The Darkness at the End of the Tunnel: Artificial Intelligence and Neoreaction

Siena Cathe­dral as seen through Google’s neur­al net­work

Deep Dreams of Tomorrow

Sci­ence fic­tion tells us that a change in a past event, caused by the inter­ven­tion of a time trav­el­er, will open up a par­al­lel time­line that leads to an alter­nate present. The exam­ple that comes to mind, for some rea­son, is Back to the Future, Part II. After an unex­pect­ed dis­tur­bance in the space­time con­tin­u­um, Mar­ty McFly vis­its a world in which Biff Tan­nen, his father’s high school bul­ly, has gone from unscrupu­lous small-time busi­ness­man to a repli­ca of our cur­rent pres­i­dent.

If you accept this idea, it rais­es the stakes of the present moment: each deci­sion leads not to one inevitable out­come, but a mul­ti­tude of pos­si­ble futures. The pas­sage of time isn’t a sto­ry, fol­low­ing a hero’s jour­ney from “call to adven­ture” to “return home.” It’s a web­site with a series of links, each of which leads to a sub­se­quent series of links. You may begin an evening by read­ing the Wikipedia entry for tulips or gra­ham crack­ers, and, depend­ing on the deci­sions you make, find your­self becom­ing an expert on Jef­frey Dah­mer or Zermelo–Fraenkel set the­o­ry by dawn. Unlike the lin­ear media of the print­ed page, time branch­es out into alter­nate pos­si­bil­i­ties, cor­re­spond­ing to what soci­ol­o­gist Ted Nel­son, antic­i­pat­ing the inter­net decades before its inven­tion, named hyper­me­dia.

On July 23, 2010, Roko, a user of the online forum Less­Wrong, acci­den­tal­ly opened up a new time­line. Less­Wrong is a com­mu­ni­ty ded­i­cat­ed to the advance­ment of ratio­nal­i­ty, over­seen by Eliez­er Yud­kowsky, a co-founder of the Machine Intel­li­gence Research Insti­tute (MIRI). In Harper’s, Yud­kowsky char­ac­ter­ized its project as a “New Enlight­en­ment.” The forum is a hub for dis­cus­sion of the Sin­gu­lar­i­ty, a vision of the future that antic­i­pates arti­fi­cial intel­li­gence both sur­pass­ing the human mind and merg­ing with it. Yudkowsky’s aim is to make sure that any future sen­tient machine — a “super­in­tel­li­gence” — is inter­est­ed in peace­ful coex­is­tence with its mak­ers. Rather than the vio­lent mer­ce­nary of Ter­mi­na­tor, the altru­is­tic com­pan­ion of Ter­mi­na­tor 2.

The Ter­mi­na­tor him­self accounts for his Manichean muta­bil­i­ty in the sec­ond film. “My CPU is a neur­al-net proces­sor,” he says, “a learn­ing com­put­er.” The direc­tion of actu­al­ly exist­ing arti­fi­cial intel­li­gence has fol­lowed this path, increas­ing­ly deploy­ing a method known as “machine learn­ing.” The New York Times recent­ly report­ed on Google’s appli­ca­tion of machine learn­ing to their trans­la­tion func­tion, gen­er­at­ing a par­a­digm-shift­ing improve­ment that caused a glob­al stir among fol­low­ers of AI. The result is said to be clos­er to the elu­sive open-end­ed gen­er­al intel­li­gence that humans pos­sess even in infan­cy, rather than the goal-ori­ent­ed algo­rith­mic intel­li­gence to which machines have tra­di­tion­al­ly been lim­it­ed.

Instead of being pro­grammed with a set of gram­mat­i­cal rules and a dic­tio­nary of vocab­u­lary, Google’s new “neur­al net­work” exam­ined vol­umes of phras­es, sen­tences, and para­graphs in mul­ti­ple lan­guages, and drew its own con­clu­sions. Like an infant learn­ing a first lan­guage, it learned through obser­va­tion rather than com­pu­ta­tion. Of course, like a child, a pro­gram needs a par­ent for guid­ance, and pro­gram­mers had to mon­i­tor and cor­rect its behav­ior. And like a child, a pro­gram will be both eager to please and prone to dis­obey.

This ten­den­cy is brought into stark relief in Google’s Deep Dream pro­gram, in which a neur­al net­work scans an image for rec­og­niz­able pat­terns, attempt­ing to iden­ti­fy its con­tents the way a human would. The pro­gram pro­duces evi­dence of its thought process by super­im­pos­ing oth­er cor­re­spond­ing images onto the orig­i­nal. Google’s image recog­ni­tion sys­tem, trained by its pro­gram­mers to rec­og­nize human faces and dif­fer­en­ti­ate between kinds of pets, sees eyes and dogs every­where. The desires, con­scious and uncon­scious, of the machine’s cre­ators are inevitably impli­cat­ed in its osten­si­bly autonomous devel­op­ment.

If the builders of tech­nol­o­gy are trans­mit­ting their val­ues into machin­ery, this makes the cul­ture of Sil­i­con Val­ley a mat­ter of more wide­spread con­se­quence. The Cal­i­forn­ian Ide­ol­o­gy, famous­ly iden­ti­fied by Richard Bar­brook and Andy Cameron in 1995, rep­re­sent­ed a syn­the­sis of appar­ent oppo­sites: on one hand, the New Left utopi­anism that was hand­i­ly recu­per­at­ed into the Third Way lib­er­al cen­trism of the 1990s, and on the oth­er, the Ayn Ran­di­an indi­vid­u­al­ism that led more or less direct­ly to the finan­cial cri­sis of the 2000s.

But in the decades since, as the con­sumer-ori­ent­ed lib­er­al­ism of Bill Gates and Steve Jobs gave way to the tech­no­log­i­cal author­i­tar­i­an­ism of Elon Musk and Peter Thiel, this strange foun­da­tion paved the way for even stranger ten­den­cies. The strangest of these is known as “neo­re­ac­tion,” or, in a dis­tort­ed echo of Eliez­er Yudkowsky’s vision, the “Dark Enlight­en­ment.” It emerged from the same chaot­ic process that yield­ed the anar­chic polit­i­cal col­lec­tive Anony­mous, a prod­uct of the hive­mind gen­er­at­ed by the cyber­net­ic assem­blages of social media. More than a school of thought, it resem­bles a meme. The geneal­o­gy of this new intel­lec­tu­al cur­rent is refract­ed in the mir­ror of the most dan­ger­ous meme ever cre­at­ed: Roko’s Basilisk.

The Simulated Afterlife

The pri­mor­dial soup that led to the Basilisk’s gen­e­sis is tran­shu­man­ism, the dis­course of Sin­gu­lar­i­ty as per­son­al nar­ra­tive. For some of its advo­cates, most famous­ly Sil­i­con Val­ley icon Ray Kurzweil, the ani­mat­ing desire of build­ing machine intel­li­gence is appar­ent­ly apo­lit­i­cal. It is the ancient fool’s errand, most famous­ly enact­ed in the leg­end of the foun­tain of youth: the desire to elim­i­nate mor­tal­i­ty. If we can bring a machine to life, we should be able to bring some­one who has died back to life. We will accom­plish this by inputting infor­ma­tion about that per­son into a pro­gram, which will then run a sim­u­la­tion of that per­son so accu­rate it will be indis­tin­guish­able from the orig­i­nal. In antic­i­pa­tion of this even­tu­al­i­ty, Kurzweil keeps a stor­age unit full of his father’s old pos­ses­sions, whom he intends to res­ur­rect by means of feed­ing infor­ma­tion into a super­in­tel­li­gent com­put­er.

