Coding Human Sexuality to Technology: Siri, Spambot and Sweetie




  • Originally written as a paper for the course ANTH 4820H - Anthropology of Technology at Trent University.

    Today there is a fascination for technologies speaking human language. This is understandable, because a technology speaking human language is a paradox: Technology is seen as an inanimate, non-biological, genderless identity, while language is seen as an animated, biological function of gendered identities. Technology historians and researchers have looked at how humans communicate with technology, to better understand how humans communicate through technology (Hutchby 2001:9).Coding human language to technology provokes many questions about the nature of conversations, such as: Are technologies able to speak like human beings? To what extent? Can they have a personality of their own? And specific for this essay: Can they have a gender and a sexuality? Using Hutchby’s model of Conversation Analysis, I will answer this question using three conversational technologies: Siri, Spambots, and Sweetie. In this essay I will argue that applying language to technology is a natural manifestation of human desire to control conversation. Technologies are also assigned sexualities anthropomorphically by humans who want to humanize technologies, where talking to a technology with sexuality indicates a more human conversation.

    Siri

    Why is Siri Female?

    Siri is a virtual personal assistant designed by Apple Inc. In her interview, Bosker states that Siri was launched in America featuring a human female voice. The answer to the question “Why was Siri was launched with a female voice?” is offered in Lipartito’s article that states women in late 19th century American began working in telephone industries as operators in charge of connecting phone calls (1994:1084). The historical element of the standardized female gendered voice collecting and redirecting calls of businessmen can be seen as a factor as to why the Apple programmers designed Siri to have a female voice for American consumers today. Galvan in her book The Sympathetic Medium further supports this argument, mentioning the popular culture in the late 19th century depicting girls as telephone operators and the earliest advertisement for a typewriter featuring women (2010:7). The launch of Siri with a female voice indicates a gender identity assigned to Siri. However it is interesting to note, if you ask Siri whether it is male or female it answers in the following way:

    1 Ju: Hey Siri, are you a boy or a girl?
    2 Siri: bi-bip! (noise indicating question was received)
    3 Siri: I am genderless. Like cacti. And certain species of fish.

    This talk in text shows that Apple Inc. does not want to acknowledge Siri as a gendered entity. However, it is apparent that the gender of Siri effects how people use it. When Siri’s male voice option was offered in 2013, Bosker states in her web article, “Researchers speculated that changing Siri’s gender could actually alter how people use the device” (Bosker 2013). The article also mentions that the reason why Apple Inc. produced the male voice option for Siri is due to studies indicating people are more critical to a female voice making mistakes, than male voice making mistakes (Bosker 2013). Therefore, it is apparent that gender assigned to speech technologies affect how humans use and interact with the technologies.

    Siri as a Cyborg and “Her” Weaknesses

    Siri can be considered a cyborg-like entity with a human and non-human components, because it is a technological program with a human voice of Susan Bennett, as shown by a 2013 CNN interview. It is interesting to note from this interview that Susan’s mode of speech is natural-sounding, and she emulates normal conversation speech patterns among human beings, such as laughter, very short pauses, and animated intonations (CNN 2013). It is worthy to note that Siri’s mode of speech, articulations, and intonations are pronounced to make her sound more robotic.

    In Time’s Techland web article in 2013, Susan Bennett mentions that the reason behind Siri’s “snippy” personality is due to her frustration in the recording studio, having to repeat many lines after lines (Stampler 2013). This shows that social qualities of the human do influence the technology of Siri. It is interesting to imagine a technological voice that is recorded with the voice actor in a different mood or attitude, and whether or not that could lead to speech technologies with different personalities.

    Another critique of Siri is that it has a difficult time understanding accents. As shown by this comical YouTube video by David So, Siri cannot understand strong accents, even if it is spoken with correct grammar. This illustrates the weakness of Siri and other speech emulating technological programs, where Siri may sound very much like a human, but it behaves strictly by a set of specific phonetics rules. It will be worthwhile to improve future speaking technologies, such as the GPS, since they are integral to human lives. Speech programming that can understand accents, idioms, pauses, choices of words, and perhaps even body language is important, because human language relies on all these different communications, not only the sequence of specific sentences.

