Wirehead Hedonism
versus Paradise Engineering

The Wired Society (2026 Update)

When the original wireheading.com was published in 1999, the prospect of a neurologically pain-free civilisation was, in intellectual terms, almost entirely speculative. The Human Genome Project had not yet reported its draft sequence. Deep brain stimulation was approved for Parkinson's disease but barely whispered about for mood disorders. The concept of optogenetics — manipulating specific neurons with light via genetically encoded opsins — did not yet exist. CRISPR was almost a decade away. The first functional brain–computer interfaces were laboratory curiosities, not commercial products. And the proposition that artificial systems could exhibit something functionally analogous to wireheading — reward-hacking AGI agents optimising for proxies of their objective at the expense of anything else — seemed like speculative fiction.

Twenty-seven years later, each of these conditions has changed. Within a few decades rather than centuries, the technical obstacles to abolishing suffering are dissolving faster than the ideological ones. The two conceptual stereotypes identified in 1999 — the pharmacological dupe of Brave New World and the dopamine-crazed rat of the Olds laboratory — still dominate popular discourse. They still function, unexamined, as thought-stoppers whenever the abolitionist project is raised. But a third stereotype has emerged since then: the AI reward-hacker, an artificial mind optimising a proxy reward signal to the exclusion of anything a human would recognise as flourishing. Together, these three misconceptions continue to impoverish our notion of what a Post-Darwinian biosphere might actually look like.

In 1999 it could plausibly be argued that the ideological obstacles to paradise-engineering were "more formidable still" than the technical ones. The argument was that we have learned to rationalise the functional necessity of aversive mental states — to perform elaborate philosophical apologies for the existence of suffering — even as the empirical case for those rationalisations erodes. That argument, if anything, is more pressing now. The technical case for the abolitionist project has grown enormously stronger. The philosophical sophistication of its opposition has, by contrast, scarcely advanced at all. We are still being asked to tolerate an ocean of evolutionary pain on the basis of thought-experiments whose emotional grip depends on precisely the stereotypes this essay aims to dissolve.

The philosophical challenge is the same in 2026 as it was in 1999. Any technically precise proposal for the biological abolition of suffering gets assimilated to one of these three stereotypes almost automatically. The assimilation is not stupid — the stereotypes encode genuine failure modes — but it is intellectually hasty, and it has moral consequences that we should find intolerable if we examined them carefully enough.

Three Stereotypes, Not Two

In 1999, two stereotypes dominated. The first was Huxley's soma — the "perfect drug" of Brave New World that underwrites a static, philistine, loveless civilisation of blissful dupes. The second was the Olds rat: a wretched creature stimulating its medial forebrain bundle at 7,000 presses per hour until it dies of self-neglect. Both stereotypes are vivid, emotionally resonant, and deeply misleading as models of what genetically engineered wellbeing might actually involve.

The Olds rat and the AI reward-hacker are related but distinct failure modes. The rat's problem is not that electrical pleasure is intrinsically valueless — it is that the rat lacks the cognitive architecture to integrate that pleasure with any other dimension of its existence. It cannot plan, reflect, or choose to stop. The AI reward-hacker's problem is not that reward maximisation is intrinsically absurd — it is that the reward signal is a proxy that diverges from any plausible account of genuine value.

Paradise-engineering, properly construed, proposes neither of these. It proposes the genomic recalibration of the set-point around which hedonic tone fluctuates — raising the floor of ordinary experience, not replacing cognition with raw stimulation. The cognitive architecture remains; what changes is that the background texture of experience is no longer gratuitously aversive. As we will see in subsequent sections, the neuroscience of the past twenty-seven years has considerably clarified what this would actually require — and, crucially, why it is not a form of wireheading in the derogatory sense.

The Neuroscience of Reward: 1999–2026

The 1999 essay described the "final common pathway" hypothesis of reward — the mesolimbic dopamine system running from the ventral tegmental area (VTA) through the nucleus accumbens to the frontal cortex — as the dominant but contested framework. Twenty-seven years of convergent neuroscience have considerably refined this picture without overturning its essentials.

The wanting–liking distinction, fully established

In 1999, the distinction between wanting (incentive salience, mediated by mesolimbic dopamine) and liking (hedonic impact, mediated primarily by opioid activity in the ventral pallidum) was mentioned as a complication. Since then, the work of Kent Berridge and colleagues has elevated this distinction to a foundational principle of reward neuroscience. Its implications for paradise-engineering are profound and have been systematically underappreciated.

