul­ti­ple in­di­rect links be­tween AGW and death rates in a top-down ap­proach, but un­re­li­able due to the un­cer­tainty of cli­mate change feed­back and in­ter­ac­tions be­tween phys­i­cal, biolog­i­cal, so­cial, and poli­ti­cal cli­mate im­pacts (e.g., ecolog­i­cal cas­cade effects and co-ex­tinc­tion). Given uni­ver­sal agree­ment on the value of hu­man lives, a death toll of this un­prece­dented mag­ni­tude must be avoided at all costs. As a clear poli­ti­cal mes­sage, the “1,000-tonne rule” can be used to defend hu­man rights, es­pe­cially in de­vel­op­ing coun­tries, and to clar­ify that cli­mate change is pri­mar­ily a hu­man rights is­sue.

No comments.