My TiVo Thinks I’m Gay

Algorithmic Culture

To address why TiVo was the site of sexual anxiety, it is important to first situate it and its recommendation system within its industrial and cultural context and history. TiVo’s push toward microcasting and personalization and the desire to target and advertise to individual users rather than groups has come to define not only the contemporary television industry, but also the commercial Internet, wherein valuable user data are continually (and often unknowingly) traded for a more personalized and automated user experience, complete with comprehensive recommendations. Within this system of normalized surveillance and user exploitation, companies like Facebook, Google, Apple, and Twitter use these data to, among other things, generate user-specific advertisements that are often only nominally distinct from those recommendations that the data were originally traded for. Ted Striphas has described these technologies as a central force in the creation of an algorithmic culture, defined by the delegation of “the work of culture—the sorting, classifying and hierarchizing of people, places, objects and ideas—increasingly to computational processes” (Striphas 2015, 395). Striphas asserts that the personalization and recommendation algorithms used by companies like Google, Facebook, Twitter, Netflix, and Amazon alter “how the category culture has long been practiced, experienced and understood” (Striphas 2015, 396).

When algorithmic culture scholars discuss specific algorithms, they almost always focus on those digital recommendation systems first popularized by TiVo. There is certainly much to admire about burgeoning algorithmic culture scholarship, especially as it stands firmly against the omnipresent utopian rhetoric around digital technologies that continually presents our current media landscape ahistorically as the best and most natural one possible. Yet, too often this scholarship goes too far and becomes itself ahistorical and overly reliant on technological determinist logic. Mark Andrejevic, Alison Hearn, and Helen Kennedy have characterized our contemporary algorithmic culture in terms of the increasing scale of surveillance and opacity afforded by a reliance on data mining and big data analytics (Andrejevic et al. 2015, 380). Although these trends are certainly troubling, I would argue that they are neither new nor technologically determined. There is no question that contemporary data mining techniques have achieved incredible precision and sophistication, but American democracy, industry, and mass media have long been accused of being too massive, surveillant, opaque, and data hungry; the information disparity between those in power and those without has historical precedent. Furthermore, algorithmic culture scholarship tends to present individual algorithms as static objects with an overwhelming amount of “authoritative” (Striphas 2015, 406), if not authoritarian power, to the degree that Alexander Galloway and others have suggested they have led to a “control society” (Galloway 2004). Thereby, these scholars minimize the role of those humans who not only create those algorithms, but also continually manage and update them. As the constant updates of mobile apps attest to, social media companies constantly tweak and transform the algorithms and functions central to their sites not just to make them more accurate, but also to react to changing fads and cultural norms. Algorithms are not static objects and they do not break the dynamic relationships that structure our everyday lives. Rather, as Christian Fuchs has argued in relation to Google, they are “dialectical system[s] reflecting the contradictions of contemporary capitalism” (Fuchs 2013, 256). These technologies may affect certain cultures, but they are also—simultaneously and irrevocably—transformed by such cultures as well.

With a few notable exceptions, just as algorithmic culture scholarship tends to present algorithms as opaque, static, and despotic, it also presents users as incapable of critical reflection, transgressive actions, or the simple act of decoding that has been a central facet of Cultural Studies since its beginning. Striphas, following Tarleton Gillespie, suggests that users can only question algorithms in an instrumental mode, asking whether their results are “right” or “wrong” rather than how their assumptions were shaped or how their use leads to certain forms of exploitation (Striphas 2015, 406). In turn, Jodi Dean argues that it is extremely improbable (if not outright impossible) to use these and other social media technologies for transgressive or subversive ends as any use of them at all is inherently capitalistic and leads to a greater information disparity and profits for the industry (Dean 2009). Although these scholars certainly have a point and are pushing back against those who would uncritically frame digital technologies as our saviors, they too often underestimate the role that users have in reading and decoding the culture around them, regardless of the form that culture takes or how it came to be. I certainly agree that big data industries often believe that with enough information and the right algorithms, they can control (if not fully interpolate) their users, but I neither believe that this is a new desire nor that users are as naïve or powerless as many scholars of algorithmic culture suggest. As Cultural Studies scholars begin to play a role in critiquing algorithms and exploring how they function as a material form of culture, we should start not by assuming that sudden changes are afoot (they aren’t), or that the dominant culture is now all-encompassing because of these technologies (it isn’t). This examination of TiVo is meant to build on this scholarship by illustrating the complexity of how people produce, use, and experience recommendation systems as algorithms that do not simply enforce automated opinions (however much they may attempt to), but rather continue to leave room for moments of transgression, subversion, misreading, disidentification, recoding, queering, and ordinary agency.