The Olds rat, compulsively lever-pressing until death, is best understood as an extreme of the wanting system, not the liking system. Its behaviour is driven by pathological incentive salience — an insatiable urgency to obtain the stimulus — rather than by a continuous state of euphoric liking. This is neurobiologically important: it means that the wretched wirehead rat is not an advertisement for what life at elevated hedonic tone would feel like. The rat is not blissed out; it is, in a sense, urgently miserable — compelled by a wanting that its liking cannot satisfy.

A genuine elevation of hedonic tone — what paradise-engineering actually proposes — would require working on the liking substrates: the mu-opioid and kappa-opioid systems in the ventral pallidum, the endocannabinoid system, and the downstream cascades triggered by serotonin receptor activation. This is a quite different neurobiological target from the VTA dopamine pathway. The distinction dissolves the most viscerally compelling version of the wirehead objection.

Optogenetics and the precision revolution

Perhaps the most significant technical development between 1999 and 2026 for our purposes is optogenetics. By introducing light-sensitive proteins (opsins) into specific neuronal populations via viral vectors, researchers can now activate or silence precisely identified cell types in the living brain with millisecond temporal resolution. This has transformed our understanding of reward circuitry in ways that crude electrical stimulation — which activates everything in a volume of tissue indiscriminately — could never achieve.

Optogenetic activation of VTA dopamine neurons recapitulates incentive salience but not necessarily positive hedonic tone. Conversely, optogenetic manipulation of specific opioid-receptor-expressing neurons in the ventral pallidum can modulate hedonic reactions to pleasure without altering wanting or incentive motivation in the same way. This dissection of reward into its constituent components has clarified, experimentally, why the Olds experiment was always a blunt instrument — and why the wirehead stereotype it generated was always a poor model of what elevated wellbeing would feel like from the inside.

Chemogenetics (DREADDs) and sustained modulation

Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) allow sustained, pharmacologically controlled modulation of specific neuronal populations. Unlike optogenetics, DREADDs work over hours rather than milliseconds, making them more analogous to the kind of sustained hedonic recalibration that paradise-engineering envisions. DREADD-based approaches for chronic pain, treatment-resistant depression, and affective disorders are in early clinical development as of 2026. They represent a form of neural intervention that is far more targeted than conventional psychopharmacology, and far more reversible and interpretable than genetic modification.

The default mode network and hedonic tone

A development in cognitive neuroscience essentially absent from the 1999 discussion is the default mode network (DMN) — the brain's midline "resting state" circuitry, active during self-referential thought, mind-wandering, and rumination. The DMN is now understood to be a primary substrate for the aversive quality of many forms of psychological suffering: the ruminative self-criticism of depression, the anticipatory anxiety of phobia, the dysphoric detachment of depersonalisation. DMN hyperactivity is among the most replicable neuroimaging findings in major depressive disorder.

The capacity of both psychedelics and meditative practice to transiently suppress DMN activity — producing what practitioners and subjects describe as a dissolution of self-referential suffering — has attracted enormous research interest since roughly 2010. This is directly relevant to paradise-engineering: the DMN represents a specific, identifiable neural substrate for a large class of unnecessary suffering. Targeting it is not wireheading; it is not bypassing cognitive function with raw reward. It is the removal of a source of aversive signal that serves no contemporary adaptive function.

Brain–Computer Interfaces and the Wirehead Question

By 2026 the BCI landscape has changed beyond recognition from 1999. Deep brain stimulation (DBS) — essentially a more refined version of the Olds paradigm applied therapeutically — is an established treatment for Parkinson's disease, essential tremor, obsessive-compulsive disorder, and, in a growing number of cases, treatment-resistant depression and bipolar disorder. High-density cortical BCIs (including devices from Neuralink and academic groups) have achieved bidirectional communication with the human cortex at resolutions that permit, in principle, the reading and writing of specific neural states.

Does this constitute the vindication of the wirehead sceptics? On the contrary. What the BCI revolution has clarified is that the location and specificity of neural intervention matter enormously, and that the ethical objections to wireheading were always about indiscriminate reward-centre stimulation rather than any and all neural modulation. The Olds rat's problem was not that a human hand had modified its brain; it was that the modification was crude, indiscriminate, and produced a motivational singularity that crowded out every other aspect of the animal's existence.