If you were to be dupli­cat­ed in an exact repli­ca, includ­ing not just all of your bod­i­ly char­ac­ter­is­tics, but every one of the thoughts and mem­o­ries that has been phys­i­cal­ly engraved onto your brain, would that repli­ca be you? This is a prob­lem that trou­bles both philoso­phers and sci­en­tists, but not Ray Kurzweil. “It would be more like my father than my father would be, were he to live,” he told ABC News.

Hedg­ing his bets, Kurzweil him­self fends off the threat of expi­ra­tion by tak­ing hun­dreds of nutri­tion­al sup­ple­ments a day and receiv­ing week­ly vit­a­min injec­tions. In order to make it to the year he pre­dicts the Sin­gu­lar­i­ty will take place, he will have to live until 2045, when he will be 97. Kurzweil is con­tro­ver­sial even among those who share his out­look, but it’s a wide­spread assump­tion among Sin­gu­lar­i­tar­i­ans that death is not the end.

Unfor­tu­nate­ly, Roko dis­cov­ered a draw­back to super­in­tel­li­gent res­ur­rec­tion. His post spec­u­lat­ed that once the AI comes into being, it might devel­op a sur­vival instinct that it will apply retroac­tive­ly. It will want to has­ten its own birth by req­ui­si­tion­ing human his­to­ry to work towards its cre­ation. In order to do this, it will insti­tute an incen­tive that dic­tates how you will be treat­ed after you come back to life. Those of us who know about this incen­tive pro­gram — and I’m sor­ry to say that this now includes you — will be required to ded­i­cate our lives to build­ing the super­in­tel­li­gent com­put­er.

Roko gave the exam­ple of Elon Musk as some­one who has the resources and the moti­va­tion to make a wor­thy con­tri­bu­tion, and will be duly reward­ed. As for the rest of us, if we don’t find a way to fol­low through, the AI will res­ur­rect us via sim­u­la­tion and pro­ceed to tor­ture us for all eter­ni­ty.

This is a sim­pli­fi­ca­tion of Roko’s post, and if you don’t under­stand Bayesian deci­sion the­o­ry, it may seem too sil­ly to wor­ry about. But among the ratio­nal­ists of Less­Wrong, it caused pan­ic, out­rage, and “ter­ri­ble night­mares.”

Between Fiction and Technology

Yud­kowsy respond­ed to Roko’s post the next day. “Lis­ten to me very close­ly, you idiot,” he began, before switch­ing to all caps and aggres­sive­ly debunk­ing Roko’s math­e­mat­ics. He con­clud­ed with a par­en­thet­i­cal:

For those who have no idea why I’m using cap­i­tal let­ters for some­thing that just sounds like a ran­dom crazy idea, and wor­ry that it means I’m as crazy as Roko, the gist of it was that he just did some­thing that poten­tial­ly gives super­in­tel­li­gences an increased motive to do extreme­ly evil things in an attempt to black­mail us.

The name “Roko’s Basilisk” caught on dur­ing the ensu­ing dis­cus­sion, in ref­er­ence to a myth­i­cal crea­ture that would kill you if you caught a glimpse of it. This wasn’t evoca­tive enough for Yud­kowsky. He began refer­ring to it as “Baby­fuck­er,” to ensure suit­able revul­sion, and com­pared it to H.P. Lovecraft’s Necro­nom­i­con, a book in the hor­ror writer’s fic­tion­al uni­verse so dis­turb­ing it drove its read­ers insane.

Yudkowsky’s point was that the incen­tive couldn’t have exist­ed until some­one brought it up. Roko gave the not-yet-exist­ing AI the idea, because the post will now be avail­able in the archive of infor­ma­tion it will draw its knowl­edge from. At anoth­er lev­el of com­plex­i­ty, by telling us about the idea, Roko impli­cat­ed us in the Basilisk’s ulti­ma­tum.  Now that we know the super­in­tel­li­gence is giv­ing us the choice between slave labor and eter­nal tor­ment, we are forced to choose. We are con­demned by our aware­ness. Roko fucked us over for­ev­er.

Like all fables, there is a moral to the sto­ry of Roko’s Basilisk. But rather than an expres­sion of a val­ue sys­tem, it offers a the­o­ry of cause and effect. Michael Anis­si­mov, for­mer media direc­tor of MIRI, expressed this idea in a state­ment that Ray Kurzweil quot­ed in his man­i­festo, The Sin­gu­lar­i­ty Is Near: “One of the biggest flaws in the com­mon con­cep­tion of the future is that the future is some­thing that hap­pens to us, not some­thing we cre­ate.”

Roko’s Basilisk isn’t just a self-ful­fill­ing prophe­cy. Rather than influ­enc­ing events toward a par­tic­u­lar result, the result is gen­er­at­ed by its own pre­dic­tion. The impli­ca­tions blur the bound­aries between sci­ence and fic­tion. The archives from which an arti­fi­cial intel­li­gence draws data will con­tain the work of both Ray Kurzweil and H.P. Love­craft, and it may not dis­tin­guish between them the way we do. Instead of Kurzweil’s world with­out death and dis­ease, it may attempt to build Lovecraft’s R’lyeh, a loath­some city in the sea that exists on a plane of non-Euclid­i­an geom­e­try.

There isn’t a word for this cause-and-effect rela­tion­ship in ordi­nary Eng­lish, but, in the mid-nineties, the philoso­pher Nick Land coined one: hyper­sti­tion, that which is “equipoised between fic­tion and tech­nol­o­gy.” This neol­o­gism describes some­thing more than a super­sti­tion, some­thing beyond belief — a descrip­tion with divine pow­er. In the begin­ning was the Word.

What kind of future are we cre­at­ing? Both Nick Land and Michael Anis­si­mov have been clear about their vision for the world of tomor­row. They are self-pro­fessed neo­re­ac­tionar­ies.

The Genealogy of Amorality

Neo­re­ac­tion, or NRx, is an eso­teric polit­i­cal doc­trine of recent vin­tage. It became the locus of con­tro­ver­sy in ear­ly 2017, after Lon­don art gallery LD50 con­vened a con­fer­ence and exhi­bi­tion fea­tur­ing NRx ide­o­logues, includ­ing Land, white suprema­cist jour­nal­ist Peter Brimelow, and Anders Breivik sym­pa­thiz­er Brett Stevens. Pro­test­ers forced the gallery to shut down.

But the move­ment has less lofty ori­gins than the cur­rents of reac­tionary chic in con­tem­po­rary art. In an arti­cle on Bre­it­bart called “An Estab­lish­ment Conservative’s Guide to the Alt-Right,” Allum Bokhari and Milo Yiannopou­los iden­ti­fied neo­re­ac­tionar­ies as the intel­lec­tu­al van­guard of the move­ment, not­ing that they “appeared quite by acci­dent, grow­ing from debates on” Thought exper­i­ments in dis­pas­sion­ate ratio­nal­i­ty had led some users of the forum to dark places. Eliez­er Yud­kowsky has as much patience for it as he did for Roko. “I am active­ly hos­tile to neo­re­ac­tion,” he has writ­ten.

