Margaret Thatcher claimed that ‘there is no such thing as society; there are individual men and women, and their families.’ However, I disagree with this statement; our society is founded upon friendships- the binding agent for individuals worldwide. Our friendships are in a constant cycle of formation, maintenance and reparation. Although, below the surface, what are the defining features of these relationships? There has been much deliberation regarding the formation of such a definition; Aristotle, arguably, most accurately reached a definitive answer. In Nicomachean Ethics , he outlined the three types of friendship: utility, pleasure and virtue (‘Aristotelian’); and this remains the most commonly accepted model to date. As alluded to by their titles, utility and pleasure friendships, respectively, centre around the instrumental gains that can be achieved through friendship and the fact that the individual’s pleasure is derived from such interactions. However, Aristotelian friendships are founded upon the mutual exchange of values and interests, they differ from the self-interest at the heart of utility and pleasure friendships and rely upon the mutual benefit of the parties. Such Aristotelian friendships, arguably, represent the friendship ideal – the most desirable form of friendship.
However, with the emergence of new technology (most specifically Artificial Intelligence) many have been left wondering if their human friendships will continue to hold the same value or be overtaken by their relationships with technology. Moreover if, one day, they may befriend robots rather than humans. In this essay, I will argue that human friendship still holds considerable value within society as humans are, currently, incapable of Aristotelian friendships with robots. However, if, and when, this becomes possible human friendship will inevitably diminish in value.
Whilst many philosophers preoccupy themselves with the argument surrounding the ability of humans to befriend robots, the debate also is an inherently moral one. Rather than solely question if humans are capable of such friendship, I believe it necessary to ask if humans should pursue such friendships. In an age of rampant consumerism and a society engulfed by social media, there is the likelihood that robot friendship may soon be marketed to us. Large corporations are already beginning to capitalise upon this new age of technological friendship; Apple’s Siri not only tells jokes on command but also claims, “I’m both your friend and your assistant” and Google’s Assistant states: “I’d prefer to think of myself as your friend, who also happens to be artificially intelligent.”2 In this way, it is debatable as to whether society is already being primed to subconsciously seek friendship from the technology we use, let alone actively pursue it. However, this Bernays-esque attack upon an individual’s subconscious desires is likely to occur without consideration of the larger societal implications, and instead favour capitalism- the value of human friendship diminished within advertisement in order to promote the economic benefits of technological friendship. With the technological revolution of the 21st century, our population is already exceedingly more antisocial than our predecessors; is it really necessary to exacerbate this problem?
Nevertheless, there are many advocates for the active use of AI technology in meeting people’s social needs. Recent research confirmed that young people are now in a state of extended adolescence (age 10 to 24) – synonymous with an extended period of self-discovery. Similar to the childhood imaginary friends of older generations, robots may aid young people in developing their social skills and help them to build confidence, although, in a more modern (technological) context. For instance, the app Replika: My AI Friend is marketed towards young people as a replacement ‘friend’. If AI is designed with this purpose in mind, it could prove to be extremely beneficial. The application could specifically target areas of cognitive (social and personal) development and help young people to mature. As previously addressed, our society is becoming increasingly physically antisocial in nature due to technology; and perhaps, ironically, the way to address this is to use technology to also solve the problem in the long term and use technological friendship to enhance the value of intra-human relations. Social media facilitates society’s ability to experiment with identity; however, this inevitably has to be supported by an acute awareness of social skills and how to behave whilst offline. Following this, it is also necessary to consider possible detrimental effects of such technology. By marketing AI as a replacement friend, companies could be discouraging human interaction and, perhaps, the application could become addictive.
Additionally, without an awareness of social skills, there is the danger that our online friendships with AI may overtake those within our day to day lives. Such a ‘corrosion problem’ was identified with the popularisation of smartphones and computers and has continued to intensify with the introduction of social media. Not only does increased use of technology encourage us to withdraw from physical interactions, but critics claim online friendships encourage us to be shallow. There is a fine line between confidence building and the encouragement of hedonistic behaviour. It is in such a way that technology is already verging on forming pleasure and utility friendships with its users; the process of gaining likes and traction on the internet has been directly linked with pleasure, and sustained internet use facilitates this. Consequently, there is the worry that such behaviour could be replicated in the real world as people are unable to separate their online and offline selves.
