"Paradise-engineering is impossible. It would not be evolutionarily stable. Game-theoretic modelling demonstrates that selfishness is always the most profitable strategy possible for replicating units - whether genes or "memes" - susceptible to invasion by "defectors". Invincibly happy life-forms are inherently more vulnerable than their discontented, anxious and malaise-driven counterparts. A society of genetically pre-programmed ecstatics could not arise, let alone endure. It would be an environment open to invasion by mean-spirited defector mutants who would replace the hardwired sweethearts. Unpleasant states of consciousness will last forever."
This objection conflates two issues. Could it ever be an evolutionarily stable strategy for our descendants to be 1) innately happy? 2) innately unselfish?
The answer to the first question depends on the sort of happiness hardwired. Are we modelling a civilisation of, say, quasi-immortal superminds animated by gradients of genetically programmed well-being? Or wireheads and their genetic equivalents - a "blissed out" rather than cerebral hedonism? Clearly, the option of global wireheading [or lifelong immersive virtual realities etc] isn't an evolutionarily stable strategy, at least until the ageing process is conquered. This is because wireheads have no inclination to breed and certainly no desire to raise children. By contrast, fitness-enhancing gradients of well-being - and traditionally, ill-being - or their functional analogues can serve to motivate, protect and preserve us. Such gradients are adaptive when they are "encephalised" by evolution - and ultimately, shaped by rational design. Uniform euphoria [or chronic depression] and its insentient robotic analogues isn't adaptive. For this sort of functional architecture doesn't impel its subjects to do anything, learn anything - or nurture children. What counts from a computational-functional perspective is not our location on any absolute pain-pleasure scale, but that we are "informationally sensitive" to significant changes in our internal and external environment. Either way, genetic fitness isn't inseparably tied to a particular texture of experience, but to the way we behave and reproduce.
The controversial answer to the second question - namely that it is today's hardwired quasi-sociopathy that will prove evolutionarily unstable - sounds woolly-minded and naive, not to say biologically illiterate. Surely a civilisation founded on blissful altruists can't amount to a viable strategy? In any case, "hardwired sweetheart" scenarios aren't pivotal to the abolitionist project. They are also hugely more speculative. So why is blissful altruism an option for paradise-engineering worth exploring? Surely selfishness always wins?
Fortunately not. The (technical) genetic and metaphorical, behavioral and psychological senses of "selfish" are easy to confuse. This is because today they overlap so closely. Paradise-engineering can never be based on genetic unselfishness. But a genetic predisposition to altruism - in the metaphorical, behavioral and psychological senses of "altruistic" - can be evolutionarily stable against so-called defectors if and when it is also genetically selfish i.e. Darwinian fitness-enhancing. This is how our capacity for kindness, compassion and empathy - however meagre - arose in the first place. Even today, a genetic predisposition to individual "saintliness" isn't always a losing strategy; recall the self-sacrificing holy man who attracts devoted female admirers and becomes the proverbial father of his nation. But on the whole, a capacity to cheat, to compete and to lie has proved adaptive; humans evolved to be Machiavellian apes. Thus the proposal that unnatural selection pressure could ever cause saintliness to spread in a society of (non-clonal, genetically diverse) ecstatics looks implausible in practice. Surely alleles which promote competitiveness could never be outcompeted? Won't our descendants be, at best, happier egotists?
Now this may of course be the case. Yet decoding the human genome puts us on the brink of a major discontinuity in the mode of selection of self-replicating DNA - an evolutionary transition as profound as any in the history of life on earth. The long-term consequences of our capacity to rewrite our own code for the character of adaptive - and maladaptive - traits may be very different from what we imagine. In the Darwinian era of "natural" selection, a regime of blind, random genetic variation typically promotes an indifference to the fate of most of our fellow genetic vehicles. In the environment of evolutionary adaptation, this predisposition enhanced the inclusive fitness of our DNA. We have a "theory of mind", but our minimal capacity for empathy is limited mostly to kith and kin. So callousness has flourished. "Nice guys" get eaten or outbred. By contrast, the impending post-Darwinian era of "unnatural" selection portends genotypes that will be pre-selected/designed in anticipation of their desired effects. So genetic variation will no longer be random and undirected. Its consequences will be collectively planned - imperfectly at first, eventually perhaps via simulation and game-theoretic modelling with quantum supercomputers.
