The Reproductive
Revolution

I The Evolutionary Transition

May all that have life be delivered from suffering.

— Gautama Buddha (trad. c.566 BC – c.480 BC)

For the past four billion years, evolution has been "blind." Natural selection operates through differential reproductive success, indifferent to the quality of experience it produces. The result is a living world pervaded by predation, disease, parasitism, and a vast ambient suffering that most moral philosophies have been impotent to address. Darwinian life is, in Hobbes' phrase, characteristically nasty, brutish, and short — a description that applies with roughly equal force to a nematode worm and a chimpanzee.

Yet something unprecedented has happened. One lineage of mammals has developed the cognitive and technological capacity to understand the molecular underpinnings of its own suffering — and to contemplate, for the first time in evolutionary history, rewriting the genetic instructions that generate it. This is not a metaphor. The chemistry of mood, the architecture of the hedonic set-point, the molecular pathways of physical pain — all are increasingly legible to molecular biology. The open question is no longer whether we can intervene; it is how, and with what wisdom.

The reproductive revolution is one axis of this transition. Each generation has always been shaped by the alleles that chanced to survive. What is now emerging is parental choice over which alleles persist — not through the blunt instrument of twentieth-century eugenics, which was scientifically ignorant, coercive, and morally catastrophic, but through the probabilistic, bottom-up, voluntary selection of embryos by prospective parents using tools of extraordinary genetic resolution.

Selection pressure in the post-Darwinian world will not disappear. It will change its character. The blind roulette of natural selection is giving way — slowly, unevenly, controversially — to a form of anticipatory selection in which heritable alleles and allelic combinations are chosen in light of their expected psychological and physiological effects.

II What Is New in 2026

When the first version of this essay appeared in 2008, preimplantation genetic diagnosis (PGD) existed for monogenic disorders — cystic fibrosis, Huntington's disease, BRCA1/2 mutations — and the broader field was nascent. The genome-wide association study (GWAS) era had barely begun. Polygenic scores were theoretical. The idea of generating human gametes from skin cells was confined to mouse experiments. The notion of selecting embryos for heritable wellbeing potential would have seemed extravagant speculation.

Eighteen years later, the landscape is transformed:

None of these developments, taken individually, constitutes the revolution. Together, they represent its early infrastructure: the capillaries through which a far larger transformation will eventually flow.

III Polygenic Embryo Selection

The case for universal preimplantation genetic diagnosis, advanced in the 2008 edition of this essay, rested on monogenic examples: the SCN9A pain-sensitivity gene, the COMT Val158Met polymorphism and its effect on dopamine metabolism and reward experience. The argument was straightforward: if parents already screen for catastrophic single-gene disorders, the ethical difference between screening for a guaranteed Huntington's allele and screening for alleles that probabilistically raise suffering without adequate compensating flourishing is one of degree rather than kind.

In 2026, that argument has moved from the theoretical to the commercially contested. PGT-P — preimplantation genetic testing for polygenic disorders — has become a practical reality. The technology calculates polygenic risk scores (PRS) across sibling embryos in an IVF cohort, using thousands of genetic variants identified through genome-wide association studies, and ranks them by estimated disease susceptibility.

The Case Studies: 2026 Edition

The 2008 edition highlighted two gene examples. The following table reflects the expanded understanding as of 2026, while preserving the original cases:

The expansion from two or three key loci to polygenic scores integrating thousands of variants represents not merely a quantitative change but a qualitative one. Rather than a handful of Mendelian switches, we are beginning to model the distributed genetic architecture of the hedonic set-point itself — the constitutional baseline around which mood fluctuates across a lifetime.

What PGT-P Can and Cannot Do

Honesty requires clarity about current limitations. Polygenic risk scores for complex traits explain, even in the best-powered studies, a fraction of heritable variance. The predictive signal for disease risk is stronger and more clinically validated than for psychological traits. The claimed ability of some commercial providers to rank embryos for "intelligence" substantially outstrips what the science currently supports: the polygenic score for educational attainment, the best available proxy, captures perhaps 12–15% of phenotypic variance in large, genetically homogeneous samples, and its cross-ancestry validity is contested.

