8-2 Table of Con­tents | http://​dx​.doi​.org/​1​0​.​1​7​7​4​2​/​I​M​A​G​E​.​L​D​.​8​.​2.2 | Von­der­auPDF

Abstract | When in 2011 a world-lead­ing IT com­pa­ny expressed the inten­tion to locate its infra­struc­ture in the Swedish city of Luleå, this announce­ment imme­di­ate­ly trig­gered future sce­nar­ios and visions of a new indus­tri­al era, eco­nom­ic pros­per­i­ty, and chang­ing urban life. Such antic­i­pa­tion was sup­port­ed and shaped by munic­i­pal plan­ning and busi­ness-man­age­ment activ­i­ties that soon mate­ri­al­ized in the form of build­ing sites, region­al devel­op­ment strate­gies, and new mar­kets. Since the actu­al name and oper­a­tions of the IT com­pa­ny were kept entire­ly secret, the plan­ning and imple­men­ta­tion of “Project Gold”—as the data cen­tre project was called locally—was as much dri­ven by col­lec­tive imag­i­nar­ies as by hard facts or past expe­ri­ences. This arti­cle is based on an ethno­graph­ic study that fol­lowed the imple­men­ta­tion of Facebook’s first Euro­pean data cen­tre in Luleå. The paper ana­lyzes dif­fer­ent modes of data cen­tre infra­struc­tur­al (in)visibility and shows how imag­i­nar­ies became influ­en­tial both for imple­ment­ing the cloud in Luleå and for shap­ing the antic­i­pat­ed time and space of “post-extrac­tive moder­ni­ty.” More specif­i­cal­ly, the paper focus­es on socio-tech­ni­cal pre­con­di­tions as well as con­crete prac­tices and styles—technologies of imagination—enabling those imaginaries.

Asta Von­der­au | Stock­holm University

Technologies of Imagination: Locating the Cloud in Sweden’s North
Introduction: The Mayor’s Dream

"You know, when Swedish may­ors dream, they usu­al­ly dream of IKEA open­ing a store in town. How­ev­er, there are more than 30 IKEAs in this coun­try but there is only one Facebook—here, in Luleå!” (Inter­view, April 2014). This is how Karl Petersen, munic­i­pal com­mis­sion­er (or as he called him­self, “a may­or”) of Luleå, an indus­tri­al town locat­ed near the Arc­tic Cir­cle in Sweden’s North, intro­duced me to his municipality’s most ambi­tious indus­tri­al devel­op­ment project ever, a Face­book data cen­tre. Petersen admit­ted to hav­ing been igno­rant of IT infra­struc­tur­al needs before he was con­tact­ed by the Luleå Busi­ness Agency with the idea of sell­ing the Nordic cold to the cloud indus­try. He was proud of hav­ing real­ized in good time that “the cloud has a big future because data cen­tres are need­ed all over the world and they need to be cooled” (Inter­view, April 2014). This busi­ness idea, uncon­ven­tion­al for a Swedish munic­i­pal­i­ty, led to unex­pect­ed suc­cess when the world’s largest social-net­work­ing ser­vice announced its deci­sion to locate its first Euro­pean data cen­tre in Luleå.[1]

Facebook’s announce­ment made Luleå known in Swe­den and abroad as a place “where the cloud hits the ground,” as my inter­locu­tors would put it. It also brought overnight fame to the munic­i­pal com­mis­sion­er and his col­leagues. Among the count­less inter­views and news reports that result­ed from a fren­zy of nation-wide and even glob­al media atten­tion, a cross-pro­mo­tion­al video enti­tled Swedish Think­ing: Jour­ney to the Node Pole fea­tured Petersen him­self speak­ing out for his town’s and the region’s abil­i­ty to cre­ate ben­e­fits from a harsh Arc­tic cli­mate and periph­er­al geo­graph­i­cal loca­tion. Prais­ing this abil­i­ty of “get­ting more from less” as a nation­al char­ac­ter trait, the com­mer­cial made it clear that the mayor’s dream was direct­ly linked to eco­nom­ic devel­op­ments and col­lec­tive visions going far beyond per­son­al aspi­ra­tions and the bound­aries of Sweden’s north­ern­most coun­ty, Norrbotten.

Fig­ure 1. Luleå’s munic­i­pal com­mis­sion­er Karl Petersen, fea­tured in a video pro­mo­tion for Vol­vo and the Node Pole.

Petersen’s Face­book project intro­duced a new indus­try to the region and also direct­ed the atten­tion of glob­al IT infra­struc­ture providers to Swe­den as a key Euro­pean loca­tion for cloud infra­struc­ture. This devel­op­ment was in line with the Swedish government’s plan, laid out in the Dig­i­tal Agen­da for Swe­den, to turn the coun­try into one of the strongest IT nations world­wide, and con­se­quent­ly trig­gered antic­i­pa­tion of a new indus­tri­al era ground­ed in the vision of an upcom­ing post-extrac­tive future.

Fig­ure 2. Face­book thumb in the lunch room of Luleå Busi­ness Agency, pho­tographed by author, April 2014.

This essay inves­ti­gates how the glob­al cloud mate­ri­al­izes in a par­tic­u­lar socio-cul­tur­al con­text. Based on an ethno­graph­ic study of Facebook’s data cen­tre imple­men­ta­tion in Luleå, the paper shows how this large-scale infra­struc­tur­al project forms part of an “econ­o­my of antic­i­pa­tion” char­ac­ter­is­tic of the tem­po­ral pol­i­tics of today’s (dig­i­tal) cap­i­tal­ism (Cross). It also demon­strates how in the course of locat­ing the cloud, nation­al ter­ri­to­ries are recon­fig­ured, new sub­jec­tiv­i­ties and bod­i­ly expe­ri­ences are enabled, and local future sce­nar­ios are reshaped by means of antic­i­pa­tion and imag­i­na­tion. Focused on the every­day infra­struc­ture-mak­ing process­es, this study engages anthro­po­log­i­cal­ly with ques­tions of loca­tion of pow­er, agency, and prac­tice in IT economies.

I have fol­lowed the plan­ning and imple­men­ta­tion of Facebook’s data cen­tre over a peri­od of almost two years. Due to the com­plex­i­ty of large-scale infra­struc­tur­al projects and because of the data centre’s dif­fi­cult acces­si­bil­i­ty, my field­work activ­i­ties did not aim for long-term research stays at sin­gle sites. Instead, it con­sist­ed of var­i­ous “poly­mor­phous engage­ments” (Gusterson)—that is, of mul­ti-sit­ed inter­ac­tions around Facebook’s Data Cen­ter project in dif­fer­ent insti­tu­tion­al and indus­tri­al con­texts, engag­ing with diverse expert cul­tures, dis­cours­es, humans, and tech­nolo­gies. This includ­ed, for instance, inter­views with region­al and nation­al deci­sion mak­ers, munic­i­pal employ­ees, data cen­tre engi­neers and man­agers, and rep­re­sen­ta­tives of indus­tri­al orga­ni­za­tions such as Data Cen­ter Dynam­ics; obser­va­tions at data cen­tre build­ing sites, at IT-relat­ed indus­tri­al and pub­lic events, and par­tic­i­pa­tion at con­fer­ences; and analy­sis of local and inter­na­tion­al media and polit­i­cal doc­u­ments such as maps, region­al devel­op­ment pro­grams, and nation­al IT strate­gies.[2]

Sit­u­at­ed with­in the inter­dis­ci­pli­nary field of infra­struc­ture stud­ies, this paper inves­ti­gates Facebook’s data cen­tre as an aspect of cloud infra­struc­ture. As such, a data cen­tre is more than just a piece in a func­tion­al tech­no­log­i­cal sys­tem, it is a con­test­ed and fun­da­men­tal­ly rela­tion­al tech­no-social con­fig­u­ra­tion (see Hu; Larkin; Dour­ish and Bell; Parks; Starosiel­s­ki; Bowk­er et al.). The con­crete strate­gies and prac­tices of local­iz­ing the cloud are here under­stood as a form of infra­struc­tur­ing (Niewöhner)—a con­tin­u­ous bring­ing togeth­er, relat­ing, and coor­di­nat­ing of tech­nolo­gies, com­mu­ni­ties of actors, orga­ni­za­tion­al struc­tures, and moral val­ues. A per­spec­tive on infra­struc­tur­ing empha­sizes the com­plex­i­ty of work inher­ent in infra­struc­ture-mak­ing process­es. The infra­struc­tur­ing of the cloud will most­ly be described from the per­spec­tive of local author­i­ties and experts, includ­ing politi­cians, busi­ness devel­op­ers, con­sul­tants, munic­i­pal city plan­ners, archi­tects, and envi­ron­men­tal and IT experts. Acknowl­edg­ing the dif­fer­ing tasks and con­cerns of these actors regard­ing data cen­tre imple­men­ta­tion process­es, and con­trast­ing such con­cerns with Facebook’s own cor­po­rate inter­ests when it comes to infra­struc­tur­al secre­cy and con­trol­ling (in)visibility, I ana­lyze some of the cloud’s local social effects.

