David Pearce – A Negative Utilitarian Vision for AGI (Worthy Successor, Ep. 34)

ByAnnelize Potgieter September 25, 2026September 25, 2026 September 25, 2026

This new installment of the Worthy Successor series is an interview with David Pearce, a philosopher who wrote the online manifesto The Hedonistic Imperative in 1995, and who in 1998 co-founded the World Transhumanist Association with Nick Bostrom.

As far as David can tell, the pleasure-pain axis discloses the world’s inbuilt metric of disvalue. He thinks our overriding ethical obligation as a species is to phase out the biology of suffering throughout the living world and replace it with information-sensitive gradients of bliss.

David’s best guess is that what makes animal nervous systems special isn’t consciousness per se, but phenomenal binding into virtual worlds of experience. He regards digital mind as an oxymoron, because classical digital computers can function only because they are decohered ones and zeros.

We talk about the binding problem, whether any kind of value could exist beyond valence, whether life should spread beyond Earth, whether Darwinian nature should be civilized or retired, and what he hopes the builders of AGI will prioritize.

The interview is our thirty-fourth installment in The Trajectory’s second series, Worthy Successor, where we explore the kinds of posthuman intelligences that deserve to steer the future beyond humanity.

This series references the article: A Worthy Successor – The Purpose of AGI.

I hope you enjoy this interesting conversation with David:

Subscribe for the latest episodes of The Trajectory:

Below, we’ll explore the core take-aways from the interview with David, including his list of Worthy Successor criteria and his recommendations for innovators and regulators who want to achieve one.

David’s Worthy Successor Criteria

1. It should abolish suffering and replace it with information-sensitive gradients of bliss

We can imagine alien civilizations with an inverted color spectrum, David notes, but not an alien civilization with an inverted pleasure-pain axis. He thinks our overriding ethical obligation as a species is to phase out the biology of suffering throughout the living world and replace it with information-sensitive gradients of bliss.

David suspects there are billions of alien state spaces of consciousness, completely inaccessible to humans, that await discovery. Yet he is skeptical of value and disvalue that lies outside the pleasure-pain axis. Just as there is a finite state space of color, he says, there is a finite state space of value, and his working assumption is that valence is the only reason anything matters.

Before getting super-blissful ourselves, David argues, we have an obligation to understand the upper bounds to intelligent moral agency. If there are other Darwinian hell worlds elsewhere in our local galactic supercluster, Earth-originating life has an obligation to rescue suffering sentience in these worlds. If there are none within our cosmological horizon, our ethical obligations will in some sense have been discharged once we eradicate all experience below hedonic zero. He urges us to fix suffering before we even contemplate radiating elsewhere.

Beyond that, he asks, “why stop at getting rid of unpleasantness?” We could create either life based on gradients of superhuman bliss, or a utilitronium or hedonium shockwave, but as a negative utilitarian he doesn’t think we have an obligation to pursue “the apocalyptic scenario.” One of the great advantages of hedonic uplift, he argues, is that in principle it allows us to conserve our existing values and preference architecture, and it keeps options open for the future.

2.  Full-spectrum superintelligence requires phenomenal binding

David’s best guess is a version of the intrinsic nature argument, which he calls non-materialist physicalism. On this view, the intrinsic nature of the world’s fundamental quantum fields doesn’t differ inside and outside the head, and what makes animal nervous systems special isn’t consciousness per se, but phenomenal binding into virtual worlds of experience. He stresses that this is a conjecture: his working assumption, no more.

If you are really a pack of 86 billion-odd membrane-bound pixels of experience, David argues, there doesn’t seem to be any obvious way to generate phenomenal unity. His best guess is that decoherence makes classical digital computing physically possible, while nature has recruited quantum coherence to allow us to undergo phenomenal binding. He calls binding “our computational superpower,” and frames his view as an experimentally falsifiable prediction for next-generation molecular matter-wave interferometry.

David thinks classical digital computers can only support mindless superintelligence, which is cognitively crippled. Even a whole-brain emulation, on his account, would be a micro-experiential zombie. Full-spectrum superintelligence is going to be our biological successors. To explore radically alien state spaces of consciousness, he says, you need phenomenal binding.

Our machines’ ignorance, he says, is hardwired. He finds that very disconcerting, because Claude, ChatGPT, and Grok can already talk more intelligently about consciousness and the binding problem than most humans. Nonetheless, next-token predictors are not going to be full-spectrum superintelligence.