Harpoons and Casting Nets: TiVo’s Industrial Context

Even as TiVo worked to usher in an algorithmic culture that appeared to be new and disruptive to the status quo of the television industry, it was also deeply enmeshed in this industry and its efforts to remain dominant within a digital landscape. Indeed, various companies including AOL, Time Warner, Microsoft, Apple, and WebTV had been developing “interactive televisions” (iTVs) throughout the 1990s at tremendous costs. In 1998, after TiVo had been announced but before its release, a Forbes article asserted that “hundreds of digital Captain Ahabs have in vain thrown golden harpoons at Interactive television, yet the whale has eluded them all,” but suggested that TiVo may have finally struck gold. William Boddy has noted that many of these failed efforts focused on making the television more like a simple computer for reading e-mail and surfing the web (Boddy 2004). Indeed, these technologies were largely marketed to those who did not own a computer and had no Internet access—a market that was rapidly shrinking by the early 2000s. Parks argues that the advertisements for these products framed normal television as a passive medium and iTVs as active, which in turn triggered “gendered assumptions” and “class-based discourses related to digital access and speed, and broader issues of cultural taste and social distinction” (Parks 2004, 134). Indeed, these companies tried to shame their largely lower class and rural customers into purchasing iTVs by making their current, computer-less existences appear unacceptable.

Rather than present itself as a computer alternative, TiVo became successful by marketing its simple interface and status as an updated, digital videocassette recorder —a tool for making TV viewing easier and more manageable. Indeed, though television viewing tends to be framed as passive, the large increase in television channels (from, on average, 41.1 receivable channels in 1995 to 102.1 in 2002) and content in the late 1990s through today has made the act of picking something to watch an often complicated decision that TiVo proposed to ameliorate (“Average Number of TV Channels Receivable by US Household Drops in 2004” 2004). Rather than depict television as a passive (dumb) medium, TiVo presented contemporary television as a complicated landscape that even the savviest viewer could not easily navigate. One TiVo commercial from 2000 used a first-person perspective to display the viewer calmly walking through a television network office with two bodyguards, who (equally calmly) toss an executive out a window; the commercial then commands the viewer to “program your own TV network” (“Tivo Network Executives Commercial” 2000). With the ever-increasing amount of content on television, TiVo suggests that we (regardless of our sexuality, gender, race, age, or class) must demand that this content be both more heterogeneous and more personalized. As “you,” in first-person perspective, are barely noticed by any of the workers in the office approaching the executive, this commercial simultaneously suggests that TV networks do not respect viewers and also that this lack of regard has specifically led to the need for viewers to actively take control of their media—to overthrow the system.

These efforts at presenting itself as subversive and revolutionary worked and TiVo quickly became synonymous with interactive television: for a while, the term TiVo even became both a general term for all PVRs and a verb that referred to the act of recording something on a PVR (Robins 2005). At the same time, TiVo presented itself as the consummation of a long trend in the television industry from broadcasting to narrowcasting to microcasting. Ron Becker has argued that since the 1970s, television marketing has “broken Americans up into narrower and narrower psychographic groups targeted through direct mail, lifestyle magazines, and cable channels” (Becker 2006, 5). As American culture became more fragmented, advertisers began targeting their messages to niche audiences they believed would most likely be interested in their products.

