S c d Pa #A IR2021:
T 22 d A a C c
A c a I R a c
Vi al E e / 13-16 Oc 2021
ALGORITHMIC IMAGINARIES IN THE MAKING: BRAZILIAN
UBERTUBERS ENCOUNTERS WITH SURGE PRICING ALGORITHMS
Ana Guerra
Uni ersidade Federal de Minas Gerais Carlos Frederico de Brito D Andr a Uni ersidade Federal de Minas Gerais
I d c
Surge pricing (SP) is a central mediator of Uber dri er s e er da e perience. It can be described as an algorithmicall dri en mechanism that uses price adjustments as financial incenti es to redistribute the orkforce on a territor (Rosenblat; Stark, 2016).
SP algorithms are presented b Uber as entities capable of measuring, translating, and responding to the market conditions roughl summari ed as dri ers suppl and riders demand. A heatmap ith colored shades sho n in the Uber Dri er app orients the real-time decisions made b the platform orkers (Grohmann; Qiu, 2020).
This algorithmic go ernance is routinel questioned and appropriated b dri ers.
Although SP ma trigger feelings of e citement, it is also often associated ith frustration and deception, for e ample hen SP disappears as soon as the reach high-demand areas. Recei ing requests from riders outside surge areas or the non-pa ment of the e tra alue are some of the issues faced and discussed b Uber dri ers.
This paper argues that dri ers e er da practices and efforts to understand and contest surge pricing are producti e of algorithmic imaginaries and alid forms of kno ledge (Bucher, 2018) that, based on personal and shared e periences, inform ho the orient their actions to ards Uber s algorithmic s stems. We approach this b mapping and anal ing ho popular Bra ilian Uber dri ers discuss surge pricing on their YouTube channels. Our focus is understanding ho these Ubertubers dri ers ho built their YouTube presence based on their identit as Uber dri ers e plain and demonstrate ho SP orks, producing and sharing their o n isual inscriptions (Latour, 1999).
Inspired b Critical Algorithmic Studies,this paper seeks to contribute to the
understanding of the situated kno ledge triggered b algorithm-oriented e periences.
Suggested Citation (APA): Guerra, A., and de Brito D’Andréa, C.F., (2021, October). Algorithmic Imaginaries in the Making: Brazilian Ubertubers Encounters with Surge Pricing Algorithms. Paper presented at AoIR 2021: The 22nd Annual Conference of the Association of Internet Researchers. Virtual Event: AoIR. Retrieved from http://
spir.aoir.org.
Such e periences are part of hat constitutes algorithms ontogenic nature (Kitchin, 2017), as the are continuousl edited and updated, as ell as transformed b user s practices. B in estigating the algorithmic imaginaries of SP, e are interested in "the
a in hich people imagine, percei e and e perience algorithms and hat these imaginations make possible" (Bucher, 2017, p. 31), taking into account the producti e capacities of people s encounters ith these sociotechnical s stems1.
Ube D e Y T be
The rele ance of ideos produced b Ubertubers has been pre iousl pointed out b Chan (2019), ho argues that, b positioning themsel es as sociotechnical e perts, these dri ers assume a pedagogical discourse that promises to unpack Uber s s stems to optimi e dri er's earnings and producti it . Thus, Ubertubers are e pected to be
self-entrepreneurs not onl as dri ers but also in ideo production and audience engagement.
In pre ious research (Guerra, 2018), e obser ed ho o er 100 Bra ilian Ubertubers organi e a shared kno ledge based on e er da practices. Personal stories about
orking as an Uber dri er, e planations about changes and updates in the app, ad ice about ho to a oid robberies, and calls for collecti e political action are the most common issues in their ideos. Thus, through a performance of e pertise, Ubertubers s stemati e dri ers algorithmic imaginaries and make them accessible to a ider audience.
F d g
For this stud e anal ed 59 ideos posted b the se en most popular Bra ilian UberTubers (all ith more than 50,000 subscribers b the time of channel selection). All of them are men and li ing in the South or Southeast of Bra il. The time frame of this data collection October/2016 to October/2020 allo ed us to follo ho dri ers perception of SP and SP itself changed o er time.