Accord­ing to Jamie Cross, large-scale infra­struc­tur­al projects form part of a glob­al econ­o­my of antic­i­pa­tion char­ac­ter­is­tic of con­tem­po­rary cap­i­tal­ism. These economies are under­pinned with an abun­dance of dreamed-of futures in rela­tion to which the present is con­stant­ly rene­go­ti­at­ed (Cross 8). Infra­struc­tur­al projects rep­re­sent are­nas of fan­ta­sy and desire; they are affec­tive spaces “in which the lived sen­sa­tion of future prospects can seize bod­ies, per­sons and selves” (Cross 9). Accord­ing­ly, I under­stand the Swedish Face­book data cen­tre and infra­struc­ture in gen­er­al as such an object of fan­ta­sy and desire that pro­vokes a “sens­ing of moder­ni­ty” (Mrazek)—that is, aspi­ra­tions and expec­ta­tions of a new time, “a process by which the body as much as the mind appre­hend what it is to be mod­ern, muta­ble and pro­gres­sive” (Larkin 337). The recon­fig­u­ra­tion of ter­ri­to­ries and sub­jec­tiv­i­ties that results from cloud infra­struc­tur­ing thus includes not only mate­r­i­al and bod­i­ly but also imag­i­na­tive enti­ties (Har­vey 17).

Imagination—understood as a prac­tice of syn­thet­ic knowl­edge pro­duc­tion, “the bring­ing togeth­er of diverse forms of appear­ance and the abil­i­ty to relate them” (Sneath et al. 11)—is thus inher­ent to process­es of infra­struc­tur­ing and the social trans­for­ma­tions such process­es enable. In what fol­lows, the focus is on dif­fer­ent actors’ attempts to real­ize imag­ined futures and on the fric­tions result­ing from the col­li­sion between dif­fer­ing visions of what such futures might entail. Hence, an ethnog­ra­phy of cloud infra­struc­tur­ing allows to iden­ti­fy what Sneath, Hol­braad, and Ped­er­sen call “tech­nolo­gies of imag­i­na­tion” or the “spe­cif­ic mate­r­i­al and spa­tial means by which par­tic­u­lar imag­in­ings are gen­er­at­ed” (6). In the case of Facebook’s Swedish data cen­tre, such tech­nolo­gies of imag­i­na­tion were instru­men­tal for con­nect­ing indus­tri­al mate­ri­al­i­ties and cor­po­rate images of the cloud to local­ly sit­u­at­ed ontolo­gies, expe­ri­ences, and his­tor­i­cal hori­zons of mean­ing, and for shap­ing the glo­cal geo­gra­phies of the cloud as a space of antic­i­pat­ed post-extrac­tive modernity.

Technologies of Imagination

Cor­po­rate (In)Visibility 

Cloud tech­nolo­gies are objects of pop­u­lar imag­i­na­tion and cor­po­rate mar­ket­ing. In order to under­stand how par­tic­u­lar tech­nolo­gies of imag­i­na­tion become instru­men­tal for local process­es of cloud infra­struc­tur­ing, it is there­fore nec­es­sary to inquire how the cloud is being brand­ed by IT com­pa­nies and rep­re­sent­ed in pop­u­lar media. Fur­ther­more, how do these pop­u­lar images of the cloud relate to the actu­al tech­no­log­i­cal oper­a­tion of data stor­age and processing?

Cloud com­put­ing is a cur­rent­ly dom­i­nat­ing com­pu­ta­tion and data-pro­cess­ing mod­el on which most social net­works and online ser­vices are based. Data cen­tres rep­re­sent the core of cloud infra­struc­ture and one of the fastest-grow­ing indus­tries world­wide. Their ener­gy con­sump­tion rep­re­sents over three per­cent of glob­al elec­tric­i­ty usage and is grow­ing at a rate of four­teen per­cent annu­al­ly. This ener­gy-con­sum­ing heavy indus­try, how­ev­er, is still not in the pub­lic eye. Media reports and IT com­pa­nies rep­re­sent the cloud, and the Inter­net more gen­er­al­ly, as being nat­ur­al, imma­te­r­i­al, and fluid—think of the lan­guage reg­u­lar­ly used for describ­ing such tech­nolo­gies and the metaphors of the “cloud,” “data streams,” or “IT ecolo­gies” (see Blum; Hogan). Such images are sup­port­ed and instru­men­tal­ized by IT com­pa­nies in mar­ket­ing cam­paigns that often present the cloud and data cen­tres as sus­tain­able by con­trast­ing them with “old­er” and ”dirt­i­er” industries.

Dema­te­ri­al­ized per­cep­tions of the cloud are ground­ed in a spe­cif­ic rela­tion between the cloud as a (vir­tu­al) net­work of data pro­cess­ing and data cen­tres as the cloud’s sin­gle infra­struc­tur­al site. While data cen­tres are locat­ed in con­crete geo­graph­ic places with­in nation­al ter­ri­to­ries, their tech­no­log­i­cal oper­a­tions can range far beyond those ter­ri­to­ries and leg­isla­tive bound­aries. The cloud and its data cen­tres thus enter­tain a rela­tion­ship of “inti­ma­cy at a dis­tance” by form­ing part of each oth­er while being dif­fer­ent­ly sit­u­at­ed in spa­tial terms at the same time (see Rossiter). The cloud’s mate­ri­al­i­ties, its social effects, and the envi­ron­men­tal impact of the data cen­tre indus­try are uneven­ly dis­trib­uted across spa­tial dimen­sions. This arrange­ment sup­ports dema­te­ri­al­ized per­cep­tions of the cloud. Users of cloud ser­vices, for instance, locat­ed far away from the actu­al stor­ages of their data, may remain unaware of the cloud’s mate­r­i­al exis­tence and its resource needs.

Mov­ing beyond the cloud’s tech­no­log­i­cal log­ics, in The Pre­his­to­ry of the Cloud Tung-Hui Hu explains that such dema­te­ri­al­ized metaphor­ic rep­re­sen­ta­tions are not only instru­men­tal for pop­u­lar descrip­tions but actu­al­ly form a major part of the cloud itself, which “has nev­er real­ly been about com­put­ing because com­put­ing is just one part of a larg­er cul­tur­al fan­ta­sy” (145). Accord­ing to Hu, the busi­ness mod­el of the cloud is based on pop­u­lar visions of the Inter­net as a vir­tu­al, bor­der­less, and open space, and on imag­i­nar­ies of an all-embrac­ing con­nec­tiv­i­ty envi­sioned to be inevitable for our society’s future devel­op­ment. As has been not­ed by var­i­ous schol­ars, such dema­te­ri­al­ized images of the cloud obscure its infra­struc­tur­al and indus­tri­al mate­ri­al­i­ties as well as its prob­lem­at­ic social and envi­ron­men­tal con­se­quences, includ­ing the enor­mous elec­tric­i­ty and water needs of data cen­tres, the increas­ing pol­lu­tion through waste heat, or the low num­ber of job oppor­tu­ni­ties the cloud indus­try offers to local com­mu­ni­ties (see Hogan; Cubitt et. al.; Gregg; Dour­ish and Bell). They also incen­tivize per­ma­nent con­nec­tiv­i­ty and data pro­duc­tion, nec­es­sary for gen­er­at­ing Big Data, and hence inte­gral to pre­vail­ing IT rev­enue mod­els. This is one key rea­son why larg­er IT com­pa­nies aim to con­trol the (in)visibility of cloud infrastructure.