With commercials displaying TV executives tossed out windows, TiVo presented itself to the public as an antagonistic and anarchic force against not just television, but also the mass-market more generally. As one New York Times Magazine article proclaimed, “if no one watches commercials, then there is no commercial television” and if there is no commercial television, there is no mass-market (Crain 2000). Yet, in the trade press, TiVo continually asserted that it was actually on the mass-market’s side and was the answer to the advertising industry’s microcasting prayers. Depending on the audience, TiVo was either part of the status quo, deeply subversive, or possibly both. Although TiVo certainly made the thirty-second commercial a less viable format, it also heavily encouraged product placement, interactive ads that would allow viewers to immediately purchase products, and “more appealing ads” that users would not want to skip over (Hall 2016). Daryl Simm, the CEO of Omnicom Media argued that the creation of such adverting would lead to viewer loyalty, which in turn would lead to a greater return on investment (Crain 2000). Lauding TiVo’s efforts (in a distressingly ironic twist on Newton Minow’s famous anti–mass media and mass-marketing speech about the televisual “wasteland”), Advertising Age suggested that “TV has always been a vast wasteland when it comes to understanding how programming affects advertising” (Crain 2000). By 2003, TiVo strengthened its focus on making personalization marketable by collecting data on its viewers and selling them to advertisers. In contrast to Nielsen’s Ratings, TiVo President Martin Yudkovitz asserted that TiVo’s data’s “level of granularity is absolutely unheard of. You can tell at any moment what parts of the country start to tune out of a program, at what time, by demographic and time of day” (Elkin 2003). Although TiVo’s efforts may have been viewed as revolutionary and at times transgressive, they were also a continuation of larger trends throughout the television and advertising industries.

Casting for Sexual Desires

As television marketers fixated ever more narrowly in their search for lucrative eyeballs, they were guided not just by class divisions, but also other splits in contemporary culture: most notably, those related to sexuality. Becker argues that sexual identity is central to a transformation in advertising and that niche marketing led to a huge increase in homosexual narratives on television. His work, which largely focuses on the period right before TiVo’s PVR, helps to situate TiVo in relation to the television industry and the connections it made between narrowcasting and the greater visibility of gays on the networks. In the 1990s, marketers and television networks focused on “the slumpy demographic (socially liberal, urban-minded professionals)” as the most valuable niche audience (Becker 2006, 81). Networks tried attracting this group by airing gay characters and plotlines, designed to appeal to their liberal, multicultural, and “hip” politics; by 1997, there were thirty-three queer characters on prime-time network television (Becker 2006, 104). Notably, much of this programming was directed not at a gay audience, but rather at the gay-friendly, but straight audience “looking to reconcile their liberal ideals with their bourgeois materialism” (Becker 2006, 119). Gay content, which in the 1990s was rarely ever actually sexual, was sold as edgy, but never over the line.

At the same time, niche marketers were also becoming more interested in gay and lesbian demographics because of their perceived high value as consumers—a perception that Becker points out was exacerbated by highly questionable research methods and grossly distorted statistics (Becker 2006, 128). In the process, they helped shape which homosexual and queer identities could be made legible. Like Becker, Lisa Henderson describes in detail the complicated ways in which class structures act as a dominant force in how gay—often specifically queer—communities form and how homosexuality is understood. Class distinctions shape queer identity and define what characteristics are positive (wealth, family, normalcy, and “fabulousness”) and which are negative (“performative excess and failures of physical control”) in ways that “appeal to a range of consumers and still flatter those at the crest of advertising trade value” (Henderson 2013, 34). In the process, homosexuality became a legitimate and legible identity because of its connection to upward class mobility and consumerism—a transformation that actually ignored the vast majority of gay people.

Although many narratives used glamorized images of what Henderson calls “good queerness” to reinstantiate the dominance of heteronormative values, the “My TiVo Thinks I’m Gay” plotline illustrates how they also simultaneously depicted a breakdown between heterosexual and homosexual identities and desires. “My TiVo Thinks I’m Gay” relied on the “mistaken-sexual identity plot-twist”—one of the most popular gay narrative devices of the 1990s—in which a gay character is mistaken for being straight, or vice versa. As the question in this plot was always “What to do if TiVo thinks you are gay,” it is clear that though these plotlines may have involved gay content, they were always aimed at a straight audience anxious about its own sexuality and privilege. As microcasting became ever more targeted toward specific people rather than groups or demographics, questions of identity construction became even more relevant.

