https://d2ng68cf36kxsk.cloudfront.net/store/public/post/903/post_images/Slide314.JPG">

Uber starts knowing who you are, where you are, and what time you ask the question: these points of reference and position are drawn mechanically from context – “asignifying semiotics” Félix Guattari calls these.^§AS1.3.2b But “where to?” is a question that no amount of artificial intelligence can figure, unless, knowing your ways, it’s somewhere you have been before, in which case, Uber will suggest that as a possibility.

I can see Uber cars, like little cockroaches crawling over the map of Camden Town on my phone. A driver accepts my request. I guess the Uber algorithm has calculated he was closest and available.

Thirty-nine minutes to the airport, price £38.86, the car four minutes away, now three minutes, now two. I see the little icon inching along a tangle of one-way streets, so phone the driver to say which corner I am on. I have his number plate, the make, model, and color of his car, and a picture of that kind of car. But I call to be sure he knows where I am. It’s a personal thing these days, to give a cellphone number, and intrusive to call with voice, but the relationship we are about to have is this personal. Phoning has become all-the-more personal since the rise of texting; voice calls have become rarer out of deference for other people’s right, as Richard Kern says, to manage the terms of their participation in a conversation.²⁵⁴

Abdullahi has been driving with Uber for one and a half years, has a 4.7 rating, and speaks English, Arabic, and Somali – so the app tells me. Thirty people have checked the diamond icon, “Excellent service,” six the “Great Chat” icon. But he has only three recommendations for “Great Route Choice,” an odd thing because there is no choice – the driver can’t take routes other than the one prescribed by Uber. Which incidentally means that he can’t pull that old taxi driver dodge, taking a longer route for a bigger fare. Uber micromanages the route, offering at least this peace of mind to the rider, mitigating in this particular way the generalized anxieties that are intrinsic to the differential interests of sellers and buyers.^§AS2.4.1

GPS^§AS1.3.2b also lowers the barriers to access that were present in the old taxi industry. In London, taxi drivers have to pass a grueling map test, known in the trade as “the Knowledge.”²⁵⁵ GPS has not only made this kind of knowledge redundant; the scale of empirical detail captured in GPS, now that we have it, has made any kinds of memorized spatial knowledge inadequate.

And another nice thing about Uber: If you are in a country where you don’t speak the language, no speech is needed, because the driver’s textual and visual interface is in one language and yours in another. Conversational turn-taking works because its grammar is multimodal (place, time, map) and multilingual (a translation interface). The mechanics of conversation have shifted from speech in natural language, to a shared ontology^§MS3.3a of streets, persons, and the scripted universal dialogue of being driven that operates in a semantic architecture that is beyond natural language. If there is text on my screen, its natural language has become circumstantial and arbitrary in a way that it never was when hailing a cab and having to tell the driver in speech where you wanted to go.

Sometimes you even wish that you didn’t belong to a shared speech community, when the Uber driver’s life situation – their three jobs and their medical bills with no insurance – are just too grueling, or their social opinions too unnerving. But Uber even steps in on the rider’s behalf to regulate this, because when the ride is over, “Conversation” is one of the rating buttons.

My conversation with Abdullahi brings our transitory relationship into the context of his wider life. Uber takes 25 percent of the fare, a lot I thought, just for a piece of software which manages the conversation that coordinates my riding with his driving. Better than being a mini-cab driver, he says. He was this before, but now he’s more in control of his own life. No longer a worker, he’s a small business person now. In the “gig economy,” even the minimal labor rights of traditional employees are cast asunder.

No, it’s more complicated, he’s with forty other Somali drivers in an “Ayuuto” collective. Each contributes £100 per week, and every couple of weeks the collective buys a new car for one of its participants. “Ayuuto”: Somalized version of “help.” There’s a hashtag for it on Twitter, pointing to moments in a socialism of sorts, albeit with Somalian characteristics. This is how, to use anthropologist David Graeber’s formulation,^§AS2.4.1 two entirely different “systems of moral accounting,” Ayuuto and Uber, are held together in uneasy tension, each with its own characteristic patterns of multimodal conversation.²⁵⁶

Abdullahi drops me off at the airport. I’ve been writing with my thumbs, and now I pay with my fingerprint. On a newer phone it would have been face recognition. Or, as we would say in this grammar, this turn in the conversation is a viscerally embodied communication of my personhood, singularly instantiated bodily characteristics that can’t be faked.

But before the conversation comes to an end, I have to rate Abdullahi, and he has to rate me. If I get ratings that are bad enough – like, I am drunk and have thrown up in his car – I will get banned from Uber. There is a whole genre of reputation software to keep track of patterns in behavior.²⁵⁷ And rating the driver: no stars to five stars; and suggested improvements could be “professionalism,” “driving,” “navigation,” “car quality,” “music,” “pickup.” Uber’s algorithms are listening, of course, and anticipating possible follow-up conversations where the company may need to arbitrate: “I was involved in an accident >,” “I would like a refund >,” “I lost an item >,” “my driver was unprofessional >.” The range of predictable and allowable conversations has been pre-scripted.

