Abstract
The question of whether technologies based on artificial intelligence (AI) could ever achieve a level of machine intelligence that is indistinguishable from human intelligence hovers over many debates about the future of such technologies. This idea often informs the narratives of cultural products, such as movies and books, that portray AI as able to project both agency and humanity, posing an existential threat to our culture. Some people may find themselves worrying and feeling unsure about the future of these technologies although they seem perfectly content to interact with the mellifluous voices of chatbots, voice assistants, or the speedy responses from social media bots.
Communicative AI—voice assistants and smart speakers, such as Google Assistant, Amazon’s Alexa, and Apple’s Siri—seem to be on the threshold of a confidence level at which most users have decided that it is safe to socially engage with a machine. These technologies can do so because they have successfully deceived users into attributing intelligence to a computer without feeling threatened. This idea can be traced back to the Turing test, which valued computers based on their ability to deceive people into thinking they were human.
In Deceitful Media, Simone Natale, associate professor of Media Theory and History at the University of Turin, argues that deception is a constitutive element of AI rather than an unwelcome by-product. He argues, “Deception is as central to AI’s functioning as the circuits, software, and data that make it run” (p. 1). Being deceived is a response by the user to the affordances coded deep inside these types of technologies.
Natale has conducted years-long research into the cultural myths surrounding the rise of thinking machines, proposing the concept of “banal deception” to explain the mechanisms and practices embedded into these technologies that allow them to be integrated ubiquitously into their users’ routines. As Natale posits, “Banal deception entails mundane, everyday situations in which technologies and devices mobilize specific elements of the user’s perception and psychology . . . to attribute agency to things or personality to voices” (p. 6). Yet, banal deception does not treat audiences as passive or naïve. Characteristics such as a humanlike voice or specific language expressions are meant to produce calculated reactions from users, although Natale argues that the audiences are “actively exploiting their own capacity to fall into deception” (p. 7) while remaining completely aware of the differences between humans and computers.
Throughout Deceitful Media, Natale demonstrates that banal deception is utilized by those who produce and code these technologies to achieve a positive, engaged reaction from the audience. The book presents five key characteristics that allow communicative AI to be a meaningful, useful tool.
First, from ELIZA, considered the first chatbot in history, to Siri and Alexa, AI has had a typical character, at least as far as users are concerned. Recent literature indicates that audiences find talking AI to be more ordinary than expected, rendering these technologies imperceptible to humans rather than causing users to feel that they are conversing with a robot. Second, success in deceiving users begins with the functionality of these technologies. The audience can establish empathetic sentiments with domestic technologies, such as Jibo or Replika, because these comprehensible systems represent pragmatic benefits for users, while the complexities of coding and software are hidden behind the cuteness and friendliness of the voice assistant. Third, users approach these technologies obliviously. For Natale, the audience may engage with these technologies both deliberately and unconsciously, maintaining (in their minds) a level of control over the hyper-friendly assistant. Fourth, and unlike humanoid robots, disembodied voice assistants have a low-definition characteristic, which requires an extra level of involvement and participation from the audience. As Natale argues, “Voice assistants encourage users, by relying on them to apply their own stereotyping, to contribute actively to the construction of sense around the disembodied voice” (p. 114). Finally, and perhaps most important, banal deception highlights the agency behind those who design, develop, and program this software to understand the effect that these technologies have in the real world (p. 72).
Perhaps this is one of the most relevant arguments raised in Deceitful Media—namely, the question that we should be asking is not necessarily whether these computers and AI technologies will be more intelligent than people. Rather, Natale argues, we “need to consider the cultural and social consequences of deceitful media providing the appearance of intelligence” (p. 15). Deceitful Media challenges the reader to wonder how communicative AI may change not only our relationships with the machines that we integrate into our everyday lives but also our relationships in our social lives. Perhaps the predominant challenge will not be whether machines become smarter than people but that we may lose the ability to differentiate between the two.