Becker posits that these narratives reinforced “the notion that the line between gay and straight was far from clear” (Becker 2006, 185). Although they taught the lesson that we all share the same basic humanity, regardless of sexuality, they did so by marginalizing any queer politics and/or desires to claim any differences between gay and straight cultures. These narratives and the increasing normalization of gay identities created what Becker calls a “straight panic [that] refers to the growing anxiety of a heterosexual culture and straight individuals confronting the shifting social landscape where categories of sexual identity were repeatedly scrutinized and traditional moral hierarchies regulating sexuality were challenged” (Becker 2006, 4). As these narratives began to not just validate but also valorize homosexual identities throughout the late 1990s, they destabilized the privileged normative position of heteromasculinity.

Fictional accounts of TiVo users treat themes of straight panic and sexual identity confusion, reflecting the ways in which TiVo and many other algorithmic technologies present themselves as transgressive while simultaneously reinforcing certain dominant, heterosexual norms. For instance, in an episode of the HBO series The Mind of the Married Man directed by Bruce Paltrow, TiVo is playfully framed as a menace to patriarchal, retrograde definitions of heteronormativity—a position that HBO itself and much “quality” television have at times and with limited success tried to occupy throughout the history of narrow and microcasting with series like Sex and the City (1998–2004), Six Feet Under (2001–2005), The Wire (2002–2008), The L Word (2004–2009), and Looking (2014–). This The Mind of the Married Man episode features a TiVo introduced by its ability to know that “if you record Star Trek, [it] assumes you like that kind of thing and then when you aren’t home it records The X-Files.” This product placement is highly self-conscious both because TiVo symbolized a type of luxurious spectatorship that many series wanted to be associated with and because TiVo’s fast-forward feature was itself a major cause of the growth of product placement on television more generally.

At the beginning of the episode, the protagonist Micky Barnes (Mike Binder) announces to his coworkers that because he recorded a few episodes of Will and Grace and Ellen, TiVo now thinks he is gay and will not stop recording Queer as Folk and Judy Garland specials. Micky explains that he has been trying to outfox the program to “get it to go the other way” by recording MTV Spring Break (2001), Playboy After Dark (1969–1970), and MTV Nipple Parade (a series I had to look up to make sure it was not real). Unfortunately for Micky, “The thing won’t budge. It insists I’m gay; it’s a problem.” TiVo’s recommendations become a point of crisis for Micky as he goes through the rest of the episode considering what the relationship between his actions and his perception of himself really is. Throughout the episode, Micky’s TiVo recommendations force him to question himself and his relationship with his friends and colleagues. In the process, Micky illustrates how “straight panic” became attached to TiVo and microcasting strategies more generally. This was a strategic effort to not just get the slumpy demographic to use PVRs, but also to care what their technology “thinks” of them. As this meta-narrative focuses entirely on television viewing habits, it also emphasizes the changing media landscape and vexed reaction by HBO and other premium outlets that quite suddenly catered to many new and previously marginalized audiences including the LGTBQ community while featuring testosterone (and full-frontal nudity)–driven series like Hung (2009–2011) and Entourage (2004–2011).

This issue recurs in an episode of King of Queens wherein TiVo’s efforts at recommending media are characterized as a form of sexual harassment. This episode, called “Mammary Lane,” presents three farcical sexual harassment scenarios that together lampoon victims of real harassment. These narratives include a toddler who continually grabs her babysitter’s breasts, a man who tries to be nice to an unattractive bowling alley attendant by attempting to flirt with her, and a TiVo that records gay-themed series for its heterosexual owner and thus ruins his date. The victims of such “harassment” are shown in different states of trauma, with one fearing for her life in a parking lot and another rocking back and forth on her bed while she is accused of “asking for it” from a toddler. On the one hand, these three narratives depict a postfeminist backlash/hatred of feminism (if not women more generally) that suggests how TiVo’s suggestions can harass its users, while also making fun of the serious import of harassment. Imputing harassment to a machine and a baby underscores ways in which harassment can be framed as inadvertent and as a misinterpretation of meaning; the harassers here are innocent while the harassed are crazy. On the other hand, these scenarios are so outrageously silly and self-conscious that they taunt the viewer to analyze the situation’s comedy and critique the postfeminist rhetoric of harassment and victimization that it relies on. In either case, the episode presents as absurd the fear that TiVo’s surveillance strategies may have unforeseen consequences, and that its “opinion” of users would be trusted more than the users themselves.