But mostly, the experience is at this point reduced to a number, a rating, and in a culture of obsessive and overvalued calculation, this becomes the crux of the conversation. After the rating there is no process for dispute resolution, and if fault is admitted, there is no way to make amends. There is nothing more to discuss. A ride-sharing driver in California tried to educate his passengers by putting up in his car a picture of a four-out-of-five star rating, with the caption, “This driver sucks, fire him slowly; it does not mean ‘average’ or above ‘average.’”²⁵⁸

Far from solving “market asymmetries,” say a trio of computational economists, mutual rating systems tend to “reputation inflation,” motivated by mutual fear grounded precisely in these asymmetries of interest.²⁵⁹ Mutual reporting is mutual policing; mutual trust is the regulation of mistrust; sharing is an alibi for the generalized antagonisms of the market; and the interpersonal is colonized by coordinating software that serves private profit.

“Sharing” is the way it feels when you are in a private person’s car, and you are asked to make life-impinging judgments of each other’s demeanor. But “renting” is what this conversation really is, and Uber are the agents, stitching the conversation together. By this conversational framing, Uber is also offloading onto customers what in another setting would have been a corporate responsibility. All this, scripted into the grammar of multimodal conversation.

Ultimately, Uber controls this conversation, and it is a corollary that they who control the conversation stand to profit more.

Jackson Cunningham got into trouble with another of the sharing behemoths, the home-sharing app Airbnb, driven by a conversational structure and business model that is similar to Uber’s. He dared to comment on a “host” in a Google listing, outside of the Airbnb rating and commenting framework. Or at least, he guesses this was where he must have overstepped the conversational mark, because Airbnb would provide no explanation why he had been banned.

Dear Jackson,

We regret to inform you that we’ll be unable to support your account moving forward, and have exercised our discretion under our Terms of Service to disable your account(s). This decision is irreversible and will affect any duplicated or future accounts.

Please understand that we are not obligated to provide an explanation for the action taken against your account. Furthermore, we are not liable to you in any way with respect to disabling or canceling your account. Airbnb reserves the right to make the final determination with respect to such matters, and this decision will not be reversed.²⁶⁰

End of conversation.

This is how Jackson discovered that, behind the dialogism of the sharing economy, there was a new dictatorship of speech, where a handful of software companies dominate driving here and staying there. Airbnb and Uber not only facilitate the conversation; they own the conversation. The only things that have changed since the Cazdens’ experience of McCarthyism^§AS1.4.1c is the grammar of control and the identities of the controllers.

---

²⁵³ David Crystal, 2008, Txting: The Gr8 Db8, Oxford: Oxford University Press.

²⁵⁴ Richard Kern, 2015, Language, Literacy, and Technology, Cambridge: Cambridge University Press, p. 248.

²⁵⁵ Eleanor A. Maguire, David G. Gadian, Ingrid S. Johnsrude, Catriona D. Good, John Ashburner, Richard S.J. Frackowiak, and Christopher D. Frith, 2000, “Navigation-Related Structural Change in the Hippocampi of Taxi Drivers,” Proceedings of the National Academy of Sciences 97(8):4398–403.

²⁵⁶ David Graeber, 2011, Debt: The First 5,000 Years, Brooklyn, NY: Melville House, p. 114.

²⁵⁷ F. Randall Farmer and Bryce Glass, 2010, Web Reputation Systems, Sebastopol, CA: O’Reilly; Phillip J. Windley and Devlin Daley, 2006, “A Framework for Building Reputation Systems” (www.windley.com/essays/2006/dim2006/framework_for_building_reputation_systems).

²⁵⁸ https://qz.com/1244155/good-luck-leaving-your-uber-driver-less-than-five-stars/

²⁵⁹ Apostolos Filippas, John J. Horton, and Joseph M. Golden, 2018, “Reputation Inflation,” in EC ’18: Proceedings of the 2018 ACM Conference on Economics and Computation, Ithaca, NY.

²⁶⁰ https://medium.com/@jacksoncunningham/digital-exile-how-i-got-banned-for-life-from-airbnb-615434c6eeba

---

Kalantzis, Mary, and Bill Cope. 2020. Adding Sense: Context and Interest in a Grammar of Multimodal Meaning. Cambridge UK: Cambridge University Press. pp. 104–108. https://doi.org/10.1017/9781108862059

Uber, Controlling the Conversation in the “Sharing Economy” §AS1.4.1d