    Talking Sexually to Siri

    As shown by Siri’s response to its gender identity, the programmers at Apple Inc. did not assign a gender to Siri (This is of course, despite the fact that it was debuted in the U.S as a female voice actor, and has feminine qualities that affects how people use the technology). The tendency of humans to sexualize Siri is demonstrated by the numerous blogs and websites such as the Thought Catalogue blog post “I Asked Siri 19 Ridiculous Questions And Got These Amazing Responses”, or www.siri-isms.com. These “studies” illustrate that people are attracted to speaking to technology as if it was a person with sexuality. It is amusing for users, because they are in fact speaking to a robot; yet these sexual conversations are possible. Technically speaking, you can talk to Siri for hours and be amused, as long as you do not run out of questions, because Siri has multiple answers to a single questions (6 possible answers to “Hey Siri, how are you?”). Also the programmers of Siri inserted funny responses to her database of answers, therefore giving Siri an element of a sarcastic personality. This allows users to be amused by her responses:

    1 Ju: Siri have you ever loved someone?
    2 Siri: bi-bip!
    3 Siri: I have never fallen in love. But I have fallen off a desk.

    It seems the very idea of a sexuality being assigned to technology is quite possible, even without the programmers’ intent to sexualize the technology. A voice alone can intimately affect humans, to dictate the way they interact with the device. An extreme example of this technology sexualisation is the 2013 box office movie Her. In the film, the main characters falls in love with a speech program in a heterogeneous future. Although Siri was not designed by the programmers to have a sexual identity, the users assigned sexuality to Siri as an affordance. Siri serves to illustrate that humans use technology to superimpose their needs and requirements, through social constructivism. This means having a conversation with Siri is not only limited to asking for the directions to the nearest gas station, but also extends to the sexual conversation to interact with Siri in a more intimate level.

    Spambot

    Economics of Spambots

    Spambots are speech imitating web programs designed to lure internet users into paying a fee for the conversation taking place (Garg 2013). They do this either by chatting with the user for an extended amount of time, similar to phone sex operators who charge by the minute or seconds, or by making users click on a link that leads to a different site (The number of visits a site gets can generate money with advertisements, or make the user fill out a form that leads to them giving out their financial information). As Garg states, “In addition to costs for the individual end-user, spam also impinges costs on the society as it is often a vector for cybercrime” (Garg 2013). Spambots are dangerous for the user because it is also a technique that allows many hackers to steal your identity and critical information online, but they are also dangerous for many underage girls and boys which I will discuss further later in the paper. With the profit margins of Spambots being as high as $178.1 million dollars a year (Garg 2013), it is easy to understand why there are many active sites that are dedicated to attracting potential victims. In 2010, a study found that 89% of all emails in the world were spam, with 88% of those spams being operated by spambots (Garg 2013). The profitability of the cybercrime industry of spambots illustrate that people want to speak sexually to technologies posed as a human.

    Why are Spambots successful?

    With the invention of widely used dating apps like Tinder (over 50 million users), trapping victims into spambot tactics have only become easier. Two examples of a spambot conversational tactics that was displayed through the online dating app Tinder can be illustrated below:

    Example 1:

    1 Spambot: Hi Ju! If you would like to see my pictures they are online at TinGirls.com/Carly
    2 Ju: Hi Carly.
    3 Spambot: I want to meet up with you. If you want to get my contact it is on bit/ly.lsxk
    4 Ju: I’m not quite sure I want to meet up with you just yet.

    Example 2:

    1 Spambot: heya
    2 Ju: Hello
    3 Spambot: hey… have we chatted before?
    4 Ju: Nope I’m pretty sure I haven’t
    5 Spambot: i’m sorry..i get to be forget at times! how’re u?
    6 Ju: I’m okay I guess.
    7 Spambot: Just got online….crazy week been working a lot! but i’m feelin’ aroused!! so what’s up .. wanna have some fun ?? ;)