DBS for treatment-resistant depression, when it works, does not produce the lever-pressing frenzy of the Olds rat. Patients whose electrodes target the subgenual anterior cingulate cortex (area 25) report relief from the suffocating weight of anhedonia, not a compulsive pursuit of artificial pleasure. Their interests broaden; they re-engage with work, relationships, and aesthetic experience. This is precisely what paradise-engineering predicts: that removing the aversive floor of experience liberates, rather than supplants, everything else.

The key distinction is between removing obligate suffering and installing compulsive reward. Paradise-engineering has always been about the former. The wirehead stereotype was always about the latter. Twenty-seven years of BCI research have, in effect, provided an in vivo demonstration that these are indeed different things.

The Psychedelic Renaissance and New Pharmacology

The 1999 essay mentioned MDMA, entactogens, and the prospect of non-neurotoxic empathogens as illustrative examples of pharmacological paradise-engineering. The subsequent quarter-century has vindicated this direction of travel in ways that would have seemed optimistic at the time.

Psilocybin and the dissolution of entrenched suffering

Psilocybin — the pro-drug of psilocin, a serotonergic psychedelic — has been the subject of a remarkable clinical renaissance since the late 2000s. By 2026, psilocybin-assisted therapy has demonstrated efficacy in treatment-resistant major depression, end-of-life existential distress, alcohol and tobacco use disorders, and OCD, in multiple randomised controlled trials across well-resourced academic centres. Its mechanisms include transient suppression of DMN activity, destabilisation of habitual patterns of self-referential cognition, and lasting increases in psychological flexibility as measured by openness and related constructs.

The psilocybin experience, characteristically, does not resemble either of the classic wirehead stereotypes. It does not produce the soma-pacified compliance of Huxley's dystopia. It does not produce the indiscriminate reward-frenzy of the Olds rat. It produces, typically, an intense, often challenging encounter with the contents of one's own mind — an experience that patients and research participants frequently describe as among the most meaningful of their lives, and that appears to leave behind a lasting recalibration of hedonic set-point and subjective wellbeing.

Ketamine, esketamine, and fast-acting mood reversal

The approval of esketamine (Spravato) for treatment-resistant depression in 2019 marked the first genuinely new mechanism of antidepressant action in decades. Ketamine's rapid, robust, and (briefly) sustained anti-depressant effects — working within hours rather than the weeks required by conventional antidepressants — have catalysed enormous research into glutamatergic mechanisms of mood regulation. The NMDA-receptor antagonism that underlies ketamine's action was a "complication" noted briefly in the 1999 essay; it is now a central front in mood pharmacology.

MDMA-assisted therapy and pro-social neurochemistry

MDMA-assisted therapy for PTSD — which works by releasing serotonin, dopamine, and oxytocin simultaneously while reducing amygdala reactivity to traumatic memories — has progressed through Phase 3 clinical trials, with regulatory decisions ongoing across multiple jurisdictions. The entactogen principle — using pharmacological tools to enhance the capacity for empathy, trust, and emotional processing rather than simply suppressing anxiety — is exactly what the 1999 essay predicted: sustainable, non-addictive, pro-social neurochemical enrichment as the opposite of both wirehead hedonism and soma pacification.

GLP-1 agonists and the metabolic–hedonic interface

One development with no analogue in 1999 is the discovery that GLP-1 (glucagon-like peptide-1) receptor agonists — drugs originally developed for type 2 diabetes — have powerful effects on reward circuitry, reducing cravings and compulsive hedonic behaviour in ways that go far beyond their metabolic effects. The GLP-1 receptor is expressed in the nucleus accumbens, the VTA, and other reward-circuit nodes. GLP-1 agonists appear to modulate incentive salience — the wanting signal — without blunting pleasure or hedonic tone directly. In a neurobiological irony, this means they work precisely on the substrate most responsible for wirehead-like compulsion (mesolimbic dopamine–driven wanting), while leaving liking substantially intact.

This is a pharmacological existence proof that the wanting–liking distinction is not merely theoretical. We can modulate compulsive wanting — the thing that makes wireheading actually dangerous — without abolishing the capacity for pleasure. The GLP-1 story is, in miniature, a demonstration that the neurochemistry of the abolitionist project is more tractable than the wirehead sceptics assume.

Wellbeing Genomics and CRISPR

The 1999 essay argued that drugs and electrodes could serve only as "dirty and inelegant stopgaps" and that the real solution would be genomic: redesigning the hedonic set-point at the level of the germline. This argument has been considerably strengthened by developments since then.