The episode introduces TiVo through Spence Olchin (Patton Oswalt), who has just purchased it for his bachelor pad. In an attempt to impress a female neighbor he has a crush on, Spence asks her,

Want to see my new purchase? It’s my TiVo; it’s this thing that records TV shows but the cool part is that it is intuitive. Once you program in some shows you like, it gets to know your taste and then, automatically picks out other shows for you.

After explaining that he programmed in “Sex and the City, Six Feet Under and a few other favorites,” he checks to see what TiVo recorded for him automatically and finds The Adventures of Priscilla, Queen of the Desert (1994); Judy Garland, Live at Carnegie Hall (1961); Decorating With Style (1996–2002); and Queer as Folk (2000–2005). In a moment of (straight) panic, Spence yells, “Oh my God, TiVo thinks I’m gay!” His attractive neighbor then abandons him to his hetero-anxiety. TiVo appears to have an agency of its own, which makes Spence uncertain as to whether the machine is judging him or if it is instead simply mimetically reacting to his own actions.

In an odd turn, TiVo becomes a status symbol even as it makes Spence extremely uncomfortable. Rather than associate homosexuality with upward mobility, the implication that homosexuals use TiVo makes the technology—and contemporary television more generally—appear classy. Spence makes his neighbor check out his suggestions because he hopes that she will be impressed both by his expensive new machine and the suggestions themselves, which he hopes will portray him as the interesting heterosexual that he imagines himself to be. Although Spence challenges the suggestions TiVo has for him, he never questions TiVo’s role as a self-representational technology even though he does not like the representation it has come up with. Yet, he questions the generated self-representation by imparting TiVo with agency. By placing these anxieties and the threat of TiVo suggestions to normative conceptions of sexual identity within such a benign context, they are made to appear hyperbolic and silly. This situation nods to the reality that TiVo can indeed create anxiety around a user’s sense of sexual identity, while also making clear that this is not an issue that should be taken seriously. The humor of this episode leaves the troubling nature of this recommendation system intact.

Along with sexual anxiety, TiVo’s appearance here also points to class pressures that are an explicit part of King of Queens as a whole. Henderson outlines how, in both broadcast and niche television, representations of homosexuality tend to “borrow liberally from the class fantasies of everyday life, especially fantasies of mobility and having” (Henderson 2013, 16). This process is inverted on King of Queens as Spence is classed in part due to his queered desires. The characters in the series are largely high school educated White heteronormative members of the lower to middle class with aspirations for more. Spence himself is a nerdy tollbooth worker from rural West Virginia. With an interest in science fiction and comic books, he is perhaps the most cultured of the series’ ensemble. These interests are underscored by his declared penchant for HBO “quality” series, which Michael Newman and Elana Levine (2011, 71) argue constitute a complete rejection of the style, address, and viewer positioning of King of Queens and other similar multi-cam sitcoms associated with a dated broadcasting strategy. They also point out the TV industry used PVRs to legitimate itself as a higher art form by suggesting that it was no longer just for the “‘saps’ and ‘dipshits’ who tune in each week to follow a narrative, with nothing better to do than to be at home, in front of the set, sitting through the commercials” (Newman and Levine 2011, 1). They demonstrate how narrowcasting, microcasting, and video recording have always been classed. Yet, with its prodigious use of product placement coupled with an anxious depiction of PVRs, King of Queens illustrates the vexed nature of these legitimating efforts as television went from being deeply uncomfortable with its lower class image to fully coming out by embracing the many desires and viewers it appealed to. In this episode, heterosexual and class anxieties overlap as an interest in quality television and luxury goods is marked as gay.