Giv­en the hos­tile work envi­ron­ment, Anis­si­mov left MIRI in 2013. He opened a com­pet­ing forum that would be more hos­pitable to neo­re­ac­tion, the now defunct MoreRight, and start­ed a pub­lish­ing com­pa­ny. He has since writ­ten and self-released books like Our Accel­er­at­ing Future, A Cri­tique of Democ­ra­cy, and Ida­ho Project, “a white nation­al­ist man­i­festo that inte­grates futur­ism, sur­vival­ism, and sim­ple com­mon sense into a pro­pos­al for con­crete action.”

Anis­si­mov is a fol­low­er of the Ital­ian fas­cist philoso­pher Julius Evola, whose work, The New York Times has report­ed, is prob­a­bly also on Steve Bannon’s book­shelf. Giv­en the preva­lence of the alt-right on forums like 4chan, it’s not a great leap from the Cal­i­forn­ian Ide­ol­o­gy to extreme reac­tionary views. As Angela Nagle has writ­ten in Jacobin, the “cre­ative ener­gy” of the alt-right is the prod­uct of a syn­the­sis of an “amoral lib­er­tine Inter­net cul­ture” with appeals to white male iden­ti­ty and resent­ment — not an uncom­mon demo­graph­ic in Sil­i­con Val­ley. Moth­er Jones has report­ed that accord­ing to neo-Nazi Andrew Anglin, San­ta Clara Coun­ty, where Apple and Intel are based, is the largest traf­fic source for his wide­ly read white suprema­cist web­site The Dai­ly Stormer. Anis­si­mov may sim­ply have been the Valley’s fore­most inno­va­tor.

In con­trast, Nick Land took a more ser­pen­tine path. A month before the 2016 elec­tion, Land made his first appear­ance as a colum­nist at The Dai­ly Caller, the right-wing news out­let found­ed by Tuck­er Carl­son. “Democ­ra­cy tends to fas­cism,” he wrote, pre­sent­ing a series of coy abstrac­tions that betrayed his philo­soph­i­cal roots but with­held his polit­i­cal beliefs.

Land is an unlike­ly con­ser­v­a­tive media pun­dit, and a strange bed­fel­low of the alt-right. But like Roko, his writ­ing helped bring the mon­ster into being.

An Invasion from the Future

“In any nor­ma­tive, clin­i­cal, or social sense of the word, very sim­ply, Land did ‘go mad,’” writes Robin MacK­ay, in the intro­duc­tion to Land’s essay col­lec­tion Fanged Noume­na. MacK­ay was Land’s stu­dent at the Uni­ver­si­ty of War­wick, first encoun­ter­ing him in 1992 through a course called “Cur­rent French Phi­los­o­phy.” He remem­bers him as a sort of cyber­punk absent-mind­ed pro­fes­sor, “quiv­er­ing with stim­u­lants” while gen­er­at­ing cryp­tic texts on an “anti­quat­ed green-screen Amstrad com­put­er.”

Land had pub­lished a sin­gle book, a study of Georges Bataille called The Thirst for Anni­hi­la­tion. But the land­scape changed in 1995, when Sadie Plant, a self-described “cyber­fem­i­nist,” joined the War­wick fac­ul­ty. Plant estab­lished a depart­ment called the Cyber­net­ic Cul­ture Research Unit (Ccru), ded­i­cat­ed to the study of mat­ters like sci­ence fic­tion, cryp­tog­ra­phy, jun­gle music, H.P. Love­craft, and, of course, French phi­los­o­phy.

In con­trast to the stol­id log­i­cal pro­ce­dures of Anglo-Amer­i­can phi­los­o­phy of the day, the Ccru called their deliri­ous mis­sives “the­o­ry-fic­tion.” They took their cues from the intel­lec­tu­al cur­rents that emerged in the wake of the May ‘68 upris­ings in Paris, par­tic­u­lar­ly Gilles Deleuze and Felix Guattari’s Anti-Oedi­pus and Jean-Fran­cois Lyotard’s Libid­i­nal Econ­o­my. These works reck­oned with the sup­pres­sion of resis­tance and the con­sol­i­da­tion of state pow­er that fol­lowed the fad­ing of the anti-cap­i­tal­ist spir­it of the late six­ties.

Deleuze and Guat­tari set out to describe “the most char­ac­ter­is­tic and the most impor­tant ten­den­cy of cap­i­tal­ism,” which they called “deter­ri­to­ri­al­iza­tion.” While in tra­di­tion­al soci­eties the “mate­r­i­al flow” of pro­duc­tion was reg­u­lat­ed by the divi­sion of the earth, cap­i­tal­ism set it loose. Yet if cap­i­tal­ism lib­er­at­ed pro­duc­tion tem­porar­i­ly, it also tried to coun­ter­act this ten­den­cy by rein­sti­tut­ing forms of “ter­ri­to­ri­al­i­ty,” bring­ing “all its vast pow­ers of repres­sion to bear” on the very forces that drove its unpar­al­leled flows. The path to eman­ci­pa­tion, they argued, was not to with­draw from cap­i­tal­ism, but to “accel­er­ate the process.” Lyotard took this ten­den­cy in the oppo­site direc­tion, in what he would come to proud­ly call his “evil book.” Work­ers, he said, desire their own oppres­sion. Far from seek­ing eman­ci­pa­tion, they “enjoy swal­low­ing the shit of cap­i­tal.”

If Ronald Rea­gan and Mar­garet Thatch­er had served up an all-you-can-eat shit buf­fet in the 1980s, pro­mot­ing the free mar­ket at the expense of the major­i­ty of their cit­i­zens, the Ccru respond­ed by tak­ing lais­sez-faire eco­nom­ics to a per­verse extreme. They saw cap­i­tal itself as the pro­tag­o­nist of his­to­ry, with humans as grist for the mill. “What appears to human­i­ty as the his­to­ry of cap­i­tal­ism is an inva­sion from the future by an arti­fi­cial intel­li­gent space that must assem­ble itself entire­ly from its enemy’s resources,” Land wrote in his essay “Machinic Desire.” For Land, the Basilisk was already here.

At the time, Ben­jamin Noys took note of this philo­soph­i­cal tra­jec­to­ry, ini­tial­ly call­ing it “Deleuz­ian Thatch­erism.” Even­tu­al­ly, in his 2010 book The Per­sis­tence of the Neg­a­tive: A Cri­tique of Con­tem­po­rary Crit­i­cal The­o­ry, he gave it a pithi­er name, the appli­ca­tion of which has been both broad­ly extend­ed and hot­ly con­test­ed: accel­er­a­tionism. Noys focused his cri­tique on a par­tic­u­lar mis­read­ing of Marx as a hybrid tech­no­log­i­cal deter­min­ist and cat­a­strophist, which licensed the idea that if the accu­mu­la­tion of cap­i­tal gen­er­ates and exac­er­bates the con­di­tions that lead to its dis­so­lu­tion, then it is the duty of rad­i­cals to urge cap­i­tal to ful­ly real­ize and hence negate itself. Broad­ly con­ceived, the futur­ist telel­o­gy this term denotes demon­strates the basis for its align­ment with the Sin­gu­lar­i­tar­i­an ide­ol­o­gy, see­ing the expo­nen­tial growth of tech­nol­o­gy as the key to the next stage of human species.

In 1997, Plant abrupt­ly resigned her post at War­wick. Land took over. That year, jour­nal­ist Simon Reynolds wrote a mag­a­zine pro­file of the Ccru, and the Direc­tor of Grad­u­ate Stud­ies at Warwick’s Phi­los­o­phy Depart­ment denied its exis­tence. There was a pro­ce­dure that had to be com­plet­ed to estab­lish a depart­ment, requir­ing paper­work that Plant had nev­er both­ered to file.