In a sim­i­lar vein, Face­book presents itself to its users most­ly through visu­al­iza­tions of its glob­al vir­tu­al con­nec­tions. Rep­re­sen­ta­tions of data cen­tres are lim­it­ed to cleaned-up, sci­ence-fic­tion-style images that are meant to sug­gest trans­paren­cy and to evoke fas­ci­na­tion for tech­nol­o­gy with­out dis­clos­ing the actu­al work­ings or com­pli­ca­tions of com­mu­ni­ca­tion infra­struc­ture (Holt and Von­der­au).  Sim­i­lar strate­gies of dema­te­ri­al­iza­tion and delo­cal­iza­tion can be observed with regard to Luleå’s data cen­tre. Stor­ing the data of more than 800 mil­lion glob­al Face­book users, the Luleå data cen­tre does not oper­ate under Facebook’s name but as a sub­sidiary named Pin­na­cle Swe­den. Giv­en that Facebook’s Euro­pean head­quar­ters are locat­ed in Ire­land and that the social net­work is there­fore amenable to Irish leg­is­la­tion, it is ques­tion­able from a legal stand­point whether Face­book is even present in Swe­den or Luleå at all. Ques­tions of cor­po­rate invis­i­bil­i­ty extend beyond legal issues and also include, among oth­ers, the data centre’s out­side appear­ance. The Luleå build­ing resem­bles a large white box; no sounds or sights give an indi­ca­tion of the indus­try-scale data-stor­age process­es going on inside. No Face­book logos or street signs are to be found, apart from a few small flags with­in the fenced ter­ri­to­ry itself or the Face­book bikes used by employ­ees and con­struc­tion work­ers for trav­el­ling the enor­mous area at Luleå’s Datavä­gen (Data Street) [Fig. 4]. Despite the data centre’s extra­or­di­nary scale and tech­no­log­i­cal com­plex­i­ty, its rep­re­sen­ta­tives are reluc­tant to acknowl­edge the impor­tance of this infra­struc­tur­al site. The data centre’s man­ag­er Joel Kjell­gren, for instance, reg­u­lar­ly stress­es in pub­lic talks that even a com­plete data cen­tre out­age would not impede Facebook’s func­tion­al­i­ty and would remain unno­ticed by its users. While it is doubt­ful if such a major inci­dent would stay unnoticed—it cer­tain­ly would cause huge finan­cial loss­es and have a dev­as­tat­ing impact on the company’s image—such state­ments clear­ly con­tribute to Facebook’s strat­e­gy of chan­nel­ing pub­lic atten­tion from the local indus­tri­al to the glob­al and vir­tu­al dimen­sions of the social net­work.[3] Face­book thus seems to have lit­tle inter­est in being per­ceived as indus­tri­al or local, apart from its per­sis­tent pro­mo­tion of the Luleå data cen­tre as an envi­ron­men­tal­ly friend­ly and tech­no­log­i­cal­ly advanced cloud fol­low­ing Facebook’s turn towards sus­tain­able ener­gy consumption.

Con­trast­ing the cloud’s omnipres­ence and grow­ing impor­tance for today’s soci­ety with its infra­struc­tur­al secre­cy and inac­ces­si­bil­i­ty, as well as the data cen­tres’ “igno­rant rela­tion to archi­tec­ture” and the dif­fi­cul­ty of “decod­ing” them in social and cul­tur­al terms, archi­tec­ture the­o­rist Kazys Var­nelis has iden­ti­fied data cen­tres’ archi­tec­tur­al form as one embody­ing the cul­tur­al con­di­tions of dig­i­tal cap­i­tal­ism and what Deleuze terms the con­trol soci­ety. The case of Luleå’s data cen­tre indeed illus­trates IT com­pa­nies’ pow­er in con­trol­ling their infas­truc­tur­al vis­i­bil­i­ty. While ethno­graph­i­cal­ly trac­ing the every­day process­es of infra­struc­tur­ing, how­ev­er, the white box of the cloud appears much less sol­id and not just sim­ply imposed from above. As I show below, the gap that cor­po­rate secre­cy cre­ates between the glob­al vir­tu­al cloud and its local infra­struc­tur­al sites opens up a rela­tion­al space with­in which the cloud is decod­ed, nego­ti­at­ed, and con­test­ed in local­ly spe­cif­ic ways. With­in this space the cloud is reshaped in rela­tion to human and non-human rela­tions and con­cerns by means of tech­nolo­gies of imagination.

Fig­ure 3. Facebook’s first data cen­tre build­ing in Luleå, pho­tographed by author, April 2014.

Fig­ure 4. Entrance area of Facebook’s sec­ond data cen­tre build­ing, Datavä­gen 15 (Data Street 15), pho­tographed by author, Sep­tem­ber 2015.

Grasp­ing the Cloud

In our con­ver­sa­tions, Luleå’s politi­cians, city plan­ners, archi­tects, ecol­o­gists, and oth­er munic­i­pal experts empha­sized the secre­cy that marked the plan­ning process for Facebook’s data cen­tre. Dur­ing ini­tial meet­ings, Facebook’s rep­re­sen­ta­tives did not men­tion the name of their com­pa­ny and no busi­ness cards were exchanged. Com­mu­nal experts respon­si­ble for the project plan­ning were nei­ther enti­tled to know their client’s name nor its actu­al inten­tions (at least in the ear­ly plan­ning phase).[4] The infor­ma­tion need­ed for the plan­ning could only be acquired indi­rect­ly, through a group of con­tact per­sons, mak­ing the plan­ning process even more com­pli­cat­ed. Only the project’s work­ing title—“Project Gold”—and the high pri­or­i­ty it was giv­en by local and nation­al deci­sion mak­ers indi­cat­ed that large invest­ments were at stake.

In addi­tion, Facebook’s water and ener­gy needs were much high­er than for oth­er indus­tri­al projects. Cer­tain stan­dards, such as reg­u­la­tions for noise expo­sure (caused by the data centre’s 14 diesel gen­er­a­tors), did not even exist in Swe­den and had to be mod­elled on reg­u­la­tions in Ger­many and oth­er coun­tries. Despite all these com­pli­ca­tions, a detailed city plan—the main doc­u­ment for con­struc­tion projects in Sweden—had to be com­plet­ed with­in only three months instead of the usu­al one to two years. Fac­ing the extra­or­di­nary scale and pace of “Project Gold” and its secre­cy, city plan­ners and archi­tects had to be cre­ative, try­ing to envi­sion the secre­tive con­struc­tion object and to imag­ine it in con­text of its future local sur­round­ings. In inter­views they relat­ed, for instance, sto­ries of walk­ing around in the city in an attempt to imag­ine the planned build­ing, com­par­ing it with oth­er indus­tri­al build­ings and won­der­ing about its mag­ni­tude (Inter­view, Octo­ber 2014).[5]

In the ear­ly phase of its infra­struc­tur­ing, the cloud (and the data cen­tre as its spe­cif­ic local form) thus could nei­ther be deter­mined by local experts as an objec­tive­ly exist­ing mate­r­i­al enti­ty nor be eas­i­ly grasped sub­jec­tive­ly. Instead, the cloud was expe­ri­enced and became vis­i­ble in the sense pro­posed by Tim Ingold: as an assem­blage of “active mate­ri­als” that con­stant­ly emerge and trans­form through prac­tices of plan­ning, cal­cu­la­tion, and imag­i­na­tion (429). Facebook’s cor­po­rate strate­gies, aim­ing to estab­lish a cer­tain regime or order of infra­struc­tur­al vis­i­bil­i­ty for the social net­work, simul­ta­ne­ous­ly cre­at­ed dis­or­der when it came to city plan­ning and per­cep­tions of Luleå’s built envi­ron­ments. The usu­al plan­ning rou­tines and exist­ing exper­tise could not be eas­i­ly applied to this new project and new approach­es had to be invent­ed. In this par­tic­u­lar sit­u­a­tion, tech­nolo­gies of imagination—such as com­par­isons of the new indus­try to exist­ing indus­tri­al sites and imag­in­ing its build­ings in rela­tion to local environments—became instru­men­tal for the project’s suc­cess­ful imple­men­ta­tion. Such tech­nolo­gies also opened up new ways for envi­sion­ing and shap­ing the future of the region.

