June 21, 2024 ; Catergory: essays
Hello! Below is a write up of the notes I took while doing preliminary reading for my future thesis on the feminist approach to sex robots. The text below shouldn’t really be taken as definitive of my thoughts on the topic, nor should it be read as trying to conclusively argue for a specific conclusion. The notes were centred around the question “Should we ban Sex Robots?” but are more about my thoughts on the existing literature on the topic rather than coming down on any one particular side. Here’s what I read for this:
First question to ask is whether we are asking whether sex robots should be banned now, or in an ideal society? One - for example - could believe that sex robot development now is immoral because its development under patriarchy would imbue it with patriarchal and sexist notions within its design but that these notions are ultimately contingent on the existence of patriarchy.
We can consider a weaker and stronger thesis:
This distinction divides the literature into 3 main camps - the strictly positive, the reformist and the strictly negative. I’ll be keeping these in mind from this point forward.
It seems like implicitly all persons in the debate are assuming that the robots we are talking about are strictly non-conscious. I will do the same here - but not because I believe that conscious sex robots would be any better. To the contrary, having a class of conscious being who are owned for the express purpose of being personal prostitutes to their owners is insanely morally bad (should we consider these beings to have rights which seems sensible to do if they’re conscious). The debate wrt to non-conscious robots however seems not only more undecided within the literature in terms of its ethics, but also has practical importance given the actual current development and selling of sex robots.
One question in the literature is whether sex robots are normatively different from the sex toy - and therefore should be considered separately when determining the ethics of using them and using sex toys. As pointed out by both Richardson & Odlind (2023) and Devlin (2020), the history of the Sex robot is inherently interlinked with that of the sex doll, rather than the sex toy. Hence defining sex robots in line with sex toys/as a subset of sex toys does not seem correct. The primary distinction here is the necessity of the aesthetic mimicry of sexual intimacy. Sex toys needn’t (and ofttimes forego) be simulacrums of sex between persons and hence are often abstracted. The primary goal of the sex toy is as functional only. The sex doll - and by extension the sex robot - is interested in the aesthetics in combination with the functional. To remove the aesthetic component would be to no longer have a ‘sex doll’ as we’d understand it.
For now I’ll utilise the definition given my Danaher as the non-pejorative definition I take for sex robots. It is as follows:
It’s important to note that the definition is also agnostic on one important issue: whether the robots are embodied or not. This allows for AI girlfriends, metaverse companions and the like to count as sex robots. This interestingly moves the scope for sex robots away from centring imitations of sexual intimacy. This is explored more when discussing the detriments - but I believe there is good reason to not see sex robots solely through the lens of their use in sexual activity but as apart of a larger project of replacing real women.
The basic idea is that paedophiles could enact their ‘urge’ to have sex with children with these robots in order to prevent harm of real children. This claim assumes that a paedophile’s urge to have sex with children could be satiated with a mimicry of the real thing (see point on companionship). If not, a child sex robot could instead have the reverse effect - giving a paedophile a window into the ‘real deal’ and entice them to enact their actions with real children. A parallel can be drawn here with porn. As Richardson (2023) writes: “Porn provided a behavioural guide to men so that the women in porn might be viewed in ‘two dimensions’ but men who viewed it could act it out on women in their lives ‘their own three-dimensional bodies’ (MacKinnon, 1993)” and as Finlayson (2016) corroborates “What is uncontested is that a large proportion of rape and sexual abuse involves porn: abusers often use or imitate pornography while they abuse, showing it to victims and pressuring them to enact what is depicted”.
Claim would have more force if there were not alternatives to helping paedophiles with their affliction, including CBT and other behavioural interventions (it also seems as though the research on treating paedophiles still has a ways to go - suggesting that other possible intervention methods could be discovered). Given this, it seems as though either there would need to be no problem brought about by the use of sex robots, or the treatment efficiency of child sex robots would need to be huge (notably, this is a claim that seemingly hasn’t been tested - only purported theoretically by those who advocate this usage).
The basic idea of this point is that men would not seek out women to prostitute if they had access to sex robots. The point implies that the only problem with prostitution as an institution is that it harms the women who participate. While it certainly does do that (Finlayson, 2016) there is a greater harm caused to women as a whole by the existence of the prostitution industry. Mainly that the prostitution industry relies necessarily on the commodification of women’s bodies and sexuality as first and foremost a product to be bought and sold. The continued existence therefore of the prostitution industry therefore is not only an admission by society that such a commodification of women’s bodies and sexuality is acceptable, but also that the men who engage with this industry actively engage with and accept this proposition - an accepted proposition that no doubt harms the other women in his life.
The basic idea is pretty straightforward - sex robots should be used for older adults (usually men) to relieve loneliness they may face in relative isolation/loss of a partner. Point implies that a core component of loneliness is a lack of sexual activity - or that sexual activity is at least so important to an analysis of loneliness that treating it without a sexual substitute would be an incomplete solution. Ignoring the trivial case of asexual people, two questions arise: firstly, is it possible for a non-asexual person to live a fulfilling life without sex? If not, would sex with a robot meaningfully solve this issue? Also implies that robots can meaningfully substitute for human-robot relationships. See below for discussion.
Rests on the implicit assumption that human-human relationships can be substituted by human-robot relationships. I do not believe this holds. I take it to be core to the human-human relationship that there is reciprocity involved. This is a whole topic unto itself - and I will certainly be exploring this topic further in another future essay. I’ll simply leave my overall conclusion on that question here for understanding the rest of my point here.