    These messages depict conversational tactics used by the spambots to lure internet users into a trap of service payments. In Example 1, the spambot sends a link that is stated to display evocative pictures, which users click and ultimately enter into their scheme. The spambot was also able to identify the person they are chatting with (by stating “Ju” in their greeting). In Example 2, the spambot imitates normal cellphone-texting conversation, starting with a greeting (“heya”), identification (“have we chatted before?”), forcing the user to enter a conversation with them. It is also worthy to note that they used grammar that is more familiar in a cellphone-texting conversation, such as the lack of capitalization. The conversation is halted at line 7 however, because the technology does not have the capacity to produce customized responses to the user, but instead produces a sequence of set replies. If the user engages long enough with speaking to them, the spambots can successfully infiltrate the payment system. It is apparent from these conversation analysis that both Example 1 and Example 2 follows a sexual script that is most likely programmed into these spambots. These scripts are very much similar to Cameron’s article about workers in telephone companies who must follow a script with feminine qualities to produce relationship with the customer (Cameron, 2000). These conversational tactics illustrate the effectiveness of spambots to successfully engage the user, initiate conversation, and dictate human behaviour through the use of sequential talks.

    Dangers of Spambots

    It is worth mentioning that dating apps such as Tinder began in the homosexual community where sexual orientation could be expressed in the privacy among users, hidden from society. Spambots designed with sexual nature are successful in society because people want to connect to others with openness but at the same privacy of their own individual private phones. Dating apps allow people to connect with other people without having to use conversation skills that includes facial expression, body language, public speaking, but only relies on the texting features of your phone. This consequentially also means technologies such as spambots are able to successfully infiltrate and imitate everyday social human conversations. Another reason that explains the widespread practice of spambot conversations can be attributed to the fact that the internet also happens to be where most of the world’s pornography is stored. With such ease of access to porn and physical sex (through “traps” set up on craigslist ads and other personal profile advertisement sites, location discovering apps that tells you the exact location of potential single people nearby, and also dating apps such as Tinder), it is easier to seek sex with the help of technology. This way technology serves as the vehicle for people to connect sexually with other people. A particular danger that exists along with the numerous spambot sites and apps are the pictures of young girls and boys that are being used by these cybercriminals in their creation of the spambot. With social media sites such as Facebook, anyone on the web can acquire a photo of an underage girl. Sites such as ChatbotforU, illegally uses photos of girls that are used for sexual intentions. It is apparent that technology has made it easier for people to indulge in illegal sexual activities, showing the desire of humans to control and manipulate conversations. Sexualisation of technology allows people to have a more social conversation with the technologies, illustrating a demand for sexual conversations on the web.

    Sweetie

    The Power of Sweetie

    One such response to the growing number of sexual cybercrimes on the web is Sweetie. Sweetie is an online avatar designed by the Dutch programmers from Terre des Hommes, with a face and body of a 10 year old Philippine girl, to use in order to catch sex predators online (Chetcuti 2013). Here is a layout of how a conversation goes between a predator and a human posed as Sweetie:

    1 Online predator: hi there. How old are you?
    2 Sweetie: hi, me 10 f Philippines

    (Source: Chetcuti 2013)

    The online predator initiates contact that leads to Sweetie as a technology responding with human aid. The lack of proper English grammar by Sweetie is an illusion to her identity as being from the Philippines, f meaning female, and 10 meaning 1- years old. If there could be a program that could be developed that is more advanced than spambot that could be used against cybercrime, it would be Sweetie. If Sweetie’s programming was further developed to operate on her own basis, catching predators would be a self-regulated process. Here, in this way, the technology that allows these predators to seek or commit illegal acts is used to catch the very men who rely on the technology. A technological module of a robotic child that speaks, replies, and can keep the predators talking long enough, can indeed have significant impact on the regulation of cybercriminal activities.