Fig­ure 5. Facebook’s Swedish cloud, seen from above. Build­ings to the right include stu­dent dor­mi­to­ries and Luleå’s Sci­ence Park. (Source: Genomförandebeskrivning/The Descrip­tion of Imple­men­ta­tion, Luleå Munic­i­pal­i­ty 2011).

Fig­ure 6. This draw­ing shows how Facebook’s data cen­tre (the thin line above the trees) was expect­ed to be vis­i­ble after com­ple­tion, thus illus­trat­ing the cloud’s pecu­liar infra­struc­tur­al (in)visibility. (Source: Miljökonsekvensbeskrivning/Description of Envi­ron­men­tal Con­se­quences, Luleå Munic­i­pal­i­ty 2011).

Fix­ing the Future

The secre­cy sur­round­ing “Project Gold” and the abstract­ness of its scale and scope pro­voked euphor­ic visions among Norrbotten’s experts and author­i­ties: a new indus­tri­al era of eco­nom­ic pros­per­i­ty, new urban cul­tures, and urban sub­jec­tiv­i­ties. As my inter­locu­tors relat­ed, such expec­ta­tions allowed for unique sol­i­dar­i­ty among all author­i­ties involved, trans­form­ing an extra­or­di­nary com­plex plan­ning process into a “sim­ple cal­cu­la­tion,” as dif­fi­cul­ties and hard­ships such as long over­time hours or inter­nal con­flicts could be weighed against the project’s expect­ed gains for the city of Luleå and the Nor­rbot­ten region. This col­lec­tive antic­i­pa­tion and belief in the pos­i­tive impact of a new indus­try was espe­cial­ly fos­tered by local deci­sion mak­ers and busi­ness devel­op­ers, as it helped accep­tance of the client’s pace of demands and redi­rect­ed pub­lic atten­tion away from the project’s enor­mous ener­gy and water needs, as well as from its pos­si­bly neg­a­tive impact on a near­by nature reserve.

Expec­ta­tions of a new indus­tri­al and urban era, how­ev­er, also relat­ed to per­son­al aspi­ra­tions among many of the involved actors to change the bur­dened image of Luleå as a periph­er­al “steel city” in order to enable cul­tur­al diver­si­ty and social change. My male inter­locu­tors, for instance, specif­i­cal­ly stressed suf­fer­ing from the tra­di­tion­al dom­i­nance of steel, min­ing, for­est, and oth­er extrac­tive indus­tries in this region. These indus­tries not only rep­re­sent­ed the only oppor­tu­ni­ties for male work­ers for decades, but also con­tributed to the estab­lish­ment of a spe­cif­ic type of unskilled mas­culin­i­ty, epit­o­mized by the stereo­typ­i­cal image of a dirty, mus­cu­lar steel work­er. These inter­locu­tors expect­ed indus­tri­al change to bring about both dif­fer­ent job oppor­tu­ni­ties and cul­tur­al and bod­i­ly diver­si­ty. Oth­er respon­dents also empha­sized being tired of Luleå’s bur­dened image as Sweden’s steel city. They even found it unfair that this image of the past kept dom­i­nat­ing how the city was per­ceived by nation­al author­i­ties and the country’s South­ern regions, even after the steel indus­try ceased to be the biggest employ­er. Accord­ing­ly, they hoped that indus­tri­al devel­op­ment and its promise of a new era would free the city from its past and intro­duce a future-ori­ent­ed temporality.

As Anna Tsing notes, how­ev­er, future sce­nar­ios and visions rep­re­sent frag­ile social orders that are always “almost-but-not-quite-there” (336). Estab­lish­ing and main­tain­ing such orders is there­fore nev­er trou­ble free. As my inter­locu­tors report­ed, before Facebook’s plans final­ly became offi­cial, local author­i­ties were in con­stant fear that “Project Gold” could still be stopped, since their secre­tive client was known to nego­ti­ate simul­ta­ne­ous­ly with oth­er regions in the coun­try and react­ed sen­si­tive­ly to even the small­est restric­tions of its inter­ests. A tech­ni­cal project coor­di­na­tor, for instance, recount­ed wait­ing months before a first sig­na­ture was giv­en, doc­u­ment­ing that the project would actu­al­ly be real­ized and that the client even­tu­al­ly would order to start cut­ting trees at the planned con­struc­tion site.

The fear of los­ing an unprece­dent­ed invest­ment and expect­ed future ben­e­fits led Luleå’s author­i­ties to var­i­ous com­pro­mis­es. Local city plan­ners told me of accept­ing the unfavourably stretched data cen­tre build­ing shape even though it resem­bled a wall sep­a­rat­ing the city from a near­by nat­ur­al reserve. The city even signed an “unusu­al agree­ment,” as a rep­re­sen­ta­tive of Luleå’s admin­is­tra­tion acknowl­edged, that promised Face­book huge amounts of drink­ing water, exceed­ing the com­mon offi­cial norm and only to be har­mo­nized with Swedish law by help of legal trick­ery (Inter­view, Sep­tem­ber 2015). This extra­or­di­nary effort to serve a pow­er­ful client and its tech­nolo­gies did not end with the data centre’s actu­al imple­men­ta­tion. Dur­ing my vis­its to Luleå in 2014–2015, munic­i­pal rep­re­sen­ta­tives were still eager to remain in a “con­stant mode of sell­ing,” as my inter­locu­tors called it, to secure invest­ments and an antic­i­pat­ed future. Indeed, infra­struc­tures are simul­ta­ne­ous­ly embed­ded with­in local social struc­tures while serv­ing as struc­tur­al mech­a­nisms in them­selves (Dour­ish and Bell 28). In the course of the cloud’s infra­struc­tur­al imple­men­ta­tion, the work modes and visions of Luleå’s munic­i­pal­i­ty regard­ing urban and region­al futures were restruc­tured. While the cloud was shaped and expe­ri­enced through local­ly spe­cif­ic infra­struc­tur­ing process­es and rela­tions, its local man­i­fes­ta­tions also were per­ceived as mark­ing the begin­ning of a new urban and region­al devel­op­ment. Such local per­cep­tions also relat­ed to his­tor­i­cal hori­zons of mean­ing and to expe­ri­ences of Nor­rbot­ten as an exploit­ed region.

Recon­fig­ur­ing the Region

The long-await­ed open­ing of Facebook’s data cen­tre in the sum­mer of 2013 trig­gered new visions and expec­ta­tions among locals. A Luleå Busi­ness Agency rep­re­sen­ta­tive explains: “If five or six years ago I would have gone down the street and told peo­ple that five years from now we would have bil­lions in invest­ments and a giant store­house for everyone’s hol­i­day pic­tures, nobody would have believed me. They prob­a­bly would have thought I was crazy, would have locked me into a base­ment and thrown the key away” (Inter­view, Octo­ber 2014). In this interviewee’s opin­ion, the open­ing of the Face­book data cen­tre also made reg­u­lar cit­i­zens come to believe that a change was immi­nent, that Luleå would be trans­formed from a key site of nation­al steel pro­duc­tion into a glob­al node of IT com­pe­tence. Such expec­ta­tions were quick­ly inte­grat­ed into region­al and busi­ness devel­op­ment pro­grams, such as The Strat­e­gy for the Devel­op­ment of Nor­rbot­ten into a Lead­ing Region for Cli­mate Smart Data Cen­ters (Granberg), which described the region as a “paper­less soci­ety,” freed from the mate­ri­al­i­ties iden­ti­fied with tra­di­tion­al extrac­tive industries.