Human-robot relationships in fact corrode the quality of human-human relationships (when they’re admitted into the same relationship status/ontological class as human-human relationships). This is because in order to admit human-robot relationships into the class of valid types of relationships, we must suspend the necessity of reciprocity within human relationships - and therefore reduce relationships to their functional and aesthetic components - effectively removing the ‘humanity’ from our interlocuters.
Would this still apply if we understood our relationships with robots to be normatively different from those of humans? Robots have already acted at catalysts to change our understandings of concepts such as aliveness and consciousness. As Turkle (2011) discusses, children’s understandings of aliveness in the 60s was derived from the basis of independent motion - that their wind-up toy was not alive as they had to induce the toy’s motion. Then with the advent of robotic toys in the late 70s and 80s, the language surrounding aliveness became primarily psychological. Then by 2005, there was a distinct split between consciousness and life - with children bring able to grant that robots may be conscious without being alive. This did not require us to understand robots as anything more than what they were. As Turkle (2011) points out, robots are very good at pressing out “Darwinian buttons” - the internal cues we’ve evolved to look out for in other ‘conscious beings’. Whether intentionally or not, robots are able to hijack our psychology to the point where the line between human and machine to us isn’t as clear as we’d like it to be - a trend that almost certainly will continue to be exacerbated as technologies progress.
Richardson and Odlind (2023) argue that ‘sex robots’ really ought to be considered ‘porn robots’ which are yet another form of pornographic material used (primarily) to degrade women and girls. This ties it in with all the arguments against porn that radical feminists will traditionally make. Beyond this there is an intersection with the general trend in robotics to blend the lines between humans and property - given that these robots exist as both substitutes for real people but also at the same time as commodity - literally. Hence the pushing of porn robots also edges us closer to a normalisation of direct ownership/commodification over (primarily) women’s bodies/sexuality.
One may argue against this view of sex robots as pornography by arguing that pornography is passive rather than active - if we define sex to be “sexually explicit material” then we might want to say that for something to be ‘material’ it must be passively consumed (the person ‘consuming it’ so to speak would be a third party). However I find this a weak rebuttal. For one, VR porn is still by-in-large still obviously pornography, even with a very real sense of interactivity. In general it seems as though with advents such as VR porn, AI girlfriends, and (if you believe them to be in the pornographic tradition) sex robots, the direction porn is taking is towards higher interactivity - it would be weird then to say that porn is transforming into something that ‘isn’t porn’ in the pursuit of what its consumers want.
Richardson (2023) herself would reject this characterisation that somehow sex robots are active. Firstly because it implies that one is doing something normatively different with a robot than when say masturbating while watching porn (this distinction is categorising the act as sex) which cannot be true as the same property relation that holds with porn holds here (property in the sense of ‘consumption of tokens of women and girl’s sexuality/bodies/abuse’). Secondly, and related, is that ultimately it serves the same engagement with sexuality as porn provides: “not as an interpersonal experience, but egocentrically mediated via screens which present others as anonymous third persons.”
There’s also a more general rebuttal which will grant that sex robots are a form of pornography but then go on to argue that porn isn’t harmful - or at least isn’t necessarily harmful. Both Richardson and Finlayson I believe give good arguments for why they are intrinsically harmful but I will leave off the discussion on porn for another time - stay tuned in the near future when I talk about it at length!
Richardson and Odlind’s account still leaves some questions unanswered however. For one, they wholesale reject that ‘sex’ with a sex robot is actually sex - that sex must strictly be reciprocal. Now it should be noted that already we do make a distinction between sex as referring to the set of physical actions that constitute it and sex as an experience. For example, rape is not really considered to be sex. It exists in its own category due to having violated some facet we consider to be important to sexual experience - consent. It still however in the literal physical sense is sex. So in the case of the sex robot while one is mimicking the physical processes of sex, Richardson and Odlind find that it violates a principle of reciprocity that would be necessary for sex. I find this requirement plausible but perhaps more ought to be done to fully justify it - as the issue of whether one has sex with a sex robot comes up a lot in the literature.
Another question is to ask whether sex robots are trying to be more than just another outlet for porn. This thought came from a discussion with my supervisor who brought up the example of a sex doll called “Samantha”, who has a friendly and family mode alongside a sexy mode. This seems weird - if the doll was just a porn outlet why include these two modes? Unless it exists to do more than replace sexual experiences with real women - to aim to push out women in general from men’s lives. This is almost certainly a thread I will be following further to see where it leads.
We’ve spent the previous parts of this post dwelling on the benefits and detriments of sex robots - ultimately trying to come to some conclusion about the ethical implications of sex robots. However, even if we came down on the side against sex robots, that doesn’t imply anything about the appropriate pragmatic response to sex robots. Consider a parallel case in porn (which can seem a sensical move should we follow the argumentation of Richardson) - anti-porn feminists such as Dworkin and MacKinnon do not actually advocate for censorship of porn on the grounds of a misuse of the law (often against queer people) when enforced by patriarchal power structures (Finlayson, 2016). We may wonder then, does the same fear apply when it comes to sex robots?
I’d personally be inclined to say no and the essential difference comes in the vagueness of the definitions for porn and, in contrast, the specificity in which sex robots are defined. The ease of legislators to misuse anti-porn ordinances against other minorities is downstream of a vague definition where there is no consensus and therefore room for plausible deniability in misapplication. As seen above however, it’s not hard to give a fairly cogent definition for sex robots that isn’t particularly controversial. In general I personally find it hard to imagine how the definition above, or any variant of definition for sex robots could in fact be misconstrued in such a way to cause harm through misapplication. It seems then that despite porn being a vaguely defined term, proper subsets such as sex robots can be well-defined within it (this also presents another way in which porn and sex robots may differ).