«‘Love with robots will be as normal as love with other humans’ 1 Simone Hinds-Addow BSc Psychology Patterns of Action Dissertation May, 2014 ...»
‘Love with robots will be as normal
as love with other humans’ 1
Patterns of Action Dissertation
David Levy (2007)
Could you fall in love with a robot?
First, what is love?
And the necessary and sufficient conditions for falling in
love with a Robot?
Anthropomorphism and Humanity
Life and Intelligence
Emotions and Reciprocity Have they been met or is there still some way to go?
References Could you fall in love with a robot?
Revolutionary advances in computer software and processing speeds have led to the formation of increasingly intimate relationships between humans and machines;
people take their machines to bed, wear them to work and spend a great amount of time with them each day. As they become more sensitive to our feelings, more in tune with our desires and more gifted at expressing themselves, humans find themselves approaching an age where they may use machines not only for access to one another, but for companionship. This has led to the idea that there could come a time where people fall in love with these ever more artificially intelligent machines.
For many psychologists, who study this relationship, this isn’t a matter of if, but rather when and for artificial intelligence pioneer, David Levy that time is 2050. Levy argues that as robots become more human-like in appearance, personality and function, the relationship between robots and humans will become more personal and inevitably result in love. Whilst robots could become increasingly enticing lovers, many doubt that a robot could ever provide someone with the deepest form of human love. Nardi (2013) contends that "While a computer program could provide a captivating virtual romance… we are still going to pine for the real thing”. Robots could well provide the intimacy and affection that humans crave but will people ever be happy with that? It may be that having a partner who is a living and breathing human being, in all their imperfection and lack of programmability is an essential part of love.
First, what is love?
Evolutionary Basis The mechanisms by which people fall in love and stay in love could provide some insight as to whether humans could indeed fall in love with robots but, in order to do so we must first define what love is. David M Buss (1988) presented what is an evolutionary approach to love on the basis of two main premises. The first is that love does not reside solely in a person’s subjective thoughts, feelings and drives. Buss argues that love rather results in actions with tangible consequences and in this way forms a complex suite of adaptations. These adaptations or psychological devices are designed to solve specific problems of survival and reproduction such as exclusivity, fidelity, commitment and parental investment (Buss, 1988; Trivers, 1972). For example, a jealousy adaptation can alert an individual to increased risk of a partner’s infidelity (Buss, 2000). If a human were to fall in love with a robot however, many of these psychological devices would no longer be required. There would be no jealousy, no heartbreak and no cheating, and in this way a robot could become a desirable lover for a human.
The second premise postulates that the key consequences of love are centred around reproduction and despite a lack of empirical evidence, this provides major challenge to the question addressed in this discussion. Forming a relationship with a robot would diminish the opportunity to produce and raise offspring thus, if love has evolved to support reproduction and promote parental investment what would be its purpose here? Nevertheless, many couples may decide not to reproduce (Hakim,
1998) and many others may suffer from infertility issues. Yet, these couples still fall in love and enjoy long-lasting loving relationships. This could potentially be the case for a human- robot couple also. Furthermore, if a human-robot couple were to have a child by means of surrogacy or adoption, would love then become an adaptive emotional response?
Evolutionary psychologists further argue that love must be universal and suggest the presence of psychological circuits specifically dedicated to love within humans. This provided a starting place for neuroscience research (Miller, 2001).
Which has gone on to support the notion that love arises from distinct systems of neural activity (Damasio, 1999) and Helen Fisher (2002) proposes three core brain systems for falling in love; lust, attraction and attachment (Fisher, 1998). The initial feeling of lust is seen as the internal craving for sexual satisfaction. This ignites thoughts and feelings of sexual desire (Sherwin, 1985) and motivates individuals to seek sexual union with genetically appropriate partners. For humans, genetically appropriate partners are assumedly other humans and so this poses the question as to how a robot could be an appropriate partner if it does not possess any genetic material? However, animals could also be considered as genetically inappropriate partners. Yet, many people develop sexual attachments to them (Kraft-Ebbing,1886) and cross-species sexual activity between human and non-human animals also occurs (Laws & O’Donohue, 2008). This suggests that feelings of lust and sexual desire could perhaps be extended to robots as well.
Following on from lust is romantic attraction and characterised by increased energy and feelings of love, this focuses attention towards a preferred mate. Preferred mates are chosen on the basis of dependability, maturity and kindness (Buss, 1994) and characteristics like these could be easily programmable. Hence, this is one way by which robots could appear attractive to humans. However, physical attraction is regarded as a prerequisite for mating (Gallup and Frederick, 2010) and research shows it is of huge importance when deciding who to build relationships with (Duck et al, 2007). For this reason, robots would need to be able to evoke that initial physical attraction when they first come into contact with humans. The final result of lust and attraction is the formation of long-lasting attachments between partners and Bowlby (1969) defines attachment as lasting psychological connectedness and emotional union. Many people report having such attachments to technology, many name their devices and for some losing these devices can be hugely distressing.
Whether these attachments could lead directly onto feelings of intense romantic love for robots however, still remains to be seen.