Yet, this episode also destabilizes the dominant ideologies that frame homosexuality as a joke. The audience is encouraged to laugh at Spence’s absurd homophobia and hetero-failure. This failure lies not with the recommendations but rather with his ability to interpret and follow them. Spence inputs several quality TV series and instead of understanding TiVo’s recommendations as a list of texts associated with affluence and elitism, he construes them (and affluent culture) in a way that he rejects, or self-consciously thinks others will reject. The point is not that Spence interprets these recommendations correctly or not, but rather that they are queerly open to interpretation; one could see them as meaning most anything, or nothing at all. The laugh track erupts at moments that reveal how anxious Spence becomes due to these recommendations and what they perhaps reveal to both himself and others concerning his own status as a heterosexual. Late in the episode, Spence tries to correct TiVo’s “dumb mistakes” by recording Football, Nascar, and Basketball games. Spence does this to let TiVo know that rather than being gay, he is actually a “sports junkie,” an identity Spence assumes runs in opposition to an imagined homosexual-highbrow culture.

In the end, Spence’s attempt to become re-empowered by resting his ability to define himself on TiVo only results in him recording, watching, and consuming more media—TiVo’s goal all along. Yet, he has misunderstood the TiVo system and assumed that later actions simply nullify past (recording) transgressions, when really TiVo’s algorithms work to find the intersections between such opposing programs. The result is a set of recordings like Men’s Ice Dancing and Breaking the Surface: The Greg Louganis Story.” Although his neighbor takes TiVo’s side and suggests that it does seem to be intuitive, Spence responds with a sudden outburst: “I’m not gay. I love women, I stalk them, I’ll have sex with you right now if you want.” As the laugh track suddenly erupts, the neighbor leaves the apartment in a rush and Spence screams at his television, “Oh how dare you label me!” and then proceeds to call his TiVo gay. Later, Spence announces that he is being sexually harassed by TiVo.

By equating “labeling” with sexual harassment, Spence frames his anger as arising not from homophobia per se, but rather from having his identity judged at all. Immediately after he tries to embrace a pathological masculinity by proudly stating that he stalks women and will have sex with a woman at the drop of a hat, he ironically uses queer feminist rhetorical strategies to challenge not only TiVo’s representation of himself as gay (or interested in gay subject matter), but also any form of representation that seeks to label people within static categories. Spence would not be angry if TiVo had presented him with a list of suggestions that had stronger heterosexual connotations (like King of Queens itself) not just because he thinks of heterosexuality as being a more positive category than homosexuality (which he largely thinks of as a slur), but also more importantly because he identifies himself as a heterosexual. These fears are made only more ironic by the name King of Queens, a title also shared by drag queen television star RuPaul (Yang 2013). If Spence openly identified himself as a homosexual, TiVo labeling him as such would not upset him, or at least not for the same reasons. Instead, he may have been bothered by the broad surveillance paradigm that TiVo and technologies like it helped naturalize and domesticate—a more reasonable fear completely ignored by King of Queens.

Although TiVo’s ability to destabilize identity in such an intimate way may suggest its queer and transgressive potential, such potential is ultimately limited and made laughable. Both The Mind of the Married Man and King of Queens present a character’s panic as a momentary misunderstanding rather than as an affirmation of the complexity and instability of gender and sexual identity. These plotlines place the character’s sexual identity into question only to assert—and straighten—the lines between such identities. The idea that there is a direct relationship between enjoying a series with a gay theme or a largely gay audience and necessarily being gay is, of course, far too simplistic. Both Sex and the City and Six Feet Under (as well as all of the other series mentioned) are notable not because they exclude a heterosexual audience but because they encourage a metrosexual crossover through a shared focus on affluence and empowered consumerism. Yet, though these series and TiVo’s suggestions allow for a more fluid relationship between sexuality, taste, and empowerment, here these efforts backfire as Spence reacts against them in favor of traditional heteronormativity and its static binaries.

