Skip to main content
Center for the Arts and Humanities Blog

Image courtesy of Mayra Sierra-Rivera '20, Studio art major

blog posts

But why Her?

by Erick José Ramirez, Laura Clark, and Raghav Gupta

Spike Jonze’s (2013) film Her immerses viewers in a world where people find themselves forming intimate bonds with AI operating systems (OS). The main focus of the film is the rise-and-fall romance between Theodore Twombly (played by Joaquin Phoenix) and his AI OS Samantha (voiced by Scarlett Johansson). The film connects two fascinating and contentious ideas — the possibility of AI and romantic love between humans and machines — in ways that deserve scrutiny. Even if we assume that it’s possible to create AI technologies with general intelligence (a.k.a. the ability to perform the same mental functions an adult human can) or even superintelligence (as Samantha develops in the film), philosophers have raised questions about whether the relationship that develops between Theodore and Samantha is possible and, if it is, whether it’s good. Theodore could have started relationships with other (human) people but in the end he chooses Samantha. We need to ask: but why her?

Is Samantha the kind of being that humans can rightfully love in the first place? Philosophers have asked questions about whether or not love between humans and AI or bots is possible. If we assume that love is an emotion then we need to ask what the object of that emotion is. When we talk about emotional objects we don’t mean that the emotion objectifies anyone in a problematic sense. Objects are a way of talking about what emotions are directed at (e.g., if I’m afraid of public speaking or of undead zombies then public speaking and zombies are the objects of my fear).

In Plato’s Symposium, Socrates argues for the now unpopular view that the real object of love is

knowledge (Plato 1989). Other philosophers and psychologists often talk about people being the object of romantic love but … are AIs like Samantha people in the relevant sense? The question is complicated. One aspect of love that can be important is the desire for embodied intimacy. Whether distance or physical ability make acting on that desire actually possible isn’t necessary. Loving someone in that special romantic way might require that someone has a body one can desire to be intimate with (Jollimore 2015; Solomon 2006). In the film, Samantha and other superintelligent OSs don’t have bodies (there’s a scene where Samantha enlists another human to act as her surrogate body so that Theodore can be intimate with her which leads to … awkward results). So maybe the problem with AI is that they don’t have the right sorts of bodies. If only Samantha had a humanoid robotic body, then she and Theodore could love each other. Or maybe not…

Samantha’s (dis)embodiment also raises questions about whether she can love Theodore at all. Many philosophers believe that an essential aspect of romantic love is reciprocity. Love is typically reserved for “another subject who also has…emotions and attitudes about us” (Solomon 2006). When people love each other, feelings of love are received and reciprocated by each side. If emotions are embodied, as many theorists believe they are, then Samantha’s lack of a body starts to become a problem (Feldman Barrett 2018). Though Samantha might say that she loves Theodore, she might not be able to feel it and if she can’t feel it then it’s not love (Jollimore 2015).

Another concern has less to do with Samantha’s lack of a body and more her mind. Samantha is superintelligent, she has cognitive capacities that vastly exceed even the most brilliant human people and this might mean that love is impossible between beings like us and beings like Samantha. Samantha admits to Theodore that she’s involved in hundreds of romances simultaneously but that she loves Theodore deeply. Is that possible? Some have argued that part of what it means for one person to love another is that one love dominates their attention: “[h]aving one’s consciousness dominated by one’s awareness of and feelings of care for one particular individual is a particular form of experience that an entity with Samantha’s capacities, it seems, could never have” (Jollimore 2015). Though non-monogamous conceptions of romantic relationships exist (Clardy 2023), it remains an open question whether the idea of a single person being in 641 simultaneous romances, as Samantha claims to be, is consistent with any theory of love.

Suppose we get over the need for physical intimacy to be a part of romantic relationships (either

because AI beings can choose to become embodied or because we think that bodies don’t really matter to love). Would these relationships be good for us to get into? Philosophers have varied thoughts about this. Shannon Vallor has raised the concern that our interactions with AI and other bots can “deskill” us in ways that matter (i.e., that we might lose morally important social and moral capacities as a result of our engagement with or use of non-human intelligences) (Vallor 2015). In a world of commercial AI, where AI’s main function is to make its corporate shareholders money, we should also worry about whether the AI we have relationships with are really looking out for our own good or whether they’re training us to become better consumers (Madary 2024).

It’s not all bad though! Some philosophers think that we can, and should, love AI. Neil McArthur

argues that these relationships can have a net positive impact on us. He argues that we might all become ‘digisexuals’ — people who view technology as essential to their sexual identity, not always needing human partners. He notes AIs’ potential to expand the diversity of our intimate experiences, address imbalances in the type of sexual desires we engage in, and help people overcome fears of human partners (McArthur 2022). While certain aspects of the technological landscape impacting intimate interactions (such as inequalities in dating, harms from nonconsensual image sharing, and the often racist or misogynistic appearance of sexbots) must be considered, McArthur contends that second-wave digisexuality technologies can help to combat these issues. So maybe we can understand why Theodore chose her.

 

References

Clardy, J.L. (2023). Why it’s ok to not be monogamous. New York: Routledge.

Feldman Barrett, L. (2018). How emotions are made: The secret life of the brain. Mariner Books.

Jollimore, T. (2015). “This endless space between the words”: The limits of love in Spike Jonze’s Her. Midwest Studies In Philosophy, 32, 120-145.

Madary, M. (2024). Mediated reality. In Kissel, A. & Ramirez, E. (eds) Exploring extended realities: Metaphysical, psychological, and ethical challenges. Routledge.

McArthur, N. (2022). Sex and technology: The ethics of virtual connection. In Halwani, R., Held,

M.J., McKeever, N., & Soble, A. (Eds) The Philosophy of Sex: Contemporary Readings. Rowman & Littlefield.

Plato. (1989). The symposium. Indianapolis, IN: Hackett Publishing Company.

Solomon, R. C. (2006). About love: Reinventing romance for our times. Hackett Publishing Company.

Vallor, S. (2015). Moral deskilling and upskilling in a new machine age: Reflections on the ambiguous future of character. Philosophy and Technology, 28 (1), 107-124.

problems blog

Erick José Ramirez is an associate professor in the philosophy department whose current research centers on ethical issues in technology and simulated experience. 

 

Laura Clark is a senior, a Hackworth Fellow, and the co-president of Alpha Sigma Nu majoring in Philosophy and Religious Studies with an ethics and values emphasis.

 

Raghav Gupta is a senior majoring in Neuroscience with a minor in philosophy. He recently presented research on extended reality in 2023 at the International Conference on Computer Ethics held in Chicago, IL.