Regard­less of how real­is­tic such strate­gies were, they expressed the attempt to rebrand the city, the region, and Swe­den as a cool and there­fore ide­al infra­struc­tur­al site, a com­man­do post of the glob­al cloud. The Node Pole, a new­ly estab­lished semi­pub­lic com­pa­ny, was a key actor that man­aged this re-brand­ing process and served as facil­i­ta­tor between pri­vate busi­ness­es and the state, lob­by­ing for the data cen­tre industry’s inter­ests, pro­mot­ing this indus­try in local and nation­al media, and pre­sent­ing North­ern Swe­den at inter­na­tion­al indus­tri­al fairs or con­fer­ences. The Node Pole includ­ed busi­ness devel­op­ers as well as IT and ener­gy experts who took the role of liai­son between city plan­ners and the client dur­ing the secre­tive data cen­tre plan­ning phase. Par­al­lel to their engage­ment with the Face­book project, these experts gath­ered a pool of exper­tise pro­vid­ed by region­al and inter­na­tion­al com­pa­nies that were inter­est­ed in sell­ing their con­struc­tion, con­sul­tan­cy, tech­nol­o­gy recy­cling, and oth­er ser­vices to the data cen­tre indus­try. In col­lab­o­ra­tion with Luleå and oth­er near­by cities, it start­ed offer­ing to sell land for poten­tial data cen­tre sites, pro­mot­ing sce­nar­ios of new indus­tri­al region­al futures in local media and pub­lic pre­sen­ta­tions. Facil­i­tat­ing suc­cess­ful and fast tech­no­log­i­cal solu­tions, the Node Pole also pro­vid­ed the region­al devel­op­ment dis­course with ideas of “remote intimacies”—new spa­tial frames for region­al devel­op­ment that were relat­ed to the poten­tial­ly glob­al ter­ri­to­ri­al­i­ty of data pro­cess­ing. This encour­aged region­al deci­sion mak­ers to think about the region and the city not only as a part of a nation­al state but as a part of the bor­der-cross­ing geo­gra­phies of the cloud.[6] The fact that the region­al elec­tric­i­ty grid had been con­nect­ed to glob­al cloud infra­struc­ture because of Facebook’s data cen­tre was depict­ed in the media as an actu­al relo­ca­tion of Nor­rbot­ten and Luleå from Sweden’s periph­ery to the cen­tre of the glob­al cloud. The Node Pole con­se­quent­ly became a wide­ly pro­mot­ed brand name for an imag­ined space of post-extrac­tive moder­ni­ty: a glo­cal geog­ra­phy of IT com­pe­tence and invest­ment possibilities.

On the one hand, these pro­mo­tion­al nar­ra­tives mere­ly per­pet­u­at­ed a hege­mon­ic per­cep­tion of the North­ern periph­ery that over cen­turies legit­imized Norrbotten’s exploita­tion by Sweden’s nation­al cen­tre. Such nar­ra­tives pre­sent­ed the scarce­ly pop­u­lat­ed but resource-rich region as an emp­ty space, free from local require­ments or frictions—a cor­nu­copia of nat­ur­al resources, tech­no­log­i­cal inno­va­tions, and secu­ri­ty arrange­ments offer­ing an abun­dance of cheap hydro pow­er and lim­it­less eco­nom­ic expan­sion far away from urban con­flicts and the pub­lic eye (Sör­lin, “Intro­duc­tion: Polar Exten­tions”). A rep­re­sen­ta­tive of Luleå’s neigh­bour­ing town, Boden, described its poten­tial as a data-stor­age site: “Boden is the most bor­ing place in the world for peo­ple but it is great for servers” (Inter­view, April 2015). On the oth­er hand, how­ev­er, these nar­ra­tives also empow­ered the region vis-à-vis the nation­al cen­tres by envi­sion­ing Nor­rbot­ten as an inde­pen­dent actor with­in a glob­al data econ­o­my and by stat­ing its prox­im­i­ty to major IT hubs such as Frank­furt or Lon­don, based on com­pa­ra­ble region­al fibre cable net­works, high-speed inter­net traf­fic, or the avail­abil­i­ty of inter­na­tion­al flight con­nec­tions. Cal­cu­la­tions and graphs devel­oped for brand­ing pur­pos­es com­pared the aver­age tem­per­a­ture in Luleå with tem­per­a­tures in Lon­don, Frank­furt, or New York in order to under­line the com­pet­i­tive­ness of Norrbotten’s cloud infra­struc­ture. The idea of region­al eman­ci­pa­tion from the state was also sup­port­ed by the expec­ta­tion that serv­er farms would be the first indus­try to make local use of nat­ur­al resources such as the Arc­tic cold and hydro ener­gy with­out hav­ing to export these resources to Sweden’s South, as had been the case with the tra­di­tion­al heavy indus­tries in this region.

In sum, “relo­cat­ing” Nor­rbot­ten and Luleå to the cloud and recon­fig­ur­ing these ter­ri­to­ries with­in the imag­ined geo­gra­phies of the Node Pole trans­formed the usu­al under­stand­ing of cen­tre and periph­ery and re-eval­u­at­ed the region and the city in geopo­lit­i­cal terms as an IT hub. The new indus­tri­al devel­op­ment changed the local actors’ under­stand­ing of the city’s ter­ri­to­ri­al­i­ty as flex­i­ble and deter­mined by spa­tial and tem­po­ral regimes of data oper­a­tion and not just by nation­al bor­ders (Rossiter). It was still to be deter­mined, how­ev­er, what the region and the city actu­al­ly had to offer on the glob­al IT market.

Fig­ure 7. An infor­ma­tion­al event for region­al busi­ness rep­re­sen­ta­tives pro­mot­ing par­tic­i­pa­tion in data cen­tre devel­op­ment, orga­nized by the Node Pole and the Luleå Busi­ness Agency, Luleå, Octo­ber 2014. Pho­tographed by author.

Pack­ag­ing a “Unique Sales Proposition”

While for region­al politi­cians and indi­vid­ual cit­i­zens cloud imple­men­ta­tion meant open­ing up new visions for Norrbotten’s future, for local and nation­al busi­ness devel­op­ers the cloud became pal­pa­ble first of all as a new mar­ket. In the con­text of the imag­i­nary space of the Node Pole, the Arc­tic cold—a for­mer sym­bol of Norrbotten’s and Luleå’s periph­er­al geo­graph­ic position—now appeared as a nat­ur­al resource. As Richard­son and Weszkalnys note, being social­ly pro­duced, nat­ur­al resources come into being not only as a result of tech­ni­cal inven­tions and phys­i­cal pro­duc­tions, but also through acts of epis­te­mo­log­i­cal and onto­log­i­cal cre­ativ­i­ty (129). To reshape the Arc­tic cold as a valu­able resource and new com­mod­i­ty, tech­nolo­gies of imag­i­na­tion were need­ed in order to re-con­cep­tu­al­ize the local cold in rela­tion to glob­al data flows.

In order to be com­mod­i­fied, the cold had to be pack­aged with oth­er local ben­e­fits into a spe­cif­ic region­al prod­uct. Pack­ag­ing, accord­ing to rep­re­sen­ta­tives of Luleå’s Busi­ness Agency, turned out to be par­tic­u­lar­ly chal­leng­ing as it demand­ed bring­ing things togeth­er that had pre­vi­ous­ly been seen as unre­lat­ed. As they lat­er would acknowl­edge, it was not clear yet what they were actu­al­ly sell­ing when first knock­ing at Facebook’s and Google’s doors dur­ing ear­ly vis­its to the Unit­ed States in 2009: “We under­stood that we had a fair chance to sell cheap elec­tric­i­ty and our cold cli­mate, and we had plen­ty of space and also intel­lec­tu­al capac­i­ty, so we tried to pack those things togeth­er. Admit­ted­ly, we were pack­ing some­thing which we didn’t own. We did not own the land, we did not own the cli­mate, and we nei­ther owned elec­tric­i­ty. But we still tried to come up with a work­able USP—a unique sales propo­si­tion” (Inter­view, Novem­ber 2014).