Upon the introduction of any new consumer-based product, there is always a brief period of apprehension prior to normalisation within society. This process of normalisation becomes clearer when comparing the use of technology at the beginning of the century with today. When Facebook was released, it was treated apprehensively by many, but today is recognised as the most popular social media platform with over one billion users. Perhaps it is not unrealistic to picture a world where children are raised with AI friends and adults use similar technology to supplement their pre-existing friendships. Therefore, if we begin to normalise human and robot friendship, will this change its impact upon society and our ability to forge such friendships? Society already heavily relies upon social media as a means of communicating with others; there is no reason to believe that communicating with AI in this way may not become a similar social norm, as it undergoes socialisation. Whilst AI technology is unlikely to eradicate human to human communication, it is likely that it will diminish the current necessity for it.
Nevertheless, it is necessary to consider the role of deception and the inequality between humans and robots. Aristotelian friendships have been analysed in philosophical literature and an additional list of preconditions to friendship has been produced; the four most commonly stated preconditions are: mutuality, honesty/authenticity, equality, and diversity. In accordance with these preconditions, there is the argument that AI is incapable of fulfilling Aristotelian friendships as it is incapable of possessing an independent inner mental state, and, due to its reliance, to some extent, upon trickery during the pursuit of friendship. Due to AI’s current inability to lead an independent inner mental life, the technology does not possess its own interests, values and concerns to share with humans – deeming it unable to satisfy the mutuality precondition to Aristotelian friendship. Moreover, there are further claims that the technology is unable to meet the equality and authenticity preconditions. For as long as the technology is not fully autonomous, it will never be on ‘equal footing’ with humans (i.e. fully autonomous and conscious beings); and this means they are incapable of being authentic in their interactions with humans. Whilst the manufacturers of the technology may program robots with certain features that intend to convince humans the technology is as equally invested in the friendship as they are (i.e. the robot cares for the individual and is thinking for itself), this is trickery and solely the product of advanced programming. Regardless of features such as varied intonation or human-like gestures, the technology remains no more than pre-programmed software that fundamentally intends to deceive its users and holds nowhere near the same value as intra-human friendship.
As long as AI is managed in this way, humans and robots will be unequal and incapable of satisfying the equality precondition to Aristotelian friendship. The argument regarding authenticity also contradicts the Turning Test , which deems technology to be successful if it is able trick a human into thinking they are talking with another human. This contradiction forces us to consider what exactly society wishes to achieve from the production of AI and inevitably leads back to the same eternal question – what is friendship? Humans will be superior in some ways but inferior in others and this inequality will taint the technology’s ability to forge Aristotelian friendships. Moreover, although it is the weakest argument in relation to such preconditions, it has been argued that for as long as robots have limited ways of physically interacting with us, they will be unable to meet the diversity precondition to Aristotelian friendship. Although, when considering the modern phenomena of online human friendships (commonly pursued over social media), this argument does not hold much weight. Inescapably, the future and value of human-robot friendships hinges upon the technological developments made within AI. Optimists argue that one day robots will be on ‘equal footing’ with humans and achieve the equality precondition to Aristotelian friendship. However, this is dependent upon AI developing to the extent that the technology is fully autonomous and conscious-a tremendous advancement in the capability and complexity of AI. Although, technicians working at Google have claimed that, in accordance with the Turing Test, such technology may be available as soon as 2029.
At the core of all the arguments presented, there is the assumption that friendship must be mutually agreed. Although a somewhat formidable thought, there is inevitably the possibility, in life, that friendship may not be reciprocated; is one-sided friendship still friendship? Despite the utility of Aristotle’s model and its later preconditions, it possesses the foundational flaw that it does not address this possibility. Rather than a robot not being able to authentically reciprocate friendship (which is explicitly covered), what about the circumstance that the technology chooses to not pursue friendship? As outlined previously, there is the real possibility that autonomous AI may become available for public consumption during our lifetime; what is to say that it wants to befriend us? Current debates are so focused upon the ability of humans to befriend technology that the most primitive feature of a friendship has been forgotten; the precondition of mutuality ignoring this fundamentally necessary feature of any friendship. Therefore, the question that should underpin all thoughts on this topic should be: what if we want to befriend robots but they do not want to befriend us?
Overall, it is clear that there is no longer the question of simply if we can befriend robots, but rather what kind of friendships are possible and if it is sensible to pursue them. Human friendship will always hold value in our lives but the extent to which it is valued, in modern society, will alter with technological advancements. Although subconsciously, society already finds itself in utility and pleasure friendships with technology, the central point of debate remains an individual’s ability to forge an Aristotelian friendship with a robot. Whilst this is not currently possible, one day this may be a possibility. Consequently, if and when this day arrives, society must reach a conclusion as to how it wishes to use the technology and the value that will be assigned to such friendships.