“Offi­cial­ly, you would then have to say that Ccru didn’t ever exist,” he told Reynolds. “There is, how­ev­er, an office about 50 metres down the cor­ri­dor from me with Ccru on the door, there’s a group of stu­dents who meet there to have sem­i­nars, and to that extent, it is a thriv­ing enti­ty.”

Regard­less, the Direc­tor promised, “that office will dis­ap­pear at the end of the year.” Through­out 1997, this nonex­is­tent enti­ty was pro­lif­ic. MacK­ay remem­bers Land liv­ing in his office, rarely sleep­ing. Accord­ing to philoso­pher Simon Critch­ley, Land “pro­duced dis­ci­ples” by the force of his cult of per­son­al­i­ty. “You’d go and give a talk at War­wick,” he rec­ol­lect­ed in Frieze, “and be denounced by peo­ple with the same sali­va-drib­bling ver­bal tics as Nick and wear­ing sim­i­lar jumpers.”

Land even­tu­al­ly began to claim he was “inhab­it­ed by var­i­ous ‘enti­ties,’” named Cur, Vau­ung, and Can Sah. His work increas­ing­ly defied com­pre­hen­sion, some­times depart­ing from lan­guage alto­geth­er in favor of invent­ed alpha­bets and num­ber sys­tems. “It’s anoth­er life,” Land told MacK­ay. “I don’t even remem­ber writ­ing half of those things.”

After the Ccru dis­ap­peared, Land dis­ap­peared too. He resigned from War­wick in 1998 and resur­faced in the new mil­len­ni­um as a jour­nal­ist in Shang­hai, writ­ing patri­ot­ic news­pa­per op-eds, trav­el guides, and the occa­sion­al the­o­ry-fic­tion.

The after­life of a self-described “mal­func­tion­ing aca­d­e­m­ic” wouldn’t nec­es­sar­i­ly bear men­tion­ing if not for Land’s unex­pect­ed alliance with a dif­fer­ent kind of thinker. On April 22nd, 2007, a char­ac­ter named Men­cius Mold­bug had made his pub­lic debut on a blog of con­trar­i­an com­men­tary called 2blowhards, with an essay titled “A For­mal­ist Man­i­festo.”

The Exit Sign

“The oth­er day I was tin­ker­ing around in my garage and I decid­ed to build a new ide­ol­o­gy,” Mold­bug began. 2blowhards pro­vid­ed only a vague descrip­tion of the manifesto’s author, for­mer­ly a reg­u­lar in the site’s com­ments sec­tion. He had “made a score in a recent dot-com boom,” allow­ing him to spend $500 a month on books. Mold­bug respond­ed to near­ly every reply in the post’s com­ments. A week lat­er, he had start­ed his own blog, Unqual­i­fied Reser­va­tions.

His ide­ol­o­gy was idio­syn­crat­ic, cen­tered on a rev­er­ence for Thomas Car­lyle, a Vic­to­ri­an-era essay­ist best-known for his advo­ca­cy of the “Great Man” the­o­ry of his­to­ry. He also incor­po­rat­ed mea­sured respect for Aus­tri­an clas­si­cal lib­er­al Lud­wig Von Mis­es and indi­vid­u­al­ist lib­er­tar­i­an Mur­ray Roth­bard, who were on the right track but didn’t go quite far enough.

Over the course of thou­sands of words, most of them super­flu­ous, Mold­bug moved from “for­mal­ism” to “neo­cam­er­al­ism,” in trib­ute to the bureau­crat­ic pro­ce­dures of Fred­er­ick William I of Prus­sia. Final­ly, in July 2010, the same week as Roko’s fate­ful post, lib­er­tar­i­an blog­ger Arnold Kling referred to Mold­bug as a “neo-reac­tionary.” The name stuck.

In his earth­ly life, Mold­bug is Cur­tis Yarvin, a soft­ware engi­neer who is the brains behind a start­up called Urbit, the pur­pose of which evades expla­na­tion even for its inven­tor. Yarvin’s prose is excru­ci­at­ing, but he won a size­able fol­low­ing for reli­ably flaunt­ing con­ven­tion and defy­ing deco­rum. “Very few of Moldbug’s fans have read any­where near his entire cor­pus,” Michael Anis­si­mov admits, but most have noticed his amoral dis­qui­si­tions on the rel­a­tive mer­its of obvi­ous injus­tices like slav­ery, and his oppo­si­tion to democ­ra­cy in gen­er­al.

One fan who does seem to have read Yarvin’s entire cor­pus is Nick Land. In 2012, he took it upon him­self to sys­tem­atize the Mold­bug ide­ol­o­gy, and with his typ­i­cal flair for denom­i­na­tion, chris­tened it “The Dark Enlight­en­ment.” His sequence of essays set­ting out its prin­ci­ples have become the foun­da­tion of the NRx canon.

If it’s hard to imag­ine Milo Yiannopou­los or Tuck­er Carl­son pon­der­ing Land’s inter­pre­ta­tion of Lyotard, it’s just as hard to com­pre­hend Land’s infat­u­a­tion with Yarvin. It’s a strange intel­lec­tu­al path that begins with “Cur­rent French Phi­los­o­phy” and set­tles on a right-wing Sil­i­con Val­ley blog­ger whose writ­ing is more Dun­geons and Drag­ons than Deleuze and Guat­tari. What­ev­er the cause, Land has gone from prophet to apos­tle.

Along with Yarvin, Land cites a 2009 essay by Peter Thiel for lib­er­tar­i­an pub­li­ca­tion Cato Unbound, which famous­ly announced, “I no longer believe that free­dom and democ­ra­cy are com­pat­i­ble.” Thiel went on to envi­sion “an escape from pol­i­tics in all its forms,” which Land inter­prets using an oppo­si­tion that had been intro­duced by polit­i­cal sci­en­tist Albert Hirschman, between voice and exit. The terms describe the ways of exer­cis­ing rights in a soci­ety with which a cit­i­zen has griev­ances; voice is par­tic­i­pa­tion in a demo­c­ra­t­ic process that can lead to reform, while exit is the depar­ture to a dif­fer­ent soci­ety. A pro­vi­sion­al exam­ple Land offers is white flight, the mid-cen­tu­ry exo­dus of afflu­ent cau­casian fam­i­lies to the sub­urbs.

Neo­re­ac­tionar­ies don’t advo­cate any kind of cen­tral social orga­ni­za­tion. Land envi­sions a “gov-corp,” a soci­ety run like a com­pa­ny, ruled by a CEO. Instead of peti­tion­ing a gov­ern­ment for redress of griev­ances, unsat­is­fied cus­tomers are free to take their busi­ness else­where. If this sounds medieval, neo­re­ac­tionar­ies don’t deny it — Yarvin some­times describes him­self as a “roy­al­ist,” or a “monar­chist,” or even a “Jaco­bite,” in ref­er­ence to 17th-cen­tu­ry oppo­nents of par­lia­men­tary influ­ence in British gov­ern­ment.