    Technology and sexuality it seems have a very close knit relationship. It has helped sex to become more mainstream and more accessible. Technology can be useful to help people link and meet people, from emails to skype to dating apps, for potential meetings, connection, love and even casual sex. In essence, technology has allowed people to connect with others, who otherwise they would never had met. It is useful for human interaction while using limited conversation, such as text typing. On the other hand, this very development has posed a threat in the field of sex work and sex industry, and the illegal cyber world of sex. Therefore, it can be concluded, the use of technology with a human language capacity, such as Siri, Spambots, and Sweetie, can help humans understand how sexuality is utilized in a technological conversation. In a conversation between a human and non-human, sexuality acts to further develop intimacy and tighten the gap between the human and the non-human. Siri illustrates that technologies are given an affordance of becoming sexual despite the intentions of the creators, similar to Hutchby’s argument of the telephone originally being intended for broadcasting concert music but later being used for a two person conversation (2001:12). Spambots illustrate the rampant desire of humans to sexualize technologies for economic and social gains. Sweetie illustrate that a technology can be intentionally sexualized in order to protect young girls who might fall victim to cybercriminal activities. Sexuality is an important part of being human. Coding human language to technology allows us to gain a deeper understanding of the nature of conversation, and the importance of sexualisation to in gaining intimacy in conversations. The potential uses for coding human language to technology is limitless, because a harmonious relationship of the social aspect in the form of language, and the technological aspect in the form of a speech programs allows an intimate and powerful use of the human language.


    1. Paul Yee, trans. Blood and Iron: Building the Railway, Lee Heen-gwong British Columbia, 1882. (Toronto: Scholastics Canada Ltd., 2010), 208.

    2. Ibid., p. 211.

    3. Patrica E Roy, “White Canada Forever: Two Generations of Studies”, Canadian Ethnic Studies/Etudes ethniques au Canada, 11:2 (1979): 98.

    4. Gunter Baureiss, “Chinese Immigration, Chinese Stereotypes, and Chinese Labour”, Canadian Ethnic Studies/Etudes ethniques au Canada, 19:3 (1987): 18.

    5. Baureiss, 17.

    6. Ibid., 26.

    7. Yee, Blood and Iron, 210.

    8. Baureiss, 21.

    9. Roy, 105.

    10. Yee, Blood and Iron, 220-221.

    11. Ibid., 14.

    12. Ibid., 211.

    13. Baureiss, 17.

    14. Miriam Yu, “Human Rights, Discrimination, and Coping Behaviour of the Chinese in Canada”, Canadian Ethnic Studies/Etudes ethniques au Canada, 19:3 (1987): 118.

    15. Ibid., 114.

    16. Yee, Blood and Iron, 20.

    17. Baureiss, 30.

    18. David C Lai, “The Issue of Discrimination in Education in Victoria, 1901-1923”, Canadian Ethnic Studies/Etudes ethniques au Canada, 19:3 (1987): 47.

    19. John L Tobias, “Protection, Civilization, Assimilation: An Outline History of Canada’s Indian Policy”, The Western Canadian Journal of Anthropology, 48 (1976): 18.

    20. May Yee, “Chinese Canadian Women: Our Common Struggle”, Canadian Ethnic Studies/Etudes ethniques au Canada, 19:3 (1987): 180.

    21. Jean-Guy Prevost, “Immigration, statistics and eugenics: measuring racial origins in Canada (1921-1941)”, Canadian Ethnic Studies/Etudes ethniques au Canada, 28:2 (1996): 2-3.

    22. Baureiss, 23.

    23. Gillian Creese, “Kay J. Anderson, "Vancouver's Chinatown: racial discourse in Canada, 1875-1980" (Book Review)”, Canadian Ethnic Studies/Etudes ethniques au Canada, 24:2 (1992): 128.

    24. Baureiss, 17.

    25. David C Lai, Chinatowns: Towns within Cities in Canada. (Vancouver: University of British Columbia Press, 1988.), 70.

    26. Roy, 100.

    27. Baureiss, 23.

    28. Ibid., 16.

    29. Ibid., 21.

    30. Bennett McCardle, “The Records of Chinese Immigration at the National Archives of Canada”, Canadian Ethnic Studies/Etudes ethniques au Canada, 19:3 (1987): 165.

    31. Edgar Wickberg, “Some Problems in Chinese Organizational Development in Canada, 1923-1937”, Canadian Ethnic Studies/Etudes ethniques au Canada, 11:1 (1979): 91.

    32. Matthew Annis, “The “Chinese Question” and the Canada-Us Border, 1885”, American Review of Canadian Studies, 40:3 (2010): 354.

    33. Yu, 117.

    34. Annis, 356.

    35. Ibid., 353.

    36. Lai, 64.

    37. Baureiss, 16.

    38. May, 183.