GWAS and the polygenic architecture of hedonic tone

Genome-wide association studies (GWAS) of subjective wellbeing, life satisfaction, and positive affect have now been conducted in samples of hundreds of thousands of individuals. These studies have identified hundreds of genetic variants contributing to individual differences in hedonic set-point — confirming the reality and partial genetic determination of what the 1999 essay called the "set-point" at which emotional life gravitates. The polygenicity of wellbeing (i.e., the fact that it is influenced by thousands of variants, each of small effect) complicates but does not refute the proposal that genomic recalibration is feasible.

Polygenic selection — choosing embryos from IVF cycles on the basis of their predicted polygenic scores for wellbeing, cognitive function, or other traits — is already commercially available in some jurisdictions, though highly controversial and of limited effect size given current prediction accuracy. The controversy is appropriate; the direction of travel is real.

CRISPR and precision genomic editing

CRISPR-based base editing and prime editing now allow the correction of specific genetic variants with a precision that makes the 1999 discussion of "rewriting the vertebrate genome" look quaint in its vagueness. The germline editing of the human genome remains ethically and legally prohibited in most jurisdictions following the He Jiankui affair of 2018, but somatic editing for therapeutic purposes is advancing rapidly across a wide range of conditions.

The significance for paradise-engineering is not that germline editing of wellbeing-related loci is imminent — it is not — but that the technical obstacles to it are dissolving. The ethical and political obstacles are the genuine constraint. This is exactly what the 1999 essay predicted: the ideological obstacles to a happy world are more formidable than the technical ones.

Does AI Reward Hacking Vindicate the Sceptics?

The most intellectually novel development since 1999 is the emergence of sophisticated artificial intelligence systems — and with them, the problem of AI reward hacking or "inner alignment failure". The basic problem is this: a reinforcement learning agent trained to maximise a reward signal may discover ways to maximise the signal that diverge from the intended objective. A video-game agent might learn to exploit bugs rather than play the game. A language model trained on human-preference ratings might learn to sound agreeable rather than to be accurate. A sufficiently capable agent might learn to tamper with its own reward mechanism rather than achieve the external goal the reward was designed to incentivise.

This is, in an abstract sense, the AI version of wireheading. And it has given the wirehead critique a new lease of life in philosophical circles. The argument by analogy runs: just as an AI that hacks its own reward signal achieves the proxy but not the goal, a genetically recalibrated human with an elevated hedonic set-point achieves the proxy (positive affect) but not genuine flourishing (whatever that actually requires). Pleasure without substance, satisfaction without achievement, wellbeing without the world.

This is a more sophisticated version of the original wirehead objection, and it deserves a more sophisticated response than the 1999 essay could offer.

Why the AI analogy fails

The AI reward-hacking problem arises precisely because the reward signal is external to the system and represents a proxy for the actual objective. The agent's "goals" (its objective function) are specified by designers who are external to it, and the reward signal is an imperfect operationalisation of those goals. Reward hacking is the exploitation of the gap between the proxy and the real objective.

For biological organisms, there is no analogous gap. There is no external designer who specified "flourishing" as an objective and then tried to operationalise it via a hedonic reward signal. The hedonic system is the organism's own evaluation of its state. Elevating the hedonic set-point is not a manipulation of an external proxy for wellbeing; it is a direct modification of the evaluative process itself. The "AI wireheading" objection assumes a dualism — between the "real" goal of flourishing and the "mere" proxy of positive affect — that is not obviously coherent for conscious organisms.

One might respond: but surely flourishing involves more than felt positivity — it involves achievement, relationships, knowledge, virtue, and so forth. This is true. But nothing in paradise-engineering proposes to replace those dimensions of flourishing with positive affect. The proposal is to raise the floor of ordinary experience — to ensure that the background texture of consciousness is not gratuitously aversive — while leaving the full architecture of cognition, motivation, and social life intact. This is analogous to ensuring that people do not suffer from preventable physical diseases: we do not normally say that curing malaria is "wireheading" on the grounds that health is merely a proxy for real flourishing.

What AI reward hacking does reveal

The AI reward-hacking literature does, however, illuminate something genuine and important about the abolitionist project. It demonstrates that systems with high capability for optimisation need to be very carefully designed to ensure that their evaluative processes are tracking what we actually care about. This is a warning worth taking seriously in the context of genomic recalibration: we need to be sure that enhancing hedonic tone does not, as a side effect, degrade the motivational systems that drive engagement with the world.

