To Plug In or Not to Plug In? The Case For and Against the Experience Machine, with help from pop culture
High in the halls of the kings who are gone
Jenny would dance with her ghosts
The ones she had lost and the ones she had found
And the ones who had loved her the most.The ones who’d been gone for so very long
She couldn’t remember their names
They spun her around on the damp old stones
Spun away all her sorrow and painAnd she never wanted to leave, never wanted to leave
Never wanted to leave, never wanted to leaveThey danced through the day
And into the night through the snow that swept through the hall
From winter to summer then winter again
'Til the walls did crumble and fall
‘Jenny of Oldstones’, a song from 'Game of Thrones'
In 1974, Robert Nozick famously introduced the Experience Machine thought experiment in his book Anarchy, State and Utopia.
As in the experimental philosophy tradition, the Experience Machine elicits responses about what people would do, but does not consider, normatively, whether people ought to plug in. With advances in virtual reality, questions about the permissibility and morality of the EM may have practical relevance in the very near future.
Here, I argue reality itself is not prudentially valuable, and with it, present a new interpretation of the EM. Elucidating the key issue of the EM would be relevant to discussing whether we should plug in, which I will then consider.
1. The Nozickean Experience Machine and Nozick’s claims
Nozick describes his EM, and the person inside of which, as such:
· The EM would “give [him] any experience [he] desire[s]”. (1979, 1989)
· He would have “a lifetime of bliss” (1979), “[feeling] pleasures […] from the inside” (1989)
· He “won’t know that [he’s] there; [he’ll] think it’s all actually happening” (1979) and he “will not remember having done this”, not “[realising pleasures] are machine-produced” (1989).
The first two features implies that (1) The experiences are what he desires; (2) The experiences have, ultimately and overall, positive affect for him (“happiness[1]”). Nozick thus assumes people desire and will choose happy experiences. This seems reasonable.
(This also allows for experiences which may have lower-order unhappiness or suffering, such as needing to overcome obstacles and work hard, but which ultimately, and overall, be considered by that person to be happy.)
Perhaps the most important feature, however, is that what we experience in real life, right now, would be exactly what the person experiences in the Machine; the EM’s ‘reality’ is indistinguishable from actual reality.
Using what he believes our choice regarding the EM is, Nozick claims “[external] reality […] is valuable in itself” (Nozick 1989, p. 109, his italics), and it is necessary to “constitute the best life” (p. 106). In other words, it is the objective external reality itself, not us, which has intrinsic value for one’s wellbeing — how well one’s life is going for the individual whose life it is (Sumner 1996, p. 20). I will now consider this claim.
2. Reality itself does not matter: What the Experience Machine is really about
Nozick uses the “happy life of deception” argument (Nozick 1989, p. 107) to show external reality has value. A similar version is as follows:
You live a happy life, with a wife, a loving family, friends, and successes. What you don’t know is actually, your ‘friends’ laugh at you behind your back, your ‘loved ones’ detest you, your wife is secretly cheating on you, and your awards are fakes, jokingly given to you. You never find out, and die happy.
The question is, did your life go well for you?
Of course, most would rather not be deceived, all things being equal. The real issue, however, is even though deception is not valued, whether it is the objective external reality that is valued.
How could some other reality, even the actual one, possibly affect your conception and consideration of your wellbeing if you never know and experience it at all?
Thus, the “happy life of deception” thought experiment is misleading: When we, a third party, are asked to evaluate the happy life of deception, we know it precisely to be a deception. We correctly consider the deceived was wronged, that there is an another, actual reality he is not privy to. (And it also assumes that this person should — a very Kantian morality).
Still, it is different from whether his life went well from his perspective. A deception would only affect his conception of his life only at the point the deception ends for him, and the actual reality is revealed to him; when he realises the reality A he had experienced and the actual reality B he is now experiencing are inconsistent. Otherwise, his reality still remains as that of a happy life for him, period.
Therefore, I argue we value not some objective external reality, but the relation between our experienced reality A and the actual underlying reality B we could also experience (If we are not being deceived, reality A is reality B). We value them being consistent, and value not being deceived or played, or at least, we are shocked by the revelation and disruption of our status quo.
The ability to also experience reality B is crucial; otherwise, we can simply claim, out of nowhere, our current lives are not consistent with ‘reality’, when there is no evidence, or more importantly, impact of that other reality. Such a claim is not only meaningless but dangerous. If we cannot know and experience reality B, then any and all possible versions of reality Bcan be accepted, even conflicting versions.