But My TiVo Really Thinks I’m Gay

Although The Mind of the Married Man and King of Queens are fiction, the anxiety concerning TiVo’s role in imagining microcast audiences in relation to sexual and class identities is not. On November 26, 2002, Jeffrey Zaslow of The Wall Street Journal reported on numerous people all experiencing the same discomfort with how TiVo rendered their identity. In his article titled “If TiVo Thinks You Are Gay, Here’s How to Set It Straight,” Zaslow, at once humorous and earnest, discusses instances in which real people—along with television characters—fought back against TiVo’s “cocksure assumptions about them that are way off base” (Zaslow 2002). In his article, Hollywood producers in particular voiced their concerns around privacy and sexuality. The inclusion of these perspectives suggests how TiVo’s recommendations potentially could affect not only the way spectators experience their media, but also how this media gets made. One of the more notable executive producers Zaslow interviewed, Basil Iwanyk, described himself as the “straightest guy on Earth” though he would later ironically produce such hyper-queer opuses as Clash of the Titans (2010), The Expendables (2010), and The Town (2010; Zaslow 2002). The knowledge that powerful auteurs who have in-depth knowledge of media production use TiVo also makes the technology appear more legitimate. Zaslow likely focused in part on Hollywood executives because in 2002, TiVo was both a luxury item priced out of reach for most people and a tax write-off for television and media producers.

The anecdotes in Iwanyk’s piece illustrate how the straight panic and sexual anxieties that surrounded microcasting audiences also affected (and were, in turn, reproduced by) media producers. For Iwanyk, the TiVo Suggestions Service became a thorn in his side when it started to “inexplicably” record programs with gay themes (Zaslow 2002). These automated recommendations seemed to be telling Iwanyk that he desired the same things as others who gave a thumbs-up to gay-themed television series, namely, other men. Iwanyk’s recommendations gave him a sudden insight into an alternative way that his film interests could be interpreted that he flatly rejected. The question for Iwanyk was never whether TiVo was queering his texts, but rather whether it was queering him. In either case, by pointing to the polysemic nature of these media products, TiVo points to how algorithmic culture and the recommendations it fosters do not just allow for reading texts against the grain in at least potentially transgressive ways, but also often encourage such activities. It is tempting to think that algorithmic culture with its focus on creating an aura of objectivity and mechanical order would be absent of moments of serendipity, or even mere surprise, but these stories of decoding and misreading recommendations suggest ways in which these systems continue to hold subversive potential.

And yet, much of this potential for misreading, surprise, and transgression comes from the obscured specifics behind how TiVo—and algorithmic culture more generally—generates its recommendations. Like Amazon, Netflix, Google, Facebook, and any other company that uses personalization technologies, TiVo primarily generates recommendations to users based on what other similar users have liked in the past. Although this logic is intuitive, it also hides the complexity of underlying secret and proprietary algorithms and the statistical models that govern exactly how similarities between users are defined. TiVo’s automated recommendations suggest that there is a link between a piece of media and the users that consume it, but the qualities that unite them are rarely definable and TiVo software does not specify exactly why it generates the recommendations it does. When TiVo recommends King of Queens to viewers of Will & Grace, it does not explain the basis behind why it finds an affinity between these shows, whether based on genre, cast, color palette, set design, and so on. Users are left to imagine what aspects tie various recommendations together, and, more importantly, what these recommendations say about themselves and their desires. This process is made more complicated by the various crossover audiences that contemporary television series try to attract.

Without explanations, Iwanyk and others like him jumped to the conclusion that the quality that linked them with their recommended series was based around sexuality and that their actions were causing TiVo to represent, or even “think” of them as homosexuals. These users viewed TiVo as a disempowering force that destabilized their tastes and sense of self. Because TiVo keeps its algorithms secret, there is no way to know whether and/or how it (or any other company) takes stereotypes into account and whether its recommendations really are meant to be based on sexuality or any other identity markers. Iwanyk’s reaction is an example of how these strategies lead users to read stereotypes back into their generated recommendations. Instead of questioning TiVo, Iwanyk questioned himself, and started watching, recording, and rating more television that he believed a heterosexual male would be interested in (i.e., series that he referred to as “guy stuff”; Zaslow 2002). Unfortunately for him, he received suggestions like documentaries on Adolf Eichmann and Joseph Goebbels: “It stopped thinking I was gay and decided I was a crazy guy reminiscing about the Third Reich” (Zaslow 2002). Iwanyk’s reaction illustrates a confusion between gender and sexuality and his assumption that homosexuality is a form of femininity. Such confusion is mapped onto and facilitated by the confusion behind how secretive algorithmic recommendations really work and what kinds of explanatory or perceptive powers they might hold, if any.