So questions of how we actually take the reproductive decisions, and on what criteria, are going to be crucial. What sort of traits do we want our offspring to have? Modelling post-Darwinian societies is immensely complex: post-humans may well rewrite their own individual genotypes ["genetic bootstrapping"] as well as the germ-line; and cloning will be trivially easy in the technical sense. Forms of "group selection" that simply weren't viable in the Darwinian Era become workable when reproductive decisions are collectivised; the "tragedy of the commons" can be forestalled. In a post-ageing world, reproduction may well be rare - and get progressively rarer as the carrying capacity of the earth [and ultimately the galaxy?] is reached. But in the era of designer babies, and taking a (very) crude genes' eye-view, a variant allele coding for, say, enhanced love-and-nurturance-inducing oxytocin expression, or a sub-type of serotonin receptor etc predisposing to unselfishness in the metaphorical, behavioral and psychological senses, may be differentially pre-selected and customised in preference to alleles promoting, say, sexual jealousy, aggressiveness or sociopathic behaviour. Genetically influenced "altruistic" traits that carry a higher payoff in the technical selfish genetic sense aren't susceptible to invasion by mean-spirited "defector" mutants - even if genetic variation were to remain random rather than directed. Thus in generations to come, the genetic and non-genetic senses of the word "selfish" may diverge. Indeed as the abolition of suffering becomes first technically feasible, and later trivially easy, then the language and institutions of traditional morality may decline into relics of a vanished age. What sort of values will replace them is hard to say. But as our descendants rewrite the vertebrate genome, and redesign the global ecosystem via nanotechnology, harsh "unnatural" selection pressure may penalise the very sorts of nasty traits that were genetically adaptive in the Darwinian Era. On this analysis, post-Darwinian superminds will be extraordinarily benevolent; but paradoxically, the science of paradise-engineering will have its origins in genetic selfishness.
Perhaps. But let's take a more pessimistic scenario. Assume that (post-)humans continue to be selfish in every sense. After all, just because allegedly we all (obliquely) seek happiness, this doesn't mean we seek happiness for everybody. Just because successful and intelligent life-forms will be able to underwrite their own happiness, why assume that they'll care about others? Let's further assume, contrary to the optimistic functionalist arguments above, that the textures of invincible happiness do inevitably make any coalition of alleles that promotes them potentially genetically vulnerable. After all, invincible well-being wasn't a viable strategy on the African savannah; why should it triumph in an era of artificial selection?
Does this pessimistic set of assumptions predict the persistence of a legacy architecture of misery and malaise? Will unpleasant states of consciousness really last for ever?
No, not necessarily, not even then. The more vulnerable enhanced well-being allegedly makes us, the more our self-interest will lie in collectively ensuring that all others are happy and well-disposed too; and in ensuring that any novel life-forms we create in the new reproductive era are constitutionally happy and benevolent. If the discontent of others potentially threatens our own well-being, then genetically underwriting their empathetic bliss serves our self-interest. If mutant psychopaths pose a potential danger [though in fact strict sociopathy tends to diminish inclusive fitness even in the primordial Darwinian era], then self-interest dictates using prophylactic germ-line therapy against genes promoting sociopathy and its sub-syndromal variants; this is one state-space of genetic options whose full exploration we can live without. In the past, natural selection ensured that selfishness in every sense of the word frequently paid. This entailed "winners" causing often severe suffering to losers. According to rank theory, the far greater incidence of the internalised correlate of the "losing" [behavioral] sub-routine, depression, compared to the "winning" sub-routine, euphoric (hypo)mania, attests to the terrible price that social animals have paid for the advantages of group living. Until now, blind genetic competition has ensured overt individual competitiveness among reproductive vehicles. There has been a sometimes physically violent struggle for the best mates and scarce resources. Winners and losers alike have been trapped on the same hedonic/dolorous treadmill. But when unlimited emotional well-being is possible for everyone at no cost to the well-being of others - and an unlimited diversity of good experiences is accessible to all via immersive VR - then only sustained malevolence, not mere egoism, will suffice to perpetuate the cruelties of the old order.
None of this proves that our descendants will really be smarter, nicer and happier - the magic trinity predicted and endorsed here. This is scenario-spinning, not true game-theoretic modelling. There are huge gaps and suppressed premises in all the above arguments for paradise-engineering. Which strategies will really prove to be stable remains to be seen. The nature of the ultimate winning strategy is open. Certainly a transformation of human nature isn't going to arise through a great spiritual awakening, an innovative package of socio-economic reforms, or a spontaneous desire to be nice to each other. But it's quite possible that, in the long run, the Darwinian genetic program based on suffering and quasi-sociopathy will lose out. Misery is not a stable strategy because by its nature rational agents seek to escape it; and soon a society of intelligent agents will have the collective capacity to do so.
4.28 ...contradiction...abolitionism can't be reconciled with an absence of compulsion... 4.29 ...why invoke nanotechnology ...isn't genetic engineering enough...? 4.30 ..."pushy" parents will choose genotypes for children who are smart, driven and successful rather than happy... 4.31 ...persistence of "natural" reproducers with Darwinian genotypes means that suffering won't be abolished... 4.32 ...cosmic HI? Some pitfalls... 4.33 ...why stress gradients of well-being? Wouldn't permanent maximum bliss be ethically better...? 4.34 ...why the headlong rush? Let's wait until we have the wisdom to understand the implications of what we're doing... 4.35 ...the Simulation Argument suggests suffering can never be abolished...
E-mail Dave : firstname.lastname@example.org