Even­tu­al­ly, busi­ness devel­op­ers came to pack­age the Arc­tic cold with Norrbotten’s cheap elec­tric­i­ty prices and Luleå’s excep­tion­al­ly sta­ble infra­struc­tur­al grid, fram­ing these and oth­er local advan­tages with the help of a pro­mo­tion­al nar­ra­tive that pic­tured Sweden’s North as a place of tech­no­log­i­cal moder­ni­ty (Sör­lin, “Rit­u­als and Resources of Nat­ur­al His­to­ry”). The abun­dance of “emp­ty space”—a sym­bol for Luleå’s and Norrbotten’s periph­er­al posi­tion and demo­graph­ic problems—was rein­ter­pret­ed as a guar­an­tee for unlim­it­ed and undis­turbed indus­tri­al growth. Using the Arc­tic cold for cool­ing glob­al data in this con­text not only meant cre­at­ing a tech­no­log­i­cal inno­va­tion but also a spe­cif­ic local ser­vice for “get­ting more from less.” The reval­u­a­tion of the Arc­tic cold thus enabled busi­ness devel­op­ers to real­ize their eco­nom­ic plans in attract­ing IT indus­try invest­ments and con­firmed Luleå’s cit­i­zens’ antic­i­pa­tion of Nor­rbot­ten enter­ing a post-extrac­tive era. The Node Pole’s pro­mo­tion­al nar­ra­tives depict­ed this process of resource-mak­ing and prod­uct-shap­ing as an affec­tive bod­i­ly expe­ri­ence. The above men­tioned video Jour­ney to the Node Pole, for instance, shows Luleå’s cit­i­zens enjoy­ing leisure activ­i­ties on the snow, stat­ing that while snow was for­mer­ly asso­ci­at­ed with freez­ing, snow and cold now are a good thing as they can be used for cool­ing the world’s data.

Pos­i­tive per­cep­tions of the new indus­try not only rest­ed upon an enthu­si­asm for tech­no­log­i­cal inno­va­tion but also on Luleå’s par­tic­u­lar his­to­ry. Over more than a cen­tu­ry, Luleå and Nor­rbot­ten were exploit­ed through resource extrac­tion and reg­u­lar­ly strug­gled with demo­graph­ic prob­lems caused by the vagaries of the nation­al steel and min­ing indus­tries. Since the begin­ning of the 20th cen­tu­ry, when dams for hydro ener­gy pro­duc­tion and the elec­tric­i­ty sup­ply of mines and steel mills were built at Lule riv­er with many locals being forced to move from their homes, nation­al indus­tri­al projects had a major impact on the region’s eco­nom­ic and demo­graph­ic devel­op­ment. One promi­nent exam­ple of this impact is the cur­rent­ly ongo­ing relo­ca­tion of the 18,000 inhab­i­tants of Norrbotten’s city of Kiruna due to the fact that the entire town is slow­ly plum­met­ing into a sink­hole caused by a for­mer iron ore mine.

In this his­tor­i­cal con­text, it is lit­tle sur­prise that the trans­for­ma­tion of the Arc­tic cold into a new nat­ur­al resource—not meant for export, but for attract­ing investment—was greet­ed with enthu­si­asm by both local entre­pre­neurs and cit­i­zens. The cloud industry’s enor­mous elec­tric­i­ty and water needs[7] and waste heat pol­lu­tion were bare­ly dis­cussed and seemed harm­less com­pared to the envi­ron­men­tal dam­age caused by tra­di­tion­al extrac­tive indus­tries: as a local jour­nal­ist explains, “this city is accus­tomed to pol­lu­tion and envi­ron­men­tal dam­ages through the SSAB steel mill. Note that Luleå’s most pop­u­lar liv­ing area is locat­ed just in front of the mill. And in the case of the data cen­tre, there is not even a chim­ney on top of it, and it’s just about elec­tric­i­ty con­sump­tion” (Inter­view, Novem­ber 2014). The mak­ing of new resources is a rela­tion­al process involv­ing dif­fer­ent actors as stake­hold­ers. As Richard­son and Wesza­kalnys observe, new resources can gen­er­ate finan­cial prof­its or loss­es, strength­en or weak­en nation states, cause envi­ron­men­tal dam­age and change local modes of liv­ing (10). Shap­ing the Arc­tic cold as a new resource of the dig­i­tal era and pack­ag­ing that resource into a unique local prod­uct empow­ered the Nor­rbot­ten region as an inde­pen­dent eco­nom­ic actor by relat­ing the Arc­tic cold to glob­al data streams. At the same time, how­ev­er, it obscured the envi­ron­men­tal effects of the cloud and the dif­fi­cul­ty of actu­al­ly prov­ing that the cloud had pos­i­tive future effects—a claim that could be eas­i­ly con­test­ed giv­en the lack of new jobs for locals, for instance. The econ­o­my of antic­i­pa­tion was, how­ev­er, cer­tain­ly ben­e­fi­cial for Face­book, since it mobi­lized the local com­mu­ni­ty around the promise of change and an imag­ined post-extrac­tive future, mak­ing the most effi­cient, fast, and fric­tion­less imple­men­ta­tion of the cloud possible.


This arti­cle has ana­lyzed the imple­men­ta­tion of the cloud in Sweden’s North as a form of “infrastructuring”—a process of relat­ing, antic­i­pat­ing, net­work­ing, and nego­ti­at­ing among tech­nolo­gies, actors, insti­tu­tions, and stocks of knowl­edge through dif­fer­ent geo­graph­i­cal loca­tions and social con­texts. The arti­cle has traced the diverse ways that cloud infra­struc­tur­al forms dis­con­nect from their tech­ni­cal func­tion­al­i­ty and link to mul­ti­ply locat­ed actors and their con­cerns enabling social change (see Larkin). The mak­ing of cloud infra­struc­ture was described against the back­drop of a glob­al econ­o­my of antic­i­pa­tion in rela­tion to which the past and present of the involved com­mu­ni­ties were nego­ti­at­ed and formed. Infra­struc­tur­ing here is not a lin­ear top-down move­ment, but a his­tor­i­cal­ly and cul­tur­al­ly sit­u­at­ed and fric­tion-laden process, facil­i­tat­ed by means of imag­i­na­tion and antic­i­pa­tion. Tech­nolo­gies of imagination—understood as knowl­edge pro­duc­tion prac­tices of bring­ing togeth­er seem­ing­ly unre­lat­ed things—were iden­ti­fied as an inher­ent part of infra­struc­tur­ing. This ethno­graph­ic per­spec­tive allows the seem­ing­ly vir­tu­al glob­al data flows to be recon­nect­ed with their mate­r­i­al and polit­i­cal chan­nels, and ren­ders their mul­ti­ple local entan­gle­ments vis­i­ble. It shows how the cloud’s mate­r­i­al forms and modes of vis­i­bil­i­ty are con­test­ed in cul­tur­al­ly spe­cif­ic ways with social and envi­ron­men­tal effects (see Tsing).

In con­trast to Facebook’s cor­po­rate strate­gies of obscur­ing its infra­struc­tur­al mate­ri­al­i­ties, it was clear­ly nev­er in the inter­est of the Swedish state, the city, or the region of Nor­rbot­ten to hide the social network’s Luleå infra­struc­ture. Rather, local actors were inter­est­ed in pub­licly claim­ing Facebook’s local pres­ence and inter­pret­ing it as a mean­ing­ful and last­ing form of belong­ing. By con­nect­ing Facebook’s infra­struc­ture to the nation­al ener­gy grid and by relat­ing the data cen­tre project to the region­al inter­ests and aspi­ra­tions of a post-extrac­tive future, the cloud was inte­grat­ed into nation­al ter­ri­to­ries and local nar­ra­tives. Cloud tech­nolo­gies pro­vid­ed Luleå not only with tech­no­log­i­cal con­nec­tions to a bor­der-cross­ing IT econ­o­my, but also with new nar­ra­tives and spa­tial frames that allowed a reimag­in­ing of the future of the city and the region beyond the Swedish state in rela­tion to the glo­cal geo­gra­phies of data pro­cess­ing. As a result, the city of Luleå and the region of Nor­rbot­ten were reframed as a hub of IT com­pe­tence with­in the imag­i­nary space of the Node Pole. Such future sce­nar­ios were, how­ev­er, not sim­ply deter­mined by cor­po­rate visions of the cloud or by region­al actors’ tech­no­log­i­cal dreams, but root­ed in par­tic­u­lar local expe­ri­ences and aspi­ra­tions that had devel­oped his­tor­i­cal­ly with regard to Luleå as a geo­graph­i­cal­ly periph­er­al indus­try town and Nor­rbot­ten as a tra­di­tion­al­ly exploit­ed region.