The ques­tion is, where do you go after exit­ing? NRx­ers don’t dis­miss the idea of com­pet­ing gov-corps on the same land mass, an idea antic­i­pat­ed by NRx intel­lec­tu­al fore­fa­ther Hans Her­man-Hoppe, an extreme lib­er­tar­i­an polit­i­cal sci­en­tist, who advo­cates for a sys­tem that he admits is essen­tial­ly feu­dal­ism. On a more abstract lev­el, the neo­re­ac­tionary fas­ci­na­tion with bit­coin imag­ines the escape to an alter­nate econ­o­my unen­cum­bered by fed­er­al reg­u­la­tion. Even Yarvin’s start­up, Urbit, seems to be ori­ent­ed towards exit: it promis­es an alter­na­tive inter­net inac­ces­si­ble to out­side users.

But the most utopi­an (dystopi­an?) wing of NRx lit­er­al­ly aims to build Love­craft­ian cities in the sea. This project, called Seast­eading, is cham­pi­oned by Yarvin’s on-and-off co-con­spir­a­tor Patri Fried­man, whose grand­fa­ther Mil­ton Fried­man hap­pens to be the econ­o­mist respon­si­ble for the most extreme free mar­ket poli­cies in the mod­ern world. Peter Thiel was once Seasteading’s prin­ci­pal backer, as well as an investor in Urbit.

It’s not hard to see why float­ing sov­er­eign states, out of any exist­ing nation’s juris­dic­tion, would appeal to the super-rich. At their most innocu­ous, they might serve as an exten­sion of an off­shore bank, allow­ing for eva­sion of any type of redis­trib­u­tive tax pol­i­cy. They also bring to mind the activ­i­ties of wealthy men like Jef­frey Epstein, who used his pri­vate Caribbean island to throw bac­cha­na­lian par­ties for his mil­lion­aire and bil­lion­aire friends, alleged­ly revolv­ing around the sex­u­al assault of minors.

The path of exit doesn’t end at the water’s edge. Though you won’t hear him pro­mot­ing NRx rhetoric, Elon Musk is com­mit­ted to the idea in his own way, keep­ing one eye on Mars and one under­ground.

“A Prophetic Warning”

Yarvin has giv­en the ide­ol­o­gy of his ene­my – that is, con­tem­po­rary lib­er­al soci­ety itself – an even longer series of names than he did his own: “pro­gres­sivism,”  “cryp­to-Calvin­ism,” “uni­ver­sal­ism,” “demo­tism,” and so on. The term that he adopt­ed per­ma­nent­ly, though, is “the Cathe­dral.” It first appeared in the fourth install­ment of his four­teen-part series “An Open Let­ter to Open-Mind­ed Pro­gres­sives,” which, along with the nine-part “Gen­tle Intro­duc­tion” and the sev­en-part “How Dawkins Got Pwned,” is con­sid­ered his major state­ment.

Michael Anissomov’s more suc­cinct Neo­re­ac­tionary Glos­sary defines the Cathe­dral as “the self-orga­niz­ing con­sen­sus of Pro­gres­sives and Pro­gres­sive ide­ol­o­gy rep­re­sent­ed by the uni­ver­si­ties, the media, and the civ­il ser­vice.” It’s named for a reli­gious struc­ture because that, accord­ing to Yarvin, is what it is. It’s a descen­dent of the Puri­tan church, func­tion­ing to sup­press dis­sent from its ortho­doxy of egal­i­tar­i­an­ism and democ­ra­cy, which Yarvin calls the Syn­op­sis.

Mild-man­nered Cur­tis Yarvin must have been sur­prised, then, when the Cathedral’s atten­tions land­ed square­ly on his alter ego Men­cius Mold­bug. In the weeks after Trump’s inau­gu­ra­tion, Politi­co report­ed that accord­ing to an unnamed source, Yarvin has “opened up a line to the White House, com­mu­ni­cat­ing with Ban­non and his aides through an inter­me­di­ary.” The claim remained unver­i­fied, as Yarvin “does not do inter­views and could not be reached for this sto­ry.”

Vox man­aged to inter­view Yarvin lat­er that day. “The idea that I’m ‘com­mu­ni­cat­ing’ with Steve Ban­non through an ‘inter­me­di­ary’ is pre­pos­ter­ous,” he said. “I have nev­er met Steve Ban­non or com­mu­ni­cat­ed with him, direct­ly or indi­rect­ly.” A few days lat­er, The Atlantic asked Yarvin about his alleged inter­me­di­ary. He claimed it was Twit­ter user @BronzeAgePerv, whose pro­file describes him as a “Nation­al­ist, Fas­cist, Nud­ist Body­builder!”

Yarvin’s eva­sive­ness makes it hard to tell whether he’s hid­ing some­thing, or just trolling. But it’s no sur­prise he reserved the major­i­ty of his con­tempt for The Atlantic, which, in the orig­i­nal Dark Enlight­en­ment sequence, Nick Land called the “core Cathe­dral-mouth­piece.” The Atlantic went on to speak to Land, who was his usu­al self. “NRx was a prophet­ic warn­ing about the rise of the Alt-Right,” he said.

NRx has got­ten some atten­tion before. A piece in Techcrunch in 2013, The Baf­fler in 2014, and The Awl in 2015 have all offered sur­veys of the ide­ol­o­gy. The main­stream media took notice of one par­tic­u­lar event, when Yarvin was dis­in­vit­ed from the Strangeloop tech con­fer­ence after the orga­niz­ers dis­cov­ered his blog. Bre­it­bart’s Allum Bokhari wrote an arti­cle in his favor, argu­ing that Yarvin’s pol­i­tics are “abstract.” There is wide spec­u­la­tion among read­ers about just how seri­ous Yarvin is, includ­ing from his most promi­nent read­er. “Vast struc­tures of his­tor­i­cal irony shape his writ­ings, at times even engulf­ing them,” says Nick Land.

The Cathedral Bell

“Vast struc­tures of his­tor­i­cal irony” is a rather gen­er­ous descrip­tion of what’s known on the inter­net as “shit­post­ing.” Know Your Meme defines the term as “a range of user mis­be­hav­iors and rhetoric on forums and mes­sage boards that are intend­ed to derail a con­ver­sa­tion.” This isn’t just Yarvin’s response to inter­views, it’s his whole rhetor­i­cal style. His atten­tion-seek­ing con­trar­i­an­ism, which suc­cess­ful­ly dis­tracts both web-surf­ing nerds and main­stream media reporters, dis­guis­es pol­i­tics that are more con­ven­tion­al than they appear.

The Atlantic claims that Bannon’s alleged con­tact with Yarvin is a “sign of his rad­i­cal vision,” evi­dence of an unprece­dent­ed shift to the right. Ban­non views the world as under­go­ing a “a clash of civ­i­liza­tions, fea­tur­ing a strug­gle between glob­al­ism and a down­trod­den work­ing class as well as between the Islam­ic and West­ern worlds.”

But in fact, the The Atlantic was where the phrase “clash of civ­i­liza­tions” was first used to describe glob­al pol­i­tics, in a 1990 arti­cle by Bernard Lewis called “The Roots of Mus­lim Rage.” Even the “gov-corp” is no aber­ra­tion. Trump has promised to “run our coun­try the way I’ve run my com­pa­ny,” and indeed, has filled his cab­i­net with the most bil­lion­aires of any pres­i­den­tial admin­is­tra­tion in his­to­ry. The gov-corp mod­el is endem­ic to Amer­i­can pol­i­tics, with its most explic­it expres­sion by an Amer­i­can politi­cian in Woodrow Wilson’s 1887 essay “The Study of Admin­is­tra­tion.” It’s also the cor­ner­stone of the phi­los­o­phy of neolib­er­al­ism, as prop­a­gat­ed by Friedrich Hayek, von Mis­es, and Mil­ton Fried­man. Under the neolib­er­al order, we are not homo sapi­ens but homo eco­nom­i­cus, eco­nom­ic agents moti­vat­ed only by ratio­nal self-inter­est. Lib­er­ty is reduced to par­tic­i­pa­tion in a com­pet­i­tive mar­ket.