The idea behind external reality being valuable in itself is that unexperienced harms could somehow negatively affect one’s well-being. However, unexperienced harms do not distinguish between actual ones and fictional ones, which can be unlimited in number and extent by our imagination, resulting in the ridiculous conclusion our wellbeing would be infinitely negative. It also implies unexperienced benefits, similarly infinite in actuality and possibility, can result in an infinitely positive wellbeing. Therein lies a contradiction.
To further show why our experiences do not need to occur in some external reality to be valuable, consider Jollimore’s “negative Experience Machine”, where a person plugged in is made to believe and experience torture, instead of happiness. If the simulated pain feels just as real and tormenting as pain caused by actual torture, it would be hard to claim the person’s life is not as bad as someone suffering actual torture, that her experience is “not intrinsically disvaluable at all” (Jollimore 2010, p. 334), just because it is simulated.
Therefore, if the EM is working perfectly (we only experience an internally-consistent ‘reality’ within, indistinguishable from our own), and we never know we are actually in an EM (that we are being ‘deceived’ from the real world), the EM’s ‘reality’ is the reality to us from the start. And if we will never know or experience the actual reality ‘outside’, with our memory of being plugged in erased, there would not be another reality for our EM-reality to be inconsistent with.
This explains why surveys have shown that while a majority of us would not plug into an Experience Machine, a majority of us would not plug out of an Experience Machine either, if it turns out we were living in one. This preference for our existing state, or status quo bias, is powerful, and any deviation from that, unsettling and undesirable.
(As for what we remembered from our real lives previously before plugging in, perhaps we can design the EM like in the Rick and Morty episode Mortynight Run, where Morty is plugged into the life of ‘Roy’, and starts his simulated life by waking up and interpreting what happened before as a dream).
Nozick may have thus conceded too much: if he was testing for our valuing of some objective, underlying reality itself, then when we are in the EM, we should know there is another underlying reality inconsistent with ours, and we have a chance to also experience that underlying reality ‘outside’. The Matrix’s Neo and The Truman Show’s Truman only seek to escape their false realities when they realise some deception or inconsistency: Truman notices a spotlight falling out of the sky, and Neo meets Morpheus.
This is not to deny, ontologically, there is an objective, underlying reality that exists when one is in the EM, or that deception is not morally wrong unless discovered, only that it is our conscious experiences[2] of being connected to someconsistent reality, not necessarily the ‘actual one’, which affect how well our lives go for us.
Therefore, if the EM does not preclude our connection to reality in a way[3] that affects our conception of how well our lives are going, the actual issue is not whether we should lead real lives, but which reality should we choose. Should we choose a reality perfectly limited to only happy experiences we want, over our current imperfect reality? (‘perfect’ refers to being perfected to ensure only desirable and happy experiences. To highlight the distinction, I will assume our current reality also has undesired, unhappy experiences.)
There are thus two aspects of the EM we should consider normatively: (1) having a reality that only has desired, happy experiences; and (2) being able to choose our reality. (For this discussion, our agent will not have any voluntary obligations to other people in the real world, just as in Nozick’s formulation.)
3. The Case for a Desired, Happy Reality
Weijers makes an insightful observation regarding Nozick’s EM: it is assumed we live in an average but real world (Weijers 2014, p. 13). What if one lives a real but painful existence?
For these people, perhaps they ought to enter the Machine, for the Machine life may be the best available for them (Belshaw 2012, p. 576). Consider the episode San Junipero in the science-fiction anthology series Black Mirror. San Junipero is a simulated reality which the protagonist Yorkie frequently visits. It turns out, in the real world, after her family had reacted poorly to her coming out, she crashed her car and became fully paralysed from a young age, and has remained like this until old age. One would be hard-pressed to deny Yorkie from entering San Junipero. She has been confined to bed, completely unresponsive, for the better part of her life. Her existence in the real world is limited, without any fulfilling experiences (it is quite obviously not a good life), while she can at least have happy and desired experiences in San Junipero (if not the good, at least a better, life, even if it is simulated).
It seems reasonable some happiness is required for a good life, if not exclusively, too much pain would be counterproductive, and our desires need to be fulfilled to some extent for a good life. For people who are excessively denied these in their current lives, the EM is permissible, even moral, as it allows them to escape their bad lives in the real world, and live a better life in a simulated world.