Fol­low­ing local process­es of infra­struc­tur­ing sheds light on the rela­tion between the urban cen­tres of dig­i­tal cap­i­tal­ism and its infra­struc­tur­al periph­eries, between vir­tu­al data flows and their polit­i­cal and mate­r­i­al chan­nels, and between transna­tion­al IT com­pa­nies and local com­mu­ni­ties. The rela­tion that I have described here as “inti­ma­cy at a dis­tance” demon­strates the mutu­al depen­den­cies between the vir­tu­al and mate­r­i­al and the local and glob­al dimen­sions of the cloud, as well as between the dif­fer­ent stake­hold­ers involved in process­es of infra­struc­tur­ing. This per­spec­tive also demon­strates the uneven dis­tri­b­u­tion of vis­i­bil­i­ty, agency, and pow­er with­in these glo­cal geo­gra­phies of the cloud. Although the intro­duc­tion of the data cen­tre indus­try and cloud infra­struc­ture in Luleå empow­ers this periph­er­al city vis-à-vis Sweden’s nation­al cen­tres, by turn­ing Luleå into an inde­pen­dent actor with­in a glob­al data econ­o­my the costs of this empow­er­ment are high and its future uncertain.

Local author­i­ties, experts, and even reg­u­lar cit­i­zens in Luleå unite in their efforts to imag­ine a post-extrac­tive future. How­ev­er, tech­nolo­gies of imag­i­na­tion that recon­fig­ure local ter­ri­to­ries, rede­fine iden­ti­ties, and shape new sub­jec­tiv­i­ties also obscure the cloud’s actu­al extrac­tive char­ac­ter and its prob­lem­at­ic social and envi­ron­men­tal effects, includ­ing the data centre’s enor­mous ener­gy needs, waste heat, or its dis­ap­point­ing effects on the local labour mar­ket. Imag­in­ings and antic­i­pa­tions of a post-extrac­tive future are main­ly based on infra­struc­tur­al promis­es that are “almost-but-not-quite-there.” There­fore, it is still ques­tion­able if locat­ing the cloud is actu­al­ly advanc­ing Luleå and Nor­rbot­ten from the nation­al periph­ery to the cen­tre of the glo­cal cloud or if it is just turn­ing them into periph­eries of the dig­i­tal econ­o­my: places to which fric­tion-laden mate­ri­al­i­ties and prob­lem­at­ic social and nat­ur­al impacts are being outsourced.

Works Cited

Blum, Andrew. Tubes: A Jour­ney to the Cen­ter of the Inter­net. Harp­er Collins, 2012.

Bowk­er, Geof­frey C., Karen Bak­er, Flo­rence Millerand, David Ribes. “Toward Infor­ma­tion Infra­struc­ture Stud­ies: Ways of Know­ing in a Net­worked Envi­ron­ment.” Inter­na­tion­al Hand­book of Inter­net Research, edit­ed by J. Hun­singer et al., Springer Science+Business Media B.V., 2010, pp. 97–117.

Cross, Jamie. Dream Zones. Antic­i­pat­ing Cap­i­tal­ism and Devel­op­ment in India. Plu­to Press, 2014.

Cubitt, Sean, Robert Has­san, and Ingrid Volk­mer. “Does Cloud Com­put­ing Have a Sil­ver Lin­ing?” Media, Cul­ture & Soci­ety, vol. 33, 2011, pp. 149–158.

Deleuze Gilles. “Post­script on the Soci­eties of Con­trol”. Octo­ber, vol. 59, 1992, pp. 3–7.

Dour­ish, Paul, and Genevieve Bell. Div­ing a Dig­i­tal Future: Mess and Mythol­o­gy in Ubiq­ui­tous Com­put­ing. MIT Press, 2011.

Granberg, Anders. Effek­ten av Face­books etab­ler­ing för Luleå Sci­ence Park som region­al nod. Luleå: Luleå Sci­ence Park, 2014, http://​doc​play​er​.se/​3​2​3​0​8​-​E​f​f​e​k​t​e​n​-​a​v​-​f​a​c​e​b​o​o​k​s​-​e​t​a​b​l​e​r​i​n​g​-​f​o​r​-​l​u​l​e​a​-​s​c​i​e​n​c​e​-​p​a​r​k​-​s​o​m​-​r​e​g​i​o​n​a​l​-​n​o​d​-​f​o​t​o​-​c​a​s​c​a​d​e​-​c​r​e​a​t​i​v​e​-​m​e​d​i​a​.​h​tml. Accessed 20 Feb­ru­ary 2017.

Gov­ern­ment Offices of Swe­den. ICT for Every­one: A Dig­i­tal Agen­da for Swe­den, 2011, http://​www​.gov​ern​ment​.se/​s​b​/​d​/​5​7​4​/​a​/​1​8​1​914. Accessed 20 Feb­ru­ary 2017.

Har­vey, Pene­lope. “The Topo­log­i­cal Qual­i­ty of Infra­struc­tur­al Rela­tion: An Ethno­graph­ic Approach.” The­o­ry, Cul­ture & Soci­ety, vol. 29, no. 4/5, 2012, pp. 76–92.

Hogan, Mél. “Face­book Data Stor­age cen­ter as the Archive’s Under­bel­ly.” Tele­vi­sion & New Media, vol. 16, no 1, 2015, pp. 3–18.

Holt, Jen­nif­fer, and Patrick Von­der­au. “Where the Inter­net Lives: Data Cen­ters as Cloud Infra­struc­ture.” Sig­nal Traf­fic. Crit­i­cal Stud­ies in Media Infra­struc­tures, edit­ed by Lisa Parks and Nicole Starosiel­s­ki, Uni­ver­si­ty of Illi­nois Press, 2014, pp. 88–114.

Hu, Tung-Hui. A Pre­his­to­ry of the Cloud. MIT Press, 2015.

Jensen, Cas­par Brun, Brit Ross Winthereik. Mon­i­tor­ing Move­ments in Devel­op­ment Aid: Recur­sive Part­ner­ships and Infra­struc­tures. MIT Press, 2013.

Larkin, Bri­an. “The Pol­i­tics and Poet­ics of Infra­struc­ture.” Annu­al Review of Anthro­pol­o­gy, vol. 43, 2013, pp. 327–343.

Mrázek, Rudolf R. Engi­neers of Hap­py Land: Tech­nol­o­gy and Nation­al­ism in a Colony. Prince­ton Uni­ver­si­ty Press, 2002.

Niewöh­n­er, Jörg. “Anthro­pol­o­gy of Infra­struc­tures of Soci­ety.” Inter­na­tion­al Ency­clo­pe­dia of the Social & Behav­ioral Sci­ences, vol. 12, 2015, pp. 119–125. SienceDi­rekt, doi: http://​www​.sci​encedi​rect​.com/​s​c​i​e​n​c​e​/​r​e​f​e​r​e​n​c​e​w​o​r​k​s​/​9​7​8​0​0​8​0​9​7​0​875.

Parks, Lisa. “Tech­nos­trug­gles and the Satel­lite Dish: A Pop­ulist Approach to Infra­struc­ture.” Cul­tur­al Tech­nolo­gies: The Shap­ing of Cul­ture in Media and Soci­ety, edit­ed by Göran Bolin, Rout­ledge, 2012, pp. 64–84.

Richard­son, Tanya, Gisa Weszkalnys. “Intro­duc­tion: Resource Mate­ri­al­i­ties”. Anthro­po­log­i­cal Quar­ter­ly, vol. 87, no.1, 2014, pp. 5–30.

Rossiter, Ned. Soft­ware, Infra­struc­ture, Labor: A Media The­o­ry of Logis­ti­cal Night­mares. Rout­ledge, 2016.

Sneath, David, Mar­tin Hol­braad, Morten Axel Ped­er­sen. “Tech­nolo­gies of the Imag­i­na­tion: Intro­duc­tion“. Eth­nos, vol. 74, no.1, 2009, pp. 5–30.