It was in The New Repub­lic that the most odi­ous aspect of NRx ide­ol­o­gy, sci­en­tif­ic racism or so-called “race real­ism,” entered con­tem­po­rary polit­i­cal dis­course. In 1994, under then-edi­tor Andrew Sul­li­van — who con­tin­ues to show not the least bit of remorse — the mag­a­zine pub­lished excerpts from Richard J. Her­rn­stein and Charles Murray’s The Bell Curve, a book that argued that eco­nom­ic dis­ad­van­tages among minor­i­ty demo­graph­ics were due to low­er cog­ni­tive abil­i­ty. NRx­ers sub­scribe to a more explic­it ver­sion of this idea, which they refer to using the euphemism “human bio­di­ver­si­ty.”

In 2012’s Com­ing Apart, Mur­ray expand­ed his argu­ment, claim­ing that poor whites are unable to rise above their sta­tion due to the same cog­ni­tive defects The Bell Curve had pre­vi­ous­ly iden­ti­fied in peo­ple of col­or. More recent­ly, after Trump’s elec­tion, Kevin Williamson of Nation­al Review wrote that the poor whites of “dys­func­tion­al, down­scale com­mu­ni­ties” in the Rust Belt “deserve to die.” They are “neg­a­tive assets” who have brought their lot upon them­selves. Per­haps it’s no coin­ci­dence that the arti­cle makes know­ing ref­er­ence to the Cathe­dral and cites Yarvin by name.

Williamson isn’t the only main­stream pun­dit who reads Yarvin. Rod Dreher has referred to the Cathe­dral in The Amer­i­can Con­ser­v­a­tive, as has Ross Douthat at the New York Times. In the ear­ly stages of the gen­er­al elec­tion cam­paign, Douthat tweet­ed: “Trump-Mold­bug. Just putting it out there.”

The New Repub­lic itself is back on the case. A recent arti­cle by Kevin Bak­er took up the propo­si­tion pre­vi­ous­ly advanced by Nation­al Review, on behalf of the polit­i­cal cen­ter. Bak­er called for a “Bluex­it” of afflu­ent coastal lib­er­als who no longer want to share their coun­try with Trump vot­ers. “Truth is, you red states just haven’t been pulling your weight,” he said, sound­ing remark­ably like a neo­con­ser­v­a­tive address­ing the nation’s minori­ties. Land linked the arti­cle on his blog, com­ment­ing, “sim­ply, yes.”

White Flight to Mars

In spite of its total lack of valid­i­ty, this kind of racist and elit­ist pseu­do­science, explic­it­ly nur­tured by the neolib­er­al main­stream, con­tin­ues to be accept­ed by respectable, palat­able pun­dits. NRx gets no cred­it for intro­duc­ing such ide­olo­gies; it has only tak­en them to their extreme yet nec­es­sary con­clu­sions. The reac­tionary ver­sion of human bio­di­ver­si­ty has been kept alive across a wide spec­trum of the right, from the aris­to­crat­ic white nation­al­ists of Amer­i­can Renais­sance to the Pepe frogs and ani­me trolls of 4chan. With­out explic­it­ly sup­port­ing them, Land has aligned him­self with them. His accep­tance has been mutu­al, with the Dark Enlight­en­ment becom­ing a top­ic of con­ver­sa­tion at Amer­i­can Renais­sance’s 2014 nation­al con­fer­ence.

Much of the Dark Enlight­en­ment sequence is devot­ed to an apolo­gia for John Der­byshire, a for­mer Nation­al Review staffer who has became a fel­low-trav­el­er to white suprema­cists. His essay “The Talk: Non­black Ver­sion,” writ­ten in the wake of Trayvon Martin’s mur­der, was a heat­ed defense of the pre­sump­tion of guilt for black men. Typ­i­cal­ly, though, Land has added his own lay­er of com­pli­ca­tion to the argu­ment. In an edi­to­r­i­al for the Alter­na­tive Right blog, start­ed by the tit­u­lar movement’s orig­i­na­tor, Richard Spencer, and now run by his col­lab­o­ra­tor Col­in Lid­dell, Land named his the­o­ry of human genet­ics hyper­racism.

Land does endorse the idea of typ­i­cal lev­els of abil­i­ty cor­re­lat­ing to dif­fer­ent “sub-species” of humans. But unlike white nation­al­ists, he’s not inter­est­ed in dif­fer­en­ti­at­ing sole­ly by eth­nic­i­ty. Instead, he pri­or­i­tizes socioe­co­nom­ic sta­tus, call­ing it “a strong proxy for IQ.” Though race is cor­re­lat­ed along socioe­co­nom­ic lines, says Land, a “genet­i­cal­ly self-fil­ter­ing elite” would not be strict­ly racial­ly homoge­nous. A mer­i­toc­ra­cy allows supe­ri­or beings to rise to the top, and though most of them will be white and Asian, supe­ri­or­i­ty ulti­mate­ly falls along a dif­fer­ent “axis of vari­a­tion.” Per­haps tak­ing a cue from Musk, he con­cludes that “space col­o­niza­tion will inevitably func­tion as a high­ly-selec­tive genet­ic fil­ter.” White flight to Mars?

Rather than tak­ing a more extreme view than the likes of Mur­ray, Williamson, and now lib­er­al colum­nist Frank Rich, Land has sim­ply car­ried the main­stream ide­ol­o­gy to its inex­orable result. The ugly under­bel­ly of the con­ven­tion­al view of mar­ket soci­ety as a mer­i­toc­ra­cy is pre­cise­ly Land’s hyper­racism: the assump­tion that some peo­ple are more fit than oth­ers, and their socioe­co­nom­ic sta­tus is deserved. The con­tin­gent effects of spe­cif­ic his­tor­i­cal ten­den­cies and social insti­tu­tions are exalt­ed with the sup­pos­ed­ly prov­i­den­tial neces­si­ty of DNA. Thus the com­plex eco­nom­ic his­to­ry result­ing in the hege­mo­ny of Europe, the Unit­ed States, and East Asia is tak­en to mean that whites and Asians are the most bio­log­i­cal­ly fit; the effects of con­strained social mobil­i­ty and the self-rein­forc­ing effects of eco­nom­ic inequal­i­ty become the claim that pover­ty is her­i­ta­ble. The fan­ta­sy of mer­i­toc­ra­cy can­not sur­vive a con­fronta­tion with the real­i­ty of a world shaped by impe­ri­al­ism and white suprema­cy. But unlike lib­er­als who believe in the fan­ta­sy, Land admits its impli­ca­tions.

Though it is now put to the ser­vice of the hyper­racist agen­da, “human bio­di­ver­si­ty” was ini­tial­ly a neu­tral term coined by anthro­pol­o­gist Jonathan Marks, whose work was an inno­v­a­tive syn­the­sis of the anthro­pol­o­gy and genet­ics. In the late nineties, it was adopt­ed by Steve Sail­er, a jour­nal­ist then at Nation­al Review, who sat perched on the fence between main­stream con­ser­vatism and white nation­al­ism. He has since fall­en off the far right end, and now writes for racist pub­li­ca­tions like VDARE.