Moreover, embryo selection operates within hard biological limits: a typical IVF cycle produces four to eight viable embryos, constraining the effective selection differential. Selecting the "best" embryo from a pool of six is not equivalent to engineering an optimal genotype from scratch. The gains are real but modest for any given family, though potentially substantial at the population level if universally applied over generations.

What PGT-P can reliably do, even now, is reduce the probability that a given child will inherit the most adverse tail of polygenic risk for serious diseases — schizophrenia, bipolar disorder, severe depression, coronary artery disease, type 2 diabetes. The reduction of suffering at the extreme left tail of wellbeing distributions is precisely where the moral case is strongest and the evidentiary standards most tractable.

IV In Vitro Gametogenesis

Preimplantation selection is constrained by the gametes available. A couple producing IVF embryos can only select among the allelic recombinations their existing genomes generate. In vitro gametogenesis (IVG) — the creation of functional gametes from pluripotent stem cells derived from any somatic cell — promises to dissolve this constraint.

The principle was established in mice: in 2016, Hayashi and colleagues demonstrated that fully functional eggs and sperm could be reconstituted from induced pluripotent stem cells (iPSCs) in the mouse, producing viable and fertile offspring. In 2023, the same group achieved chromosomal sex conversion in male mouse iPSCs and derived functional oocytes from them — a proof of concept for same-sex biological reproduction. These experiments were the foundation.

In 2025, the first human landmark: researchers derived 82 developing oocytes from human skin cells and fertilised them by IVF. The blastocyst conversion rate was low — under 9% — and chromosomal abnormalities were common. But the conceptual barrier had been crossed. The authors estimated clinical application was at least a decade away; other commentators, noting the pace of prior transitions in reproductive medicine, cautioned against complacency.

The conjunction of IVG and polygenic selection would represent a genuine phase transition. At present, the two technologies are advancing on separate tracks; their eventual convergence is not in doubt.

V The Wellbeing Genome

The original vision of the reproductive revolution was never primarily about disease prevention — it was about the positive recalibration of the hedonic treadmill. The hedonic set-point, the constitutional baseline around which an individual's subjective wellbeing gravitates regardless of circumstance, is substantially heritable. Twin studies consistently return heritability estimates of 40–50% for subjective wellbeing. The question has always been: which alleles, and in what combination, elevate or depress this baseline?

The wellbeing GWAS literature of 2026 offers partial answers. Large-scale genome-wide association studies — leveraging UK Biobank, the Netherlands Twin Register, and multinational consortia — have identified hundreds of loci associated with the "wellbeing spectrum": subjective wellbeing, life satisfaction, positive affect, neuroticism, and depressive symptoms. The genetic architecture is highly polygenic, with thousands of variants of tiny effect. Polygenic scores derived from these studies predict meaningful variance in wellbeing trajectories, brain structural correlates of happiness and resilience, and vulnerability to psychopathology.

Recent research has linked wellbeing polygenic scores to structural brain features — hippocampal volume, prefrontal cortical thickness, the connectivity of large-scale affective networks — beginning to map the neuroanatomical substrate of constitutionally high hedonic tone. This is, in essence, the beginning of a genetic map of what Pearce's Hedonistic Imperative called "superhappiness" — not the frantic euphoria of mania, but the information-sensitive high-hedonic baseline that a biologically enriched organism might one day maintain.

Genetically enhanced humans can recalibrate the hedonic treadmill, abolish ageing and disease, and phase out suffering throughout the living world.

— The Reproductive Revolution, 2008

This formulation now requires refinement. "Recalibrating the hedonic treadmill" is not a single intervention but an extraordinarily complex genomic project. The wellbeing genome is not a simple list of "happy alleles" to be selected or introduced; it is a deeply pleiotropic, context-dependent, epistatic network in which virtually every relevant allele has antagonistic trade-offs. The FAAH A allele that attenuates anxiety also reduces the ability to learn from aversive experience. The serotonin transporter long allele that confers resilience may blunt the motivational edge that some forms of creative suffering provide. The dopamine D4 receptor variants associated with novelty-seeking and reward sensitivity carry elevated risk of addictive vulnerability.