Sör­lin, Sverk­er. “Rit­u­als and Resources of Nat­ur­al His­to­ry: The North and the Arc­tic in Swedish Sci­en­tif­ic Nation­al­ism.” Nar­rat­ing the Arc­tic: A Cul­tur­al His­to­ry of Nordic Sci­en­tif­ic Prac­tices, edit­ed by Michael Bra­vo and Sverk­er Sör­lin. Sci­ence His­to­ry Pub­li­ca­tions, 2002, pp. 73–124.

---. “Intro­duc­tion: Polar Extentions—Nordic States and Their Polar Strate­gies.” Sci­ence, Geopol­i­tics and Cul­ture in the Polar Region: Nor­den Beyond Bor­ders, edit­ed by Sverk­er Sör­lin. Ash­gate Pub­lish­ing, 2013, pp. 1–19.

Star, Susan L. “The Ethnog­ra­phy of Infra­struc­ture.” Amer­i­can Behav­ioral Sci­en­tist, vol. 43, no. 3, 1999, pp. 377–391.

Starosiel­s­ki, Nicole. The Under­sea Net­work. Duke Uni­ver­si­ty Press, 2015.

Vol­vo Cars. Jour­ney to The Node Pole. https://​www​.youtube​.com/​r​e​s​u​l​t​s​?​s​e​a​r​c​h​_​q​u​e​r​y​=​j​o​u​r​n​e​y​+​t​o​+​t​h​e​+​n​o​d​e​+​p​ole.   Film. Accessed 20 Feb­ru­ary 2017.

Tsing, Anna. “The Glob­al Sit­u­a­tion.” Cul­tur­al Anthro­pol­o­gy, vol. 15, no.3, 2000, pp. 327–360.

Var­nelis, Kazys. “Eyes That Do Not See: Track­ing the Self in the Age of the Data Cen­ter.” Havard Design Mag­a­zine, vol. 38, 2014, http://​www​.har​vard​de​sign​magazine​.org/​i​s​s​u​e​s​/​3​8​/​e​y​e​s​-​t​h​a​t​-​d​o​-​n​o​t​-​s​e​e​-​t​r​a​c​k​i​n​g​t​h​e​-​s​e​l​f​-​i​n​-​t​h​e​-​a​g​e​-​o​f​-​t​h​e​-​d​a​t​a​-​c​e​n​ter. Accessed 20 Feb­ru­aty 2017.

Image Notes

Fig­ure 1: Luleå’s munic­i­pal com­mis­sion­er Karl Petersen, fea­tured in a video pro­mo­tion for Vol­vo and the Node Pole www​.youtube​.com/​w​a​t​c​h​?​v​=​Y​w​q​t​B​J​M​Z​JIU. Accessed 20 Feb­ru­ary 2017.

Fig­ure 2: Face­book thumb in the lunch room of Luleå Busi­ness Agency, pho­tographed by author, April 2014.

Fig­ure 3: Facebook’s first data cen­tre build­ing in Luleå, pho­tographed by author, April 2014.

Fig­ure 4: Entrance area of Facebook’s sec­ond data cen­tre build­ing, Datavä­gen 15 (Data Street 15), pho­tographed by author, Sep­tem­ber 2015.

Fig­ure 5: Facebook’s Swedish cloud, seen from above. Build­ings to the right include stu­dent dor­mi­to­ries and Luleå’s Sci­ence Park. Source: Genomförandebeskrivning/The Descrip­tion of Imple­men­ta­tion, Luleå Munic­i­pal­i­ty 2011.

Fig­ure 6: This draw­ing shows how Facebook’s data cen­tre (the thin line above the trees) was expect­ed to be vis­i­ble after com­ple­tion, thus illus­trat­ing the cloud’s pecu­liar infra­struc­tur­al (in)visibility.

Source: Miljökonsekvensbeskrivning/Description of Envi­ron­men­tal Con­se­quences, Luleå Munic­i­pal­i­ty 2011.

Fig­ure 7: An infor­ma­tion­al event for region­al busi­ness rep­re­sen­ta­tives pro­mot­ing par­tic­i­pa­tion in data cen­tre devel­op­ment, orga­nized by the Node Pole and Luleå Busi­ness Agency, Luleå, Octo­ber 2014. Pho­tographed by author.


[1] Luleå is locat­ed about 100 kilo­me­tres south of the Arc­tic cir­cle. The city has 45,000 inhab­i­tants and is the cap­i­tal of Nor­rbot­ten, the most north­ern Swedish province. Nor­rbot­ten is the infra­struc­tural­ly strongest region in the coun­try, tra­di­tion­al­ly dom­i­nat­ed by steel and for­est indus­tries and min­ing. Over the last five to ten years, Luleå expe­ri­enced strong eco­nom­ic and demo­graph­ic growth. How­ev­er, seen from a his­tor­i­cal per­spec­tive, the city and the region reg­u­lar­ly strug­gled with demo­graph­ic prob­lems that arose due to its periph­er­al geo­graph­ic posi­tion and harsh cli­mate as well as the ups and downs of the heavy indus­tries. Facebook’s deci­sion to locate its data cen­tre in Luleå was announced in 2011 and in 2013 the first data cen­tre start­ed oper­at­ing. As of late 2016, a sec­ond data cen­tre build­ing is still under con­struc­tion and a third build­ing is in the plan­ning phase.

[2] Face­book has reject­ed my inquiries for meet­ings and did not allow the sub­con­tract­ed com­pa­nies to give inter­views about their involve­ment in the Luleå Data Cen­ter project. My meet­ings with Face­book rep­re­sen­ta­tives at pub­lic events were, accord­ing­ly, not infor­ma­tive since they did not pro­vide any oth­er infor­ma­tion than that which was cov­ered in the media. How­ev­er, my inter­est was focused on region­al actors and their experiences.

[3] It is ques­tion­able if this indeed would be the case. Facebook’s infra­struc­ture is of course redun­dant, but tech­ni­cal prob­lems of such an extent would cause huge costs and might harm the company’s image.

[4] This ear­ly project phase cov­ers the time peri­od of 2009 to 2011. Munic­i­pal activ­i­ties dur­ing this peri­od includ­ed the sell­ing of land and the infra­struc­tur­al prepa­ra­tion of the con­struc­tion area; issu­ing envi­ron­men­tal and build­ing per­mis­sions; prepar­ing a detailed city plan; and pre­sent­ing the project to the pub­lic. While Facebook’s plans were kept entire­ly secret in the begin­ning of that peri­od, munic­i­pal experts lat­er learned that the client did not want to reveal the firm’s name even after com­ple­tion. How­ev­er, local media spread the news that the client was actu­al­ly Face­book, con­firmed in 2011 when Face­book offi­cial­ly announced its deci­sion to build the data cen­tre in Luleå.

[5]  After com­ple­tion, this data cen­tre will con­tain three serv­er halls, 28,000 square metres each, requir­ing 120 MW of pow­er. The data cen­tre imple­men­ta­tion costs amount to 800 mil­lion Swedish Kro­na (SEK).

[6] Estab­lished in the course of the Face­book data cen­tre project by Luleå’s munic­i­pal­i­ty and pri­vate busi­ness, the Node Pole rep­re­sent­ed first of all a region­al brand. In 2016, the com­pa­ny was bought by the Swedish state-owned ener­gy com­pa­ny Vat­ten­fall and the North­ern Swedish ener­gy com­pa­ny Skelleft­eå Kraft, which advanced the Node Pole to a nation­al brand.

[7] Facebook’s data cen­tre in Luleå needs one ter­rawatt of elec­tric­i­ty per year, which is com­pa­ra­ble to the ener­gy con­sump­tion of Sweden’s sixth largest munic­i­pal­i­ty of 146,000 inhabitants.

This arti­cle is licensed under a  Cre­ative Com­mons 4.0 Inter­na­tion­al License although cer­tain works ref­er­enced here­in may be sep­a­rate­ly licensed, or the author has exer­cised their right to fair deal­ing under the Cana­di­an Copy­right Act.