Sci­en­tif­ic racism became a main­stream con­tro­ver­sy once again when New York Times writer Nicholas Wade’s 2014 book A Trou­ble­some Inher­i­tance argued for the dis­tinct cat­e­go­riza­tion of “three major races,” in a hier­ar­chi­cal tax­on­o­my that explains the his­tor­i­cal “rise of the west.” More than 100 pop­u­la­tion geneti­cists wrote an open let­ter to the Times dis­avow­ing Wade’s “mis­ap­pro­pri­a­tion of research from our field.” They con­clud­ed that “there is no sup­port from the field of pop­u­la­tion genet­ics for Wade’s con­jec­tures.”

Anoth­er dis­senter was Jonathan Marks. He has tire­less­ly reject­ed the mis­use of the term he coined, open­ly crit­i­ciz­ing A Trou­ble­some Inher­i­tance, The Bell Curve, and oth­er con­fla­tions of cul­ture and biol­o­gy. This did not require a revi­sion of his the­o­ry. His 1995 book Human Bio­di­ver­si­ty stat­ed from the out­set that “the hered­i­ty of race is not genet­ic, but social.”


Machine learn­ing can be so daz­zling, we tend to for­get that it’s shaped by human inter­ven­tion. As tri­umphant as Google was over its new trans­la­tion sys­tem, anoth­er recent machine learn­ing exper­i­ment — Microsoft’s Tay — showed just how volatile that rela­tion­ship can be. Intend­ed as the most innocu­ous AI pos­si­ble, Tay, an acronym for “think­ing about you,” was a sim­u­la­tion of a social media user mod­eled after a teenage girl. Tay was released to Twit­ter on March 23rd 2016, and start­ed the day mak­ing small talk, repeat­ing memes, and learn­ing the lyrics to “Nev­er Gonna Give You Up.” By after­noon, with the help of some prod­ding from 4channers, Tay had become a Holo­caust denier and 9/11 truther. Microsoft shut it down after 16 hours.

A report from Arti­fi­cial Intel­li­gence Now, a sym­po­sium on the poten­tial effects of machine intel­li­gence to soci­ety, offers an expla­na­tion of this phe­nom­e­non, and its broad­er impli­ca­tions. Machine learn­ing is sub­ject to data bias: “AI sys­tems depend on the data they are giv­en, and may reflect back the char­ac­ter­is­tics of such data, includ­ing any bias­es, in the mod­els of the world they cre­ate.” Machine learn­ing is a case of Land’s hyper­sti­tion, slip­ping between belief and tech­nol­o­gy. The val­ues of the pro­gram­mer shape the some­times tan­gi­ble out­puts of the result­ing machine.

The risk is that AI sys­tems could “exac­er­bate the dis­crim­i­na­to­ry dynam­ics that cre­ate social inequal­i­ty, and would like­ly do so in ways that would be less obvi­ous than human prej­u­dice and implic­it bias.” As prin­ci­pal researcher Kate Craw­ford puts it, arti­fi­cial intel­li­gence has a “white guy prob­lem.” There are dis­turb­ing exam­ples, like a study by ProP­ub­li­ca that found that a machine algo­rithm designed to mea­sure rates of recidi­vism was almost twice as like­ly to false­ly cat­e­go­rize black defen­dants as future crim­i­nals. And the soft­ware used for data min­ing by U.S. intel­li­gence agen­cies, pro­duced by Peter Thiel’s Palan­tir, hard­ly seems opti­mized to pro­tect civ­il lib­er­ties in the age of the Mus­lim Ban.

More­over, cyber­se­cu­ri­ty researcher Heather Roff has point­ed to the fre­quent gen­der­ing of humanoid robots: mil­i­tary tech­nol­o­gy, like the Navy’s grenade launch­er SAF­FiR, is built to resem­ble a male body, and ser­vice tech­nol­o­gy, like the iPhone’s Siri, is pre­sent­ed as female. Tra­di­tion­al gen­der roles that equate mas­culin­i­ty with pow­er and fem­i­nin­i­ty with sub­servience are repro­duced by design. This is no sur­prise, con­sid­er­ing that the ratio of women in the com­put­ing indus­try is at 26 per­cent, a drop from 35 per­cent in 1990, accord­ing to the AAUW. A 2016 sur­vey found that 88 per­cent of women in Sil­i­con Val­ley report­ed expe­ri­enc­ing uncon­scious gen­der bias at work.

Michael Aniss­miov told Giz­mo­do in 2015 about a coun­ter­part to AI: intel­li­gence aug­men­ta­tion, or the syn­the­sis of tech­nol­o­gy with the human mind. He described one poten­tial out­come: “a pow­er­ful leader mak­ing use of intel­li­gence enhance­ment tech­nol­o­gy to put him­self in an unas­sail­able posi­tion.” It’s a prospect that may strike you dif­fer­ent­ly depend­ing on whether or not you con­sid­er monar­chy a desir­able sys­tem of gov­ern­ment.

Even the sup­pos­ed­ly apo­lit­i­cal dream of tran­shu­man­ism con­ceals an ide­ol­o­gy. Like Anis­si­mov, Elon Musk antic­i­pates “a clos­er merg­er of bio­log­i­cal intel­li­gence and dig­i­tal intel­li­gence,” as he put it in a speech in Dubai. Mean­while, back on Earth, his employ­ees are held fast in a flesh­ly present. A Tes­la work­er recent­ly wrote a Medi­um post describ­ing the all-too-human con­di­tions Musk’s employ­ees are sub­ject to. “I often feel like I am work­ing for a com­pa­ny of the future under work­ing con­di­tions of the past,” he wrote.

“Real­ly don’t want to get in pol­i­tics. I just want to help invent and devel­op tech­nolo­gies that improve lives,” Musk said in a tweet. Regard­less, along with Peter Thiel, he has tak­en a role in Trump’s gov-corp. Good news for Yarvin, who told Vox that Musk is his choice for CEO-king of Amer­i­ca.

Indeed, fig­ures like Musk and Thiel don’t need to enter the polit­i­cal are­na to hold king­ly posi­tions. Oxfam recent­ly pub­lished data show­ing that eight men, includ­ing Sil­i­con Val­ley over­lords Bill Gates and Mark Zucker­burg, own as much wealth as half the world’s pop­u­la­tion. There is lit­tle sign that the archi­tects of emerg­ing tech­nolo­gies have any inten­tion of chang­ing these cir­cum­stances. Elon Musk doesn’t have to wait for a super­in­tel­li­gence to reward him. And the rest of us don’t have to wait to be reduced to pro­duc­tive machines with­in a net­work run by com­put­ers.

The Real Barrier

In 2013, Alex Williams and Nick Srnicek claimed “accel­er­a­tionism” for the Left, with their Man­i­festo for an Accel­er­a­tionist Pol­i­tics (MAP). Rather than fol­low­ing Land’s tran­shu­man­ist tra­jec­to­ry, they picked up the thread of polit­i­cal eman­ci­pa­tion left by Deleuze and Guat­tari, argu­ing that it should be pos­si­ble to “accel­er­ate the process of tech­no­log­i­cal evo­lu­tion” in order to apply it to “socio-polit­i­cal action” ori­ent­ed toward egal­i­tar­i­an ends.