This is not a reason to abandon the project. It is a reason to pursue it with extraordinary intellectual care, empirical humility, and ongoing ethical vigilance. The alternative — leaving the genetic lottery of natural selection undisturbed — is not a neutral position. It is a decision to perpetuate, indefinitely, the suffering that random allelic inheritance generates.

VI Pitfalls

The 2008 essay identified a set of pitfalls. Eighteen years of accelerated development have confirmed some, modified others, and added new ones. An honest accounting follows:

• The Polygenic Complexity Trap. The assumption that a handful of key loci determine wellbeing was always naive. Complex traits are genuinely distributed across thousands of variants with tiny, interacting effects. Polygenic scores capture population-level trends, not individual destinies. Any reproductive application must operate within this statistical uncertainty without illusion of precision.
• Off-Target Pleiotropies. Virtually every allele that influences mood, cognition, or pain sensitivity does many other things simultaneously. Selecting for reduced pain sensitivity via SCN9A variants risks unintended effects on cardiac electrophysiology, olfaction, and peripheral immune signalling. The human genome is not a modular catalogue; it is a deeply integrated system forged across millions of years of coevolution.
• The Population Genetics Constraint. Polygenic scores trained on European-ancestry biobanks show substantially reduced predictive accuracy in non-European populations — sometimes dramatically so. A reproductive revolution built on biased genomic data would compound existing global health inequalities rather than dissolve them. Inclusive, diverse, worldwide genetic research infrastructure is a moral prerequisite.
• Commercial Capture and Misleading Claims. Companies offering embryo scoring for intelligence have made claims substantially exceeding what the science supports. Herasight and Nucleus Genomics entering the market in 2025 with bold assertions about a wide range of traits, including IQ, drew sharp criticism from geneticists. The gap between what can be responsibly offered and what can be profitably marketed is a structural problem requiring robust regulatory architecture.
• The Access Chasm. IVF remains expensive, physically demanding, and geographically uneven. PGT-P adds further cost. If these technologies consolidate advantages among the wealthy without broader access mechanisms, the reproductive revolution risks becoming a new vector of stratification rather than a universally liberating transition. The ethics of access is not secondary to the ethics of the technology itself; it is inseparable from it.
• Consent of Future Persons. Future generations cannot consent to the heritable modifications made on their behalf. This objection applies with full force to natural reproduction as well — no child consents to the alleles assembled by chance — but the deliberateness of selection places a greater burden of justification on parents and society. Reducing the probability of extreme suffering seems an uncontroversial application; selecting for contested psychological traits requires a more demanding standard of justification.
• The Value Pluralism Problem. Whose conception of a good life should guide selection? A negative utilitarian will weight freedom from suffering most heavily; a eudaimonist will weight flourishing and accomplished character; a religious conservative may regard any engineering of the human genome as hubris. Democratic societies will not converge on a single framework, and a pluralist regime will inevitably permit more than any single framework would endorse. The governance question is therefore genuinely hard.
• Germline Editing Safety. CRISPR-Cas9 off-target mutations, mosaicism in edited embryos, and the heritable transmission of errors to future generations remain unresolved safety concerns that justify the current near-universal prohibition on clinical germline editing. The He Jiankui case of 2018 — in which CCR5-edited babies were produced without adequate ethical oversight — remains a cautionary landmark. No responsible trajectory moves to clinical germline editing until safety standards vastly exceed current capabilities.

These pitfalls are serious. They do not constitute a case against the reproductive revolution; they constitute a specification of the conditions under which it can be pursued responsibly. Every powerful technology generates analogous lists — the pitfalls of anaesthesia, antibiotics, or nuclear energy did not argue against their development but against their reckless deployment.

VII Ethical Framework

The ethical case for the reproductive revolution rests, at its foundation, on a principle that most moral traditions share despite their differences: suffering is bad. Not instrumentally bad — bad in itself, as a matter of its intrinsic phenomenological character. A world containing less suffering is, all else equal, a better world than one containing more. This seems to most reflective people not a partisan axiom but an inescapable datum of moral experience.