The question of whether robots could convincingly reciprocate intense romantic love also arises here though, biological explanations can explain why reciprocity may not be necessary. Neuroimaging research has identified the specific neural circuitry of lust, attraction and attachment (Bartels & Zeki, 2000) and if correct, love is simply a series of neurochemical changes in the brain. Therefore, biochemically inducing these changes in a human whilst in the presence of a robot could in theory make them fall in love. Equally, we could also study individuals who claim to be in love with a robot by placing them under functional magnetic resonance imaging (fMRI) to see whether there is activation of the brain areas, implicated in the neurochemical changes associated with love. If this activation is similar to that seen in individuals in love with humans, this could support the possibility of humans falling in love with a robots.
Robert Sternberg’s (1986) triangular theory of love suggests a conceptualisation of love also consisting of three components; Intimacy, passion and commitment (see figure 1) and together these three components form what Sternberg describes as the greatest love of all - consummate love. Characterised by feelings of closeness and bondedness, intimacy is seen as the core component of love (Sternberg & Gracek, 1984) and emotional in its composition it also includes transparency, vulnerability, dialogue and communication. And dialogue between humans and robots in particular, has been possible for a long time. Joseph Weizenbaum (1966) designed one of the first computer programs (ELIZA) capable of producing natural language conversation and by responding with vaguely empathetic questions ELIZA was able to convincingly play the role of a therapist. So much so, people treated it as though it were a person and often felt more comfortable disclosing intimate details about themselves with the program.
However, ELIZA was simply the product of clever social engineering and so was unable to engage in sustained conversation. To achieve this a robot would require the ability to understand others, put oneself in the place of others and share personal thoughts and emotions. But if robots are incapable of genuinely having such thoughts and emotions how could they become intimate with humans? It would also be difficult to programme a robot to identify when to initiate behaviours such as kissing, hugging and holding hands and it is unlikely that this would be comparable to a human. Therefore, if intimacy is indeed a core component of love the field of human-robot interaction must address the way by which people can ‘form deep and meaningful psychologically intimate relationships’ (Kahn et al, 2010) with robots we can fall in love with them.
Primarily motivational in character, passion is the psychophysiological arousal that is brought on by another person which leads onto romance, physical attraction and sexual consummation. Sexual consummation between robots and humans is highly controversial and begs the question as to how humans could connect with robots, if they lack the characteristics which allow us to connect most intimately with each other? Well, many women use machines such as vibrators for sexual pleasure and there are already companies that sell realistic sex dolls. It may just be a case of combining the two together, and the technology for this is already there (Levy, 2007). The question of why people would want to have sex with robots in preference to, or in addition to, sex with humans also arises here but for those who have difficulty attracting a mate there is a clear benefit. This would reduce the risk of rejection and could provide a way to satisfy the universally human need for sexual intimacy. However, this does pose some ethical considerations, one of which is that people who might be violent or extremely perverse in their sexual relations with robots could be encouraged to be violent and perverse with human beings also.
The final component, commitment is primarily cognitive in nature and encompasses the decision to love someone and maintain that love for a long time. Commitment is associated with compromise, sacrifice and restricted freedom of action and these are aspects of human relationships that many people dislike. Forming a commitment with a robot however, would reduce these and for individuals who find the decision to commit quite difficult (Carter, 1987), a relationship with a robot could be a more appealing and less fear-inducing alternative to a human counterpart. Instances where humans have made commitments to virtual characters are not uncommon and reports of one male who went onto marry his virtual girlfriend have also surfaced (Moore, 2009). As a result, it has been suggested that similar relationships with robots may not be as far off as we’d think.
And the Necessary and sufficient conditions for falling in love with a robot.
Anthropomorphism and Humanity The psychological theories described above refer to love as a feeling that occurs only between human beings. Thus, it has been suggested that the humanness of robots becomes a necessary and/or sufficient condition for falling in love with them. People generally talk about machines as though they can love, fight, flirt and think. And anthropomorphising technology in this way, as though “it” were a “he” or “she” with thoughts and feelings could be a precursor for seeing machines as having their own personalities, intentions and motivations. Paired with robots which are visibly human-like, this could cause humans to fall in love with robots and believe they can love them back. And there is a trend of robots in becoming more human-like.
Duffy (2003) suggests that using a humanoid design with facial expression, movement and social interaction is an obvious strategy for effectively integrating robots with humans. The robot head is assumed to be the primary area of humanrobot interaction and so has become the focus of much research. One such study by DiSalvo et al. (2002) analysed 48 humanoid robot heads and asked participants to rate them for how human-like they were. The ratings showed that facial features such as the nose, eyelids and mouth, as well as the number of features significantly increased the perception of humanness however, the majority of the robots were not rated as being very human‐like. This questions whether an anthropomorphic design is enough to allow humans and robots to be indistinguishable from each other but Duffy (2003) believes that in order for a human and a robot to engage in meaningful social interaction, a robot with several anthropomorphic features is essential. A robot that has too many anthropomorphic features however, will lead to overly optimistic expectations about the robots abilities, causing disappointment. So Duffy suggests that the role of anthropomorphism should not be to build a “synthetic human” but instead to design robots whose appearance matches their perceived ability. For example, if current humanoid robots are expected to create several bodily movements with the use of their limbs then robots that do so will be more readily accepted.