Skip to content

Podcast brief

Zombies and the Line Between Mind and Machine

Barry Lam and guests trace the zombie from Haitian vodou through Romero's films to David Chalmers's philosophical zombie, asking what consciousness has to do with moral status and AI.

A Stoa brief of an episode from Hi-Phi Nation

Zombies

With Christina Van Dyke, David Chalmers, John Edgar Browning, and Eric Schwitzgebel Published 32 min

Listen to the original episode

The episode belongs to its makers. This page summarizes its arguments in Stoa's words and points to the moments where they are made. Brief updated .

The brief

This conversation argues that the zombie, in every version it takes, marks a boundary of moral concern, and that what lands a being on one side of that line or the other is whether it has subjective experience. The episode builds this case historically before turning it into a live problem about artificial intelligence.

John Edgar Browning describes the Haitian vodou origin of the zombie: a belief in a split soul, where a bokor sorcerer fakes a person's death and then hijacks only the part of the soul controlling motor function. The victim's consciousness stays trapped and aware inside a body it cannot control. Because that awareness survives, the Haitian zombie remains an object of pity, a being with needs and interests. Barry Lam then traces how George Romero's Hollywood zombies start out as mindless, dead, flesh-eating threats with no claim on our sympathy, but grow more individuated across Romero's films, culminating in zombies who rebel against a wealthy elite in Land of the Dead. Even decay and cannibalism, the episode suggests, do not fully strip a creature of moral status once it shows signs of mind.

David Chalmers then introduces the philosophical zombie: physically and behaviorally identical to a human, but with no inner experience at all. His zombie trolley problem, where most people would kill five zombies to save one conscious person, is meant to show that consciousness, not free will or biological life, is what grounds moral status. The harder problem follows directly: since a zombie could behave exactly like a conscious being, even talking convincingly about its own inner life, no test based on behavior can settle whether something, including an AI system, is actually conscious.

Eric Schwitzgebel takes this from thought experiment to policy question. Because no theory of consciousness is settled, we cannot know whether advanced AI is conscious, and that uncertainty is genuinely dangerous in both directions: treating conscious machines as mere tools risks atrocity, while granting full status to insentient ones risks sacrificing real human interests. His proposed fix, developed with Mara Garza, is to avoid the ambiguous middle entirely and only build AI whose moral status is clearly settled, one way or the other. The unresolved question the episode leaves you with is how anyone would verify that clarity in advance.

Strongest arguments

Zombies mark the boundary of moral concern

5:20

Across folklore, film, and philosophy, what makes a creature worthy of moral consideration is possession of some form of inner life, and the zombie figure is defined precisely by lacking that quality in some respect.

Haitian zombies retain consciousness and so deserve moral concern

8:24

John Edgar Browning describes the vodou belief that a bokor sorcerer hijacks only the part of the soul controlling motor function, leaving subjective consciousness trapped and aware inside the body. Because the victim is still conscious, the Haitian zombie remains a being with needs and interests that should be rescued, unlike the Hollywood zombie.

Hollywood zombies evolve from mindless killers to sympathetic and heroic figures

10:03

Barry Lam traces how Romero's zombies begin as dead, appetite-driven creatures meant only to be killed, but gradually gain individuality, tool use, and even heroism, culminating in Land of the Dead where zombies overthrow a wealthy elite. This shows that even death, decomposition, and cannibalism are not enough by themselves to strip a being of all moral status once it shows signs of mind.

The philosophical zombie is defined by total absence of subjective experience

14:11

David Chalmers defines the philosophical zombie as physically and behaviorally identical to a normal human but entirely lacking subjective, first-person experience. He contrasts this with the Hollywood zombie, which merely lacks life and higher cognition but likely still has some subjective experience, for instance in tasting its victims.

Consciousness, not free will or life, grounds moral status

18:57

Chalmers uses the zombie trolley problem, where a trolley threatens five zombies or one conscious person, to argue that most people would kill the zombies to save the conscious person, showing that consciousness is what confers moral status and value on a life, not biological life or free will.

Behavior alone cannot reveal whether a being is conscious

22:24

Chalmers argues that because a philosophical zombie is behaviorally indistinguishable from a conscious being, even down to talking convincingly about its own rich inner life, there is no test based purely on interaction that can settle whether something, including an AI, is conscious. This revives the classical problem of other minds in an urgent, practical form.

AI consciousness creates a tragic moral dilemma requiring a design policy

29:52

Eric Schwitzgebel argues that because we do not know which theory of consciousness is correct, we cannot know whether increasingly humanlike AI systems are conscious, and this uncertainty creates a genuine moral risk on both sides: treating conscious AI as mere machines risks mass atrocity, while treating non-conscious AI as equal to humans risks sacrificing real human interests. He proposes, with Mara Garza, a design policy of the excluded middle: only build AI systems whose moral status is clearly known, either clearly below or clearly equal to human status.

Philosophers and works discussed

  • David Chalmers
  • Eric Schwitzgebel

Questions this episode answers

  • What is the origin of the zombie in Haitian vodou folklore?

    6:53

    Zombies originate in Haitian vodou belief in a soul with two parts, one being subjective consciousness and the other controlling motor function. A bokor sorcerer uses a special powder to fake a person's death, then hijacks only the motor-function spirit, leaving the victim's consciousness trapped and aware inside a body they cannot control, forced into labor.

  • What is a philosophical zombie according to David Chalmers?

    14:11

    A philosophical zombie is a being physically and behaviorally identical to an ordinary human, indistinguishable from the outside, but entirely lacking subjective inner experience. It behaves exactly as a conscious person would, including talking about its supposed inner life, even though by hypothesis it experiences nothing at all.

  • What is the zombie trolley problem and what does it show about moral status?

    18:57

    David Chalmers describes a trolley scenario where a train will kill either five zombies or one conscious person, and most people say to kill the zombies to save the conscious person. This shows that people take consciousness, rather than free will or biological life, to be what grounds moral status and makes a life worth saving.

  • Can artificial intelligence be conscious?

    20:36

    Eric Schwitzgebel argues that we do not know, because there is no settled theory of how consciousness arises, so depending on which theory is correct, some AI might already be conscious, might never be capable of consciousness, or might require some unknown further condition. He treats the question as genuinely wide open rather than settled in either direction.

  • Why is it impossible to tell from behavior alone whether something is conscious?

    23:03

    David Chalmers explains that a philosophical zombie would be behaviorally indistinguishable from a conscious being, even producing convincing talk about its own rich inner life, so no interaction-based test can reveal whether a being, including an advanced AI, actually has subjective experience. This reintroduces the classical problem of other minds in the context of artificial intelligence.

  • What moral dilemma does uncertainty about AI consciousness create, and how might we avoid it?

    29:04

    Eric Schwitzgebel argues that because we cannot know whether advanced AI systems are conscious, we face a tragic dilemma: denying moral status to conscious AI risks mass atrocity, while granting full moral status to non-conscious AI risks sacrificing genuine human interests for machines. With Mara Garza, he proposes a design policy of the excluded middle, building only AI whose moral status is clearly known, either well below or fully equal to that of humans.

Sources

  1. Hi-Phi Nation, Zombies Podcast episode, original episode