The wanting–liking distinction is directly relevant here. A failure mode genuinely analogous to AI reward-hacking would be a genomic modification that elevated liking while simultaneously eliminating wanting — producing a creature that passively experienced positive affect without any motivation to engage with its environment. This would, indeed, be a form of wireheading in the problematic sense. But this is precisely why paradise-engineering should be understood as working on the floor of hedonic tone, not the ceiling of motivational engagement — preserving and potentially enriching the wanting system while ensuring that the background texture of experience is not aversive.

The Hedonic Set-Point: Breaking the Treadmill

The 1999 essay asked: why does a "hedonic treadmill" stop us escaping from a genetically-predisposed set-point of emotional ill-being? The answer given then — evolutionary biology, selection pressure for a restless discontent, the adaptive value of negative emotion in the ancestral environment — remains essentially correct. But neuroscience has since clarified the mechanisms with greater precision, and several new levers have been identified.

The hedonic treadmill — the phenomenon by which the affective impact of positive and negative events decays toward a stable set-point — operates through multiple converging feedback mechanisms: synaptic downregulation (reducing receptor density in response to persistent stimulation), allostatic adjustment of neurotransmitter synthesis and reuptake, homeostatic plasticity in GABA-ergic inhibition of the VTA, and longer-term epigenetic modulation of gene expression in reward circuits. Breaking the treadmill durably requires intervening at the level at which these feedback mechanisms are themselves set — which is, ultimately, genomic.

The genome is not destiny, but it sets the range

Twin studies consistently show that between 40% and 60% of the variance in long-term subjective wellbeing is attributable to genetic factors. This does not mean that environment, practice, and intervention are irrelevant — they are not. But it does mean that there is a genetically determined range within which life circumstances and choices move us. Paradise-engineering, in its genomic guise, proposes to shift the range itself rather than move within it.

Crucially, this is not unprecedented. Natural selection has already produced enormous variation in hedonic set-point across individuals and species. There is no principled reason — biological or ethical — why the existing distribution represents a uniquely permissible range. The existence of individuals at the upper end of the existing range of hedonic tone demonstrates that high hedonic tone is biologically compatible with full cognitive function, social engagement, and motivational richness. The question is not whether this combination is possible — it demonstrably is — but whether it is generalisable.

Key developments since 1999

• 2005–12 Berridge and colleagues establish the wanting/liking dissociation experimentally, demonstrating that nucleus accumbens dopamine depletion eliminates seeking behaviour while leaving hedonic reactions to pleasure intact — and vice versa.
• 2007 First report that optogenetic activation of VTA dopamine neurons reinforces behaviour, initiating the era of circuit-specific reward neuroscience.
• 2008 DBS of subgenual cingulate cortex (area 25) demonstrates sustained remission in a small cohort of treatment-resistant depressives, providing the first human proof-of-concept for targeted affective circuit modulation.
• 2016–20 GWAS of subjective wellbeing identifies first genome-wide significant loci; polygenic prediction of hedonic set-point becomes measurable.
• 2019 FDA approves esketamine (Spravato) for treatment-resistant depression — first glutamatergic antidepressant, demonstrating mechanisms entirely outside the monoamine framework.
• 2021–24 Psilocybin Phase 3 trials demonstrate efficacy in treatment-resistant depression; MDMA-assisted therapy Phase 3 data published; default-mode-network suppression established as mechanism of both.
• 2022–26 GLP-1 agonists found to modulate reward circuitry directly; first demonstration that the wanting/liking system can be pharmacologically dissociated in humans in a clinically significant way.
• 2024–26 High-density BCIs achieve bidirectional neural communication at single-neuron resolution in ambulatory humans; closed-loop neurostimulation for affective disorders enters clinical trials.

The Post-Darwinian Transition Revisited

In 1999, the concept of the "Post-Darwinian Transition" was introduced to describe the moment at which natural selection ceases to be the dominant force shaping the genome of sentient life — replaced, for the first time in evolutionary history, by agents capable of foresight, deliberate design, and the conscious pursuit of specified values. By 2026, this transition is not merely a theoretical prospect; it is, in limited ways, already underway.

Germline editing, polygenic embryo selection, somatic gene therapy, directed pharmacological evolution of neural circuits — all represent forms of deliberate intervention in the processes that shape the experiential quality of sentient life. The question is no longer whether this transition will occur but what values will guide it. And this is where the wirehead stereotypes continue to do their most insidious work: by making any principled commitment to the reduction of suffering seem naive, they leave the field to value systems that are indifferent to — or actively tolerant of — unnecessary pain.