(That said, it does not justify the EM specifically, which has only desired and happy experiences. One could potentially program another Machine with also undesired and unhappy experiences. I will consider this in Section 6.)
4. The Case for Choosing Our Reality
Still, if we could attain experiences we had not, or would not, experience in the real world, the EM may further be justified on grounds of fairness: to allow us to experience what are unable to, because of arbitrary contingencies we happen to have.
It is undeniable there are many external factors that affect our happiness, but which are not our doing, such as our socioeconomic circumstances, family and cultural background (Thomas Nagel identifies this as circumstantial luck). These can result in different experiences and thus varying (un)happiness. For Yorkie, her family’s strict beliefs would have prevented her from achieving happiness fully. One could even argue what had happened was due to random circumstances (resultant/consequential luck): events which affect our happiness could transpire in different ways because of arbitrary conditions, also outside our control. Perhaps Yorkie’s accident was caused by a wet road or malfunctioning streetlights. If it weren’t so, she would not have been paralysed and deprived of happy experiences. Even if the crash was caused by Yorkie’s own actions, it is difficult to justify she deserved lifelong paralysis due to one instance of recklessness. (She could even plead constitutive luck, where her personality is uncontrollable to some extent.)
As Rawls argued in A Theory of Justice (1971), if many of these factors are outside our control and given to us through the random lottery of life, they are arbitrary from a moral point of view. We cannot claim to deserve them, nor the happiness, and unhappiness, which result from them. It is therefore unfair for some to be unhappy, through no fault of their own; everyone should have the fair chance to attain happiness. The EM would thus be an exciting opportunity to make the pursuit of happiness more egalitarian, by correcting for these arbitrary contingencies that determine our happiness, albeit via simulation. The difference is the (simulated) reality in the EM is specifically chosen by us, to be more favourable and conducive to our pursuit of happiness, rather than being a matter of chance.
This brings us to another consideration: freedom of choice. Interestingly, while Nozick thinks most of us would not plug in, he would not view those who do as being immoral. Nozick wrote Anarchy, State and Utopia as a fierce defence of libertarian principles — we have the right to choose what we want in our lives, provided it does not infringe on the equal rights of others.
The more important concern about freedom is whether we should allow someone to choose whatever experiences they want. If we bring the libertarian ideal to its fullest expression, we should allow them to do so for any experience: It has been suggested the EM can simulate desired but socially unacceptable experiences, such as murder, which would be preferable than to allow murder in the real world (Sales 2013, p. 18). Since this is merely a simulated experience, it would not affect any real person’s utility or rights. The EM would thus be the ultimate realisation of freedom and utility.
However, there is something uneasy about allowing someone to indulge in any experience, even simulated ones. Our unease perhaps reveals the source and substance of our happiness is also relevant to what we consider should be part of a good life, not just the affective aspect. It would be hard to argue such experiences, even in simulation, are indeed part of the good life. I will now consider reasons against plugging in.
5. The Case Against Choosing Only Desired and Happy Experiences, or The Case for Undesired and Unhappy Experiences: Dignity and the Human Condition
Immanuel Kant would object to choosing whatever experiences we want. His objection, however, would be categorical (for all experiences). Kant defends human dignity: we should treat people, including ourselves, with respect, as ends-in-ourselves and not as mere means. If we plug in, we would be treating our bodies as mere instruments to attain the happiness we desire. We are “[dishonouring] human nature… and [making] humanity an instrument for the satisfaction of [our inclinations]” (Kant 1756, p. 414).
How perfecting reality, to have only happy and desired experiences, would affect human dignity relates to a larger question about what it means to be human — what features of the human condition have value and should thus be included in our lives, and whether the EM may exclude these valuable features. One can draw parallels with the philosophical debate over human enhancement, where both practices seek to improve our reality (either genetically, or in this case, our circumstances in simulation).
One possible feature undermined by the EM is vulnerability — our ability to suffer — and our struggle to deal with this vulnerability (Juengst & Moseley 2019). Since we have only happy experiences, by definition, we do not truly suffer. If the Machine is programmed specifically for you, it would likely operate within your overall capacities, so there is no true struggle. You would not need to truly strive, to put in true effort to go beyond your limitations, and in so doing, achieve true growth. Some unhappy and undesired experiences which may be excluded are insurmountable or unrecoverable failure, permanent farewells, and even dealing with our own mortality (In San Junipero, people’s consciousness can continue forever; a form of immortality). Without some true suffering and unfulfilled desires, life may thus be ‘too easy’. Perhaps this loses something valuable in the human experience: to be denied happiness and have unfulfilled desires forces us to appreciate what we have, conducing us to be humble and avoid hubris (Sandel 2004, p. 60).