After atching and transcribing the ideos, e could identif three main topics:
(a)Wha i ge ici g a d ha d e i d ?To ans er this broader question, Ubertubers describe their e perience ith SP as a mi of e citement and frustration, often depicting it as an illusion or a lure . Their anal sis also refers to SP s isual form (the reddish heat map) as a bleeding screen or a screen on fire . Man of the ideos are dedicated to e plaining, through inscriptions produced b them (screenshots and hand dra ings) ho SP orks and h it often leads to frustration (Fig. 1).
1This stud is part of Ana Guerra s academic master s thesis, hich in estigates Uber s surge pricing through technographic research methods (Bucher, 2018). This research engaged in the tracing of documents and inscriptions that cluster around SP and account both for the perspecti e of Uber and dri ers.
Figure 1: an Ubertuber uses hand dra ings to e plain ho SP orks Source: screenshot produced b the authors
(b) Tactics tobe efi f m ge ici g Ubertubers teach their audience ho to ork i hsurge pricing. Uber hides SP once dri ers accept a ride, forcing dri ers to ork in the dark , as one UberTuber puts it, but the create tactics to make SP isible before the accept the ne t trip. As a ne ride request usuall appears before the current ride is completed, the ad ice is to go offline during rides. Alternati el , one Ubertuber chooses to ork ith t o separate smartphones to permanentl track SP. Based on temporal and spatial patterns (e.g. rush hour and the cit s center-peripher mobilit d namics), this dri er de eloped his o n classification of SP ariations.
(c)Alg i hmic Lab a a lab a SP is constantl updated and arious ersions are tested on different dri ers. Ubertubers continuousl document changes on ho SP performs be it ho their e tra earnings are calculated or ho it is displa ed on the app s interface. Man of the changes are not officiall announced b Uber and are onl percei ed on dri er s e er da encounters ith the app. This documentation is done in a collecti e effort, ith Ubertubers often using screenshots and accounts pro ided b other dri ers.
Through the anal sis of the ideos, e ere able to approach ho dri ers e perience and imagine surge pricing algorithms, and ho their appropriations negotiate their orientation to ards this sociotechnical s stem. Dri er s situated encounters ith SP algorithms generate practical kno ledge fueled b producti e affects. Through the engagement ith their audiences, Ubertubers collecti i e these e periences and imaginations, potentiall transforming ho arious other dri ers incorporate surge pricing into their o n tactics and dail routines.
Finall , Dri ers channels offer us the opportunit to learn from these orkers about ho SP algorithms act, hat the supposedl do and ho the transform labor on a
micropolitical le el. Another contribution to the field is the recognition of this audio isual material as a singular archi e of SP s man ersions and facets, gi ing us access to information on algorithms e might ha e not been able to find other ise.
Refe e ce
Bucher, T. (2018).If he : Alg i hmic P e a d P li ic . O ford Uni ersit Press.
Bucher, T. (2017). The algorithmic imaginar : e ploring the ordinar affects of Facebook algorithms.I f ma i , C mm ica i a d S cie , 20(1), 30 44.
Chan, N. K. (2019). Becoming an e pert in dri ing for Uber : Uber dri er/bloggers performance of e pertise and self-presentation on YouTube .Ne Media a d S cie , 2019.
Guerra, A. (2018). Pla af mi a d abalh : m e d da ede em a
abalh d m i a Ube. (Undergraduate thesis). Uni ersidade Federal de Minas Gerais, Belo Hori onte, Bra il.
Guerra, A. (2021). I f ae a , a a i a e imagi i alg mic : ec g afa d e di mic da Ube (Master s thesis). Uni ersidade Federal de Minas Gerais, Belo Hori onte, Bra il.
Grohmann, R.; Qiu, J. (2020). Conte tuali ing platform labor.C acam , 39(1).
Kitchin, R. (2017). Thinking criticall about and researching algorithms. Inf ma i , C mm ica i & S cie , 20(1),14 29.
Latour, B. (1999).Pa d a H e: E a he Reali f Scie ce S die (1st edition). Har ard Uni ersit Press.
Rosenblat, A., Stark L. (2016). Algorithmic Labor and Information As mmetries: A Case Stud of Uber s Dri ers.I e a i al J al f C mm ica i , 10,3758 84.