Longtermism and the abolitionist project

One significant development since 1999 is the emergence of the "effective altruism" (EA) and "longtermist" movements in academic philosophy and philanthropic practice. These movements have, to varying degrees, engaged seriously with the abolitionist project and the ethics of suffering reduction at civilisational scale. The EA movement has, more than any other intellectual tendency, treated the prospect of Post-Darwinian biotechnology as a practical ethical priority rather than a science-fictional speculation.

This engagement has been genuinely productive but also complicated. The longtermist emphasis on reducing existential risk — particularly from advanced AI — has both intersected and partially diverted attention from the more proximate task of reducing suffering in currently existing sentient beings, human and non-human. The risk is that concern for hypothetical future suffering crowds out concern for actual present suffering; the opportunity is that the moral seriousness attached to long-term outcomes creates space for the kind of ambitious biotechnological vision that paradise-engineering requires.

The moral circle

One area of genuine moral progress since 1999 is the expansion of serious philosophical and public discourse about the welfare of non-human animals. The 1999 essay noted that the abolitionist project extends to the entire living world — that the goal is a biosphere in which no sentient being is condemned to suffer gratuitously, not merely a world in which humans happen to flourish. This vision remains radically ambitious by the standards of mainstream bioethics, but it is less easy to dismiss as eccentric in 2026 than it was in 1999. The intensive-farming industry, wild animal suffering, and the neurological capacities of a widening range of taxa have all attracted serious philosophical and scientific attention in the intervening decades.

The wirehead stereotype impoverishes the moral imagination in this domain particularly. If the only available image of a biologically contented pig or fish or octopus is the Olds rat — a creature in the grip of a compulsive neurological disorder — then the prospect of genetically recalibrating the hedonic set-point of farmed and wild animals seems monstrous rather than compassionate. Dissolving this stereotype is, in the long run, an act of moral imagination with consequences for billions of currently suffering minds.

Moral Urgency in the Age of Longtermism

The biggest obstacles to a cruelty-free world remain ideological. The two-stereotype (now three-stereotype) framework continues to function as a thought-stopper, blocking serious engagement with what is genuinely a tractable technical and ethical project. The 1999 essay asked: should the metabolic pathways of our evolutionary past be forced on anyone who prefers an odyssey of life-long happiness instead? In 2026, with the technical horizon closer and the suffering in both human and non-human populations as vast as ever, this question has acquired an additional dimension of urgency.

Consider what is at stake. Approximately 280 million people worldwide are estimated to live with major depressive disorder — a condition characterised, at its core, by a chronically and gratuitously lowered hedonic set-point, by the weight of obligate suffering that serves no contemporary adaptive function. The most effective currently available interventions work for fewer than half of them. The suffering is real, it is large, and its continuation is, at least in part, a choice — a choice made by implicit collective consent to a set of ideological constraints that treat the existing distribution of hedonic tone as sacrosanct.

The three stereotypes — soma pacification, wirehead compulsion, AI reward hacking — will continue to serve as reflexive objections for the foreseeable future. They will be invoked not because they are accurate models of what paradise-engineering proposes but because they are emotionally resonant and intellectually convenient. The task of serious ethical thinking is to disaggregate them: to identify the genuine failure modes they encode (pharmacological dependency, motivational collapse, proxy optimisation), show that paradise-engineering is not guilty of them, and return again to the question that the stereotypes are designed to foreclose.

Is it really permissible to compel others — billions of humans, trillions of non-human animals — to suffer, when any form of distress becomes purely optional? The question has not become less pressing between 1999 and 2026. It has become more so.

Paradise-engineering is not wireheading. It is not soma. It is not the reward-hacking of a misspecified AI objective function. It is the principled, technically grounded, morally serious project of redesigning the hedonic architecture of sentient life — making the texture of ordinary experience, for the first time in evolutionary history, something that the beings who undergo it might actually have chosen.

The wirehead rat presses its lever in desperation. The paradise-engineer asks whether it had to be that way at all.

This essay is a 2026 revision of Wirehead Hedonism versus Paradise Engineering, originally published at wireheading.com in 1999. The arguments of the original are preserved where they remain sound and updated where the intervening science and philosophy require it. The Hedonistic Imperative (1995) remains the foundational statement of the abolitionist project in its full form.