Furthermore, in a reality programmed for you, and only you, you may not develop the virtues you would have otherwise needed and developed in a reality with other real people (Sales 2013, p. 19), who also have (possibly conflicting) desires and interests you would have to consider equally with yours.
In the Black Mirror episode USS Callister, the protagonist Daly simulates and exerts control over his reality. He gets only what he wants, unfettered by anyone and sycophantically encouraged by enslaved digital clones. He becomes oppressive and cruel, lacks any empathy (for he does not need it), even suffocating one clone until she relents to obeying his commands. He becomes “an a**hole god”, as his one of his virtual crew describes him. Since only his desires and his personal happiness matter in his reality, there is no need to work collectively with others, even his digital clones. Daly does not develop the solidarity necessary to manage and reconcile everyone’s desires and interests, not just his own. If humans are social creatures meant to co-exist with other humans, perhaps we should not live a life that only fulfils our own desires and achieves only our own happiness.
Appealing to arbitrary aspects of the human condition, which could (and have) changed — with advances in medicine and hygiene, our vulnerability has significantly decreased — is questionable, and the extent to which these values should be expressed in our lives, especially if they also have negative effects, is open to debate.
However, they still represent the person we may want to be, given who we have been (Juengst & Moseley 2019). To completely abandon an imperfect reality for a perfect one in the EM is to abandon who we have been, and to deny some aspects of the good life. Echoing Nozick’s worry, “is [the person in the EM] courageous, kind, intelligent, witty, loving? […] There’s no way to tell.”
In the EM, these virtues of the good life may never be cultivated.
6. Experience Machines and the Good Life
Ultimately, whether one should plug in depends on one’s conception of the good life, and what aspects should be part of that good life. Life in the EM would be the ultimate realisation of libertarianism and utilitarianism — to choose our desired ends and live a purely happy life.
Yet, life in the EM is still but one conception. There may be other conceptions — involving unhappy experiences, unfulfilled desires, and undesired experiences — that better reflect the full spectrum of the human experience and the full capacity of the human condition, but which would not be available in the EM’s perfect reality.
So, if the EM can simulate any reality, there is no reason why it could not simulate realities with unhappiness, unfulfilled desires and undesired experiences. This would be valuable since our current real lives may also be ‘too easy’, and thus, limited, especially for the privileged.
In the same episode USS Callister, one digital clone was meanspirited in real life, but only in the terrible conditions of the simulated reality was he able to confront his character and realise his shortcomings, eventually apologising for his mistakes. Experiencing different lives and realities could thus reveal insights on a better way to live. It may allow us to develop greater empathy and solidarity with others by experiencing and understanding their lives, their suffering and their challenges. If the Greek gods envy humans for our mortality, perhaps experiencing a god-like perfect life in the EM would allow us to appreciate our vulnerability more. The EM may thus expand improve our conceptions of the good life, not restrict them.
In conclusion, if EMs are a distinct possibility in the future, we should at least use them in a way that teaches us to live better lives, possibly both inside the EM and outside in the real world. If so, in Weijers’ words, “Long Live the Experience Machine!”
Footnotes
[1] Although the concept of happiness is varied in philosophy, I will follow Nozick’s use of “happiness” in Philosophical Meditations, to mean positive, experiential psychological (“from the inside”) state(s). I avoid “pleasure” because the term has perceived narrowness. Most hedonists, however, adopt a pluralistic conception of pleasure — it can consist in diverse experiences as long as they are liked or desired (Railton 1989, p. 154). This means a hedonist can choose a life with pain and suffering, challenge and failure, but these are merely instrumental, such that overall and ultimately, his life will be desired and happy.
[2] Weijers also makes this point in Reality Doesn’t Really Matter (2011).
[3] One can still argue plugging in is a form of self-deception. However, since the EM-reality is not internally inconsistent (contradictory), even if it is distinct from the outside actual reality, it does not align with traditional philosophical conceptions of self-deception, so ethical implications may need to be considered differently.
*A previous version was submitted as a research paper for a module titled “Conditions of Happiness”.