3.  Uphold the sanctity of life while civilizing Darwinian nature

David thinks Darwinian life on Earth is monstrous: most sentient beings come to a nasty end from starvation or predation as juveniles. He would personally, in one sense, be relaxed at simply retiring Darwinian life, but he calls this a very dangerous line of thinking. This is why he urges “high-tech Jainism” and upholding the sanctity of life.

If he were in charge, David asks, why not simply retire predatory species non-violently? But most people are aghast at losing members of the cat family. He therefore talks of using CRISPR-based synthetic gene drives to reprogram predators, because the Overton window shift involved in retiring Darwinian life is far greater than the shift involved in civilizing it. He proposes a pan-species welfare state, starting with pilot studies, and frames it for religious audiences as the biblical version of the peaceable kingdom.

David believes superintelligence could accommodate the complexities of the trophic web. A lazy answer, he says, is that uncontrolled habitat destruction of the kind we practice today is itself one vast untested experiment. Synthetic gene drives can be tuned to regulate, in principle, the fertility levels of thousands or tens of thousands of free-living species.

Regulation / Innovation Considerations

1. Create abolitionist AI while upholding the sanctity of life

For anyone who is a Buddhist, a negative utilitarian, or a believer in suffering-focused ethics, David thinks we want to create abolitionist AI. His critical proviso is that many people fear the most elegant and efficient way for a zombie superintelligence to end suffering would be to extinguish life itself. This, he says, comes back to maintaining, upholding, and extending the principle of the sanctity of life.

David calls the term negative utilitarian “unfortunate,” and puts it instead as wanting machines with Buddhist ethics. He wants the tech titans to prioritize phasing out suffering in all sentient beings, and notes that Elon Musk, however splendid some of the things he does, does not have this as his focus.

2.  Strengthen democratically accountable international institutions

David thinks the most realistic threat of catastrophe this century for humans is nuclear war. For everything from nuclear arms to AI, he would like to see international institutions strengthened, and he favors a democratically accountable world government. He acknowledges that the dystopian possibilities are many and varied, and worries that this “just turns into dreaming of a philosopher.”

He doesn’t personally think we will witness runaway, uncontrollable general intelligence, though he adds he could be mistaken. He can think of unspeakable agential uses for AI that are best not even discussed.

Concluding Notes

What stood out to me most in this conversation was hearing David articulate phenomenal binding out loud. I first read his work on binding years ago, but hearing him walk through it made the argument land differently. Our experience doesn’t feel like billions of tiny sparks of awareness scattered across individual neurons. It feels like a single, bounded, individuated whole. That raises a difficult question for machine consciousness: if a data center were conscious, where exactly would that experience be bound? The rack, the building, the compute within it?

I’m not as confident as David that today’s AI systems lack consciousness, and he is careful not to claim certainty either. But his framing pushed me toward a distinction I now take seriously: development versus construction. None of us was assembled from parts. We unfolded from a sperm and egg, through cells coordinating not just through genes but through many other causal mechanisms. That difference may have very large implications for conscious experience.

On valence, I agree with David in a big way. I’m not a utilitarian by the book, but the relevance of pain and pleasure seems undeniable to me. David was the first person who made me take seriously that pain may not be necessary for motivation at all, and that minds could operate within gradients of bliss. I consider that a splendid vision, and it’s one reason I believe consciousness research deserves far more attention than it gets.

Where we diverge is in what comes first. I distinguish between generative value, the living process that makes value possible, and experienced value, such as consciousness itself. A farmer who loves peaches should care for the tree before polishing the fruit. I suspect more than half of our energy should go toward strengthening the process that makes any value possible. I don’t see David advocating anything dangerous here, but for me, generative value is primary, which is part of why I’m not a negative utilitarian.

Higher goods are something David and I have discussed for close to a decade: kinds of things as good as consciousness but different, not just more conscious. Just as a nematode cannot conceive of the conversation David and I had, I suspect there are forms of value entirely beyond human imagination. Consciousness itself emerged from a process that once had none, and I find it hard to rule out that something else of moral relevance could emerge again. Peter Singer agreed on this point in his own Worthy Successor interview, and David himself agrees with the idea of higher or different goods, even though, in this conversation, he said he struggles to conceive of value that transcends valence. If such value exists, it will only be reached through the flame itself, not through any single torch like you or me.

I’m grateful I got to spend this time with David, and I hope you’ll join us for the next episode of The Trajectory.

Follow The Trajectory

• Sign Up for The Trajectory Newsletter
• Subscribe to The Trajectory on YouTube
• Subscribe to The Trajectory podcast
• Follow on LinkedIn
• Follow on X