Applied to reproductive genetics, the argument takes the following form:

If prospective parents can substantially reduce the probability that a future child will experience extreme and sustained suffering — by selecting among embryos that exist or would be created in any case — without imposing serious harms on others, then there is a strong presumptive case for making that option available, and arguably a moral reason to take it.

The presumptive case is strongest at the extremes. An embryo with a 70% polygenic risk score for severe recurrent depression, a debilitating anxiety disorder, or the full allelic burden of schizophrenia susceptibility represents — if these statistical estimates are at all reliable — a genuine prospect of prolonged suffering. The case for embryo selection in these situations is broadly comparable to the widely accepted case for PGD in monogenic disorders. The distinction is probabilistic rather than categorical.

The case is considerably more fraught for trait selection outside the domain of suffering — cognitive enhancement, aesthetic preference, physical stature — where the interests of the future child, the parents, and society diverge more sharply, and where the risk of compounding social inequality is greatest. A responsible ethics of reproductive biotechnology distinguishes these domains carefully rather than treating all heritable selection as equivalent.

What the reproductive revolution is not — and what its most scrupulous advocates have always insisted — is a return to the coercive state eugenics of the twentieth century. That program was scientifically ignorant, selecting on crude and mostly heritable phenotypic proxies; it was politically coercive, overriding individual reproductive autonomy; and it targeted populations already subject to systematic oppression. Voluntary, evidence-based, individually initiated embryo selection, operating within a robust regulatory framework and with genuine universal access, is structurally different in every relevant respect. The comparison to historical eugenics, while rhetorically powerful, is analytically superficial.

Nevertheless, the history warrants permanent vigilance. The mechanisms by which voluntary programs become socially coercive — through insurance incentives, employer pressures, social stigma, and gradually shifting norms about what constitutes responsible parenthood — are well-documented in other domains of reproductive medicine. Governance must be designed with these mechanisms in mind.

VIII Conclusion

In 2008, this essay argued that a major evolutionary transition was in prospect: that the blind genetic roulette of natural selection was giving way to a new kind of selection pressure in which prospective parents would pre-select alleles in anticipation of their psychological and behavioural effects. In 2026, this transition is underway.

Polygenic embryo selection is commercially available. In vitro gametogenesis has produced its first human oocytes from somatic cells. The genetic architecture of subjective wellbeing is being mapped, locus by locus, through genome-wide association studies of unprecedented scale. Artificial intelligence is beginning to model the interaction landscapes that human analysts cannot. The infrastructure of the post-Darwinian transition is being constructed, not always thoughtfully, not always equitably, but unmistakably.

The central question is no longer whether the reproductive revolution will happen. It is happening. The question is whether it will be guided by wisdom or left to the intersection of commercial incentive and regulatory inadvertence. Whether it will narrow the global distribution of suffering or exacerbate inequality. Whether the prospect of engineering heritable wellbeing — not the manic excess of unchecked reward circuitry, but the information-sensitive high hedonic tone that a richly flourishing life requires — will be taken seriously as an ethical imperative rather than dismissed as dystopian speculation.

Darwinian life has been characterised by suffering because suffering has been adaptive — a motivational signal that steered behaviour toward survival and reproduction in environments of scarcity and threat. The environments of the twenty-first century are radically different. The molecular signals that made sense in the Pleistocene generate anxiety, depression, and pain that are often maladaptive in the world we have built. We are not obligated to perpetuate them simply because they are ancient.

The moral vision at the heart of the reproductive revolution is not utopian in the naive sense. It does not promise effortless perfection. It proposes, with appropriate scientific humility and ethical seriousness, that the children of the future need not inherit, without examination or consent, a hedonic baseline that the accident of prior selection has provided. That a future is possible in which extreme suffering is as genetically archaic as smallpox. That the question of what kind of life a new person will have — not its exact contours, which will always emerge from the irreducible particularity of experience, but its constitutional hedonic substrate — is at last a question that love, reason, and science can together begin to address.