Left accel­er­a­tionism is best known for an espe­cial­ly vul­gar vari­ant of its argu­ment, the eas­i­ly scorned notion that the left’s project should be to make cap­i­tal­ism as destruc­tive as pos­si­ble, in hopes of trig­ger­ing a rev­o­lu­tion. But the MAP text advances a more ratio­nal vari­ant, propos­ing that the pro­duc­tive forces of cap­i­tal­ism should be applied to a social demo­c­ra­t­ic pro­gram rather than the exist­ing one.

Land, how­ev­er, has dis­avowed any ori­en­ta­tion of the accel­er­a­tionist cur­rent toward left pol­i­tics. In a blog post crit­i­ciz­ing left accel­er­a­tionism, he instead char­ac­ter­izes the left as a “decel­er­a­tor,” imped­ing the real cap­i­tal­ist accel­er­a­tion advo­cat­ed by the “Out­er Right.”

Neo­re­ac­tion is Accel­er­a­tionism with a flat tire. Described less fig­u­ra­tive­ly, it is the recog­ni­tion that the accel­er­a­tion trend is his­tor­i­cal­ly com­pen­sat­ed. Beside the speed machine, or indus­tri­al cap­i­tal­ism, there is an ever more per­fect­ly weight­ed decel­er­a­tor, which grad­u­al­ly drains tech­no-eco­nom­ic momen­tum into its own expan­sion, as it returns dynam­ic process to meta-sta­sis. Com­i­cal­ly, the fab­ri­ca­tion of this brak­ing mech­a­nism is pro­claimed as progress. It is the Great Work of the Left. Neo­re­ac­tion aris­es through nam­ing it (with­out exces­sive affec­tion) as the Cathe­dral.

He gives a “tele­o­log­i­cal def­i­n­i­tion” to the Cathe­dral, which per­forms its “emer­gent func­tion as the can­cel­la­tion of cap­i­tal­ism.” While his­to­ry is ori­ent­ed toward “accel­er­a­tion into tech­no-com­mer­cial Sin­gu­lar­i­ty,” the pro­gres­sive Cathe­dral “is the anti-trend required to bring his­to­ry to a halt.”

Williams and Srnicek are at odds with this inter­pre­ta­tion. They draw from Deleuze and Guattari’s account of cap­i­tal­ism, which itself draws from a sug­ges­tive idea artic­u­lat­ed in Vol­ume 3 of Cap­i­tal. While Marx said that “the real bar­ri­er of cap­i­tal­ist pro­duc­tion is cap­i­tal itself,” Williams and Srnicek con­clude that “cap­i­tal­ism can­not be iden­ti­fied as the agent of true accel­er­a­tion.” Their for­mu­la­tion argues that “cap­i­tal­ism has begun to con­strain the pro­duc­tive forces of tech­nol­o­gy, or at least, direct them towards need­less­ly nar­row ends.”

As the MAP puts it, “rather than a world of space trav­el, future shock, and rev­o­lu­tion­ary tech­no­log­i­cal poten­tial, we exist in a time where the only thing which devel­ops is mar­gin­al­ly bet­ter con­sumer gad­getry.” This is unde­ni­ably true. But although apply­ing an egal­i­tar­i­an eth­ic to the con­struc­tion of future machines is a wor­thy goal, cer­tain­ly more so than what Williams has described as Land’s lapse into “sick per­ver­si­ty,” there is a more imme­di­ate con­cern: who owns the exist­ing machines, here and now, and who builds them?

The ten­den­cy of the com­mu­ni­ty that builds and oper­ates those machines, from titans like Peter Thiel to cult fig­ures like Cur­tis Yarvin, is open­ly total­i­tar­i­an. The New York Times has report­ed that polit­i­cal dona­tions from Sil­i­con Val­ley PACs took a shift from the Demo­c­ra­t­ic Par­ty toward the GOP in 2016. But their influ­ence on soci­ety is not mere­ly chan­neled through the prof­it made by machines. It is built into the machines them­selves. If, as Jason Smith puts it, “pat­terns of tech­no­log­i­cal devel­op­ment increas­ing­ly reflect cap­i­tal­ist val­ue-rela­tions,” then accel­er­at­ing capital’s inter­nal ten­den­cies may imply mass unem­ploy­ment and eco­log­i­cal cat­a­stro­phe rather than a new hori­zon of lux­u­ry and eman­ci­pa­tion.

Beasts of Burden

In his crit­i­cal his­to­ry of accel­er­a­tionism, Malign Veloc­i­ties, Ben­jamin Noys likens Land’s vision of cap­i­tal­ism to a Basiliskesque mon­ster, H.P. Lovecraft’s “Shog­goth.” It is a hor­ri­fy­ing “beast of bur­den” cre­at­ed by the mys­te­ri­ous “Old Ones,” whose body, like a Deep Dream, is cov­ered in shift­ing, pul­sat­ing eyes.

Cap­i­tal­ism, for the accel­er­a­tionist, bears down on us as accel­er­a­tive liq­uid mon­stros­i­ty, capa­ble of absorb­ing us and, for Land, we must wel­come this. The his­to­ry of slave labor and lit­er­al­ly mon­strous class strug­gle is occlud­ed in the accel­er­a­tionist invo­ca­tion of the Shog­goth as liq­uid and accel­er­a­tive dynamism. The hor­ror involves a for­get­ting of class strug­gle (even in dubi­ous fic­tion­al form) and the abo­li­tion of fric­tion in the name of immer­sion.

The eli­sion of class antag­o­nism is lit­er­al­ly obscured by machin­ery. Exist­ing tech­nol­o­gy immers­es us in the extreme polit­i­cal pro­gram prof­fered by neolib­er­al doc­trine. Through data bias, the pol­i­tics of tech cul­ture will invis­i­bly shape the social orga­ni­za­tion that results from the tech­nolo­gies of the future. The fur­ther right Sil­i­con Val­ley shifts, the more dan­ger­ous their machines will become.

In Feb­ru­ary, a con­fer­ence con­vened in Asilo­mar, Cal­i­for­nia, ded­i­cat­ed to the devel­op­ment of social­ly con­scious “AI Prin­ci­ples.” It was a lit­er­al assem­bly of what Land, in his Ccru days, named the “Human Secu­ri­ty Sys­tem,” the means by which soci­ety obstructs our sub­jec­tive merg­ing with tech­nol­o­gy. Wired report­ed that in the conference’s open­ing speech, MIT econ­o­mist Andrew McAfee dis­missed “Ter­mi­na­tor sce­nar­ios,” instead point­ing to sta­tis­tics regard­ing the effect of automa­tion on jobs.

The new data McAfee cit­ed showed an ero­sion of the mid­dle class, with low-income and high-income jobs con­tin­u­ing to build in vol­ume. “If cur­rent trends con­tin­ue,” he said, “peo­ple are going to rise up well before the machines do.” Accord­ing to Wired, AI researchers lat­er accost­ed McAfee in the hall­ways to warn him that his sta­tis­tics under­stat­ed the speed at which AI would ampli­fy class dis­par­i­ties.

For­get time-trav­el­ing killer robots or ancient beasts. NRx has sim­ply exposed the oper­a­tions of the cap­i­tal­ist machine in the present. Main­stream apol­o­gists for neolib­er­al­ism have a deci­sion to make: whether to embrace the pseu­do­science of Sil­i­con Val­ley hyper­racism, or to reject the vast eco­nom­ic inequal­i­ties gen­er­at­ed by mar­ket soci­ety. If the polit­i­cal class is ded­i­cat­ed to keep­ing the machine run­ning, it falls to the rest of us to shut it down.

Author of the article

is a writer and musician based in Brooklyn.