Conclusion

All of the above accounts, whether real or fictional, exemplify users’ desire for TiVo to get them “right.” These users felt disempowered by TiVo precisely because it did not see them as they wanted to be seen. A contrasting and more subversive reaction to TiVo and its surveillance regime, however, is the desire for it not to recognize the user correctly. Enacting this is unpleasant and requires users to record and rate highly things they have no interest in and stop viewing what they actually want to; indeed, such activities take away any benefits or pleasures that owning a TiVo (not to mention a television) might entail. And yet, several of the people in Zaslow’s article, who self-identified as gay did just that because they did not like the idea of their media both recognizing and recording their sexuality. One of the two, Ray Everett-Church, an Internet privacy consultant, said his “TiVo quickly figured out that he and his partner were gay” and though this did not negatively affect them, they decided to try to “confuse the software by punching in ‘redneck’ programs, like Jerry Springer’s talk show” (Zaslow 2002). By recording content they were not interested in viewing, this couple tried to break the algorithm with the express purpose of regaining their agency even though that largely defeated the purpose of owning a TiVo and the pleasures that might come from it. These users frame their attempts to control how they were represented by TiVo, particularly in terms of sexuality, as playfully subversive; they took TiVo up on its proposition to “program your own TV networks,” but (ironically) this programming ran counter to TiVo’s own recommendations.

Focusing on the question of what to do if TiVo thinks you are gay is a provocative, if ultimately vexed way to address the complexity of subversion, misreading and agency in digital recommendation systems and algorithmic culture more generally. The question itself, along with the reactions and texts that it generated, can and should be viewed as simultaneously subversive and docile, as it points to both the blurring lines of sexuality and digital identity, but ultimately in a way that never quite challenges the status quo of heteronormativity or digital surveillance. The responses from viewers are never to simply return their TiVos, which would potentially be much more disruptive to algorithmic culture generally, but are rather centered on manipulating the technology to reassert personal control, agency, and privacy.

Ultimately four years later, TiVo appropriated this meme of the “gay” recommended user to associate itself more closely with gay audiences by placing the question that I have focused on here in an ad in the 2006 GLAAD Media Awards Program Booklet. Along with congratulating all the nominees for awards, the ad depicted a TiVo menu full of LGBT-centric programming and asked, “Does your TiVo box think you’re gay? You decide. TiVo, TV your way.” Although this joke implies a heterosexual audience mistaken for being gay, here TiVo reworks it to instead both affirm the potential benefits of being recognized (without judgment) for being gay, and laud the increasing amount of television content that features LGBT-centric characters and plotlines. Here, TiVo uses this meme to redefine itself as a tool that no longer stands in the way of user agency and privacy, but instead ensures these desires, especially in relation to sexual identity.

Yet, by this point TiVo’s influence had already begun to wane as more companies began to offer similar services and cable and satellite companies started bundling PVRs with their services. Over the next few years, customers stopped using the term TiVo as a catchall for all PVRs and the cable companies themselves even rebranded PVRs as digital video recorders (DVRs). Among other things, this change helped obscure the early history of these technologies and left TiVo, a company that continues to sell PVR boxes and services, largely forgotten. TiVo, which holds many patents related to PVR technologies, continues to be valuable and many companies have considered purchasing or merging with it. Presently, Rovi, a maker of interactive program guides for televisions, is in talks to merge with TiVo, presumably because of these patents and the protections—and potential revenue from future lawsuits—they afford (Picker 2016). Although it would be easy to think of TiVo simply as a failure, its fall instead signifies its extreme success at transforming television by making DVRs (along with recorded and/or streamed content) a dominant mode of experiencing television and by illustrating how algorithms intersect with transforming notions of sexuality, gender, and culture.