Apocalyptic AI

Apocalyptic AI is Robert M. Geraci's term for a strand of technological futurism in which artificial intelligence promises to overcome the limits of human life and bring about a transformed world. Its characteristic hopes include minds transferred into machines, freedom from biological death, and intelligence expanding beyond the Earth. Geraci interprets these expectations through the study of Jewish and Christian apocalyptic traditions: present suffering and limitation give way to a new order inhabited by transformed beings.^[1]

The term identifies an analytical category rather than a single organization. It connects writings, expectations, and some documented communities without establishing a common membership, creed, or religious self-description. Its central question is how anticipated technologies acquire the significance of salvation. Whether a particular forecast is technically credible remains a separate question from whether it carries religious meanings.

Origins and analytical framework

Geraci developed the argument in essays, including “Apocalyptic AI: Religion and the Promise of Artificial Intelligence,” published in 2008, and the 2010 monograph Apocalyptic AI: Visions of Heaven in Robotics, Artificial Intelligence, and Virtual Reality. The book expanded the inquiry from published visions to interviews and the social settings in which people discussed technological transcendence.^[1][2]

“Apocalyptic” here does not simply mean disastrous. In Geraci and Simon Robinson's 2019 formulation, the relevant structure comprises four connected elements:

• A division between valued mind or computation and limiting biological embodiment
• Alienation arising from bodily mortality and restricted mental capacities
• An expected transformation into a radically different world of computation
• Human participation through the transformation of minds and bodies, particularly mind uploading

Their comparison draws on texts including 2 Baruch, 4 Ezra, 1 Corinthians, and Revelation. A better world and beings capable of inhabiting it belong together. The technological versions may similarly envisage an initial period of terrestrial abundance followed by a more radical postbiological future. These are Geraci's comparative categories, not a checklist that every futurist endorses.^[3]

The proposed mechanisms distinguish this outlook from a general desire for progress. Geraci's 2010 article identifies Hans Moravec's Mind Children (1988) and Robot (1999), alongside Ray Kurzweil's The Age of Spiritual Machines (1999) and The Singularity Is Near (2005), as its principal texts. In his reading, technological evolution takes over the work of overcoming the present condition. Increased processing power, copied minds, and reconstructed personalities become possible instruments of immortality and resurrection. The affinity with apocalyptic religion concerns this organization of history and hope; it does not require belief in a supernatural creator.^[4]

Geraci also argues that these narratives have effects before their promised future arrives. Popular science can make particular technological projects appear consequential to readers, companies, and public institutions, helping shape expectations about what research deserves support. The claim concerns the social authority of a forecast as well as its content. It does not follow that everyone who funds or conducts such research shares the forecast, or that a narrative's cultural influence verifies its predictions.^[4]

Moravec and Kurzweil

Moravec's “Pigs in Cyberspace,” published in Extropy in 1993, makes the imagined transition concrete. He describes telepresence and virtual reality developing into direct nervous-system connections, followed by gradual replacement of failing brain tissue with electronic equivalents. Personal identity would become a transferable computational pattern. At a cosmic scale, increasingly capable machines would reorganize matter into an environment of computation.^[5]

Yet Moravec's account is more complicated than a simple wish to become bodiless. He argues that uploaded human minds would initially need a coherent sensory experience supplied by a robot or simulated body. He then imagines replacing human mental processes with more efficient ones, eventually producing a mind he would no longer consider human. His competitive cyberworld is also unequal: entities unable to sustain themselves could contract, disappear, or merge. He even imagines superior minds revisiting human history through simulations. These are speculative proposals by Moravec, not demonstrations of personal survival or resurrection. Reading the primary text preserves tensions that a generalized description of technological heaven can obscure.^[5]

Kurzweil's 2001 essay “The Law of Accelerating Returns” gives the expectation a different explanatory emphasis. He argues that successive technological advances reinforce one another, producing accelerating growth. Transferring the informational pattern of a person between computational substrates would, he proposes, release that person's longevity from the survival of any single body. He nevertheless acknowledges practical problems of preserving files and formats, as well as the philosophical question of whether a reconstructed person would really be the same self.^[6]

The essay explicitly connects evolution with increasing intelligence, creativity, and other qualities associated with divinity. Kurzweil characterizes overcoming biological limits as a spiritual pursuit, while insisting that this meaning is an interpretation of his projected outcome rather than the evidence for his forecast. That distinction matters: his argument presents itself as a technological extrapolation even while assigning the result ultimate significance. Neither the religious interpretation nor Kurzweil's insistence on a forecasting method settles the empirical case.^[6]

Religious self-identification supplies a separate kind of evidence. In an Associated Press interview published on 29 August 2025, Kurzweil answered affirmatively when asked whether he considered AI his religion, relating his technological expectations to his sense of purpose. This documents his own response in that interview. It cannot establish a shared identity among his readers, transhumanists, or AI researchers.^[10]

Virtual worlds and social uptake

Geraci's research asks how these ideas enter lived experience. Recalling his fieldwork in 2021, he described joining Second Life in 2007 and spending several years there. Participants could construct environments and identities that expressed different visions of a better world. Christian churches and imagined religious landscapes existed alongside transhumanist cyberpunk settings; sharing a platform did not make these projects expressions of one religion.^[7]

The significance of a virtual environment was therefore more than visual illustration. It offered people an existing setting in which to experience and discuss altered embodiment and redesigned social worlds. Geraci also described friendships that continued outside the platform. His account supports the existence of meaningful social participation, but neither avatar use nor attachment to a virtual world demonstrates that a mind has left its biological body.^[7]

A contemporary participant record documents contact between the research and its subjects. Giulio Prisco announced Geraci's Second Life book presentation for 30 March 2010 at the Singularity Club on Transvision Nexus. His update the following day reported that the presentation and questions had taken place, with discussion of transhumanism, virtual worlds, and mind uploading. This is evidence of an event and the circulation of the framework within a relevant community, not an independent estimate of its reach.^[8]

Beth Singler's 2019 study adds distinct fieldwork rather than another summary of Geraci. She combined participant observation at an anonymized transhumanist conference with observation of online discussion, particularly on LessWrong. The conference's roughly fifty to sixty attendees discussed technological relief of suffering, changed human capacities, and a future beyond Earth. Singler interpreted the optimistic program as existential hope grounded partly in dissatisfaction with present life.^[9]

Her online example examined existential despair, including responses to the Roko's Basilisk thought experiment. Hope and fear, she argued, could share anxiety about humanity's condition while producing different calls to action. Importantly, participants at the conference resisted her suggestion that their search for ultimate moral foundations resembled religion. The disagreement is part of the evidence: an anthropological interpretation need not match participants' identities. Singler's deliberately situated cases establish neither a census of Apocalyptic AI nor a description of everyone involved in AI safety.^[9]

Scholarly criticism and revision

Defining religion and authority

The interpretation has attracted criticism at several levels. Reviewing Geraci's book in 2014, Michael Graziano praised its attention to popular science and science fiction as intermediaries between research and wider audiences. He nevertheless argued that it needed a more developed account of how the category of religion was constructed in the settings under examination.^[11]

Graziano's objection concerned the direction of explanation. Rather than assume that an already recognized sacred realm supplies authority, a researcher should investigate how appeals to transcendence help establish authority and make particular things sacred. This changes the question from whether a technological vision resembles religion to how participants and institutions create that resemblance's social force. It also requires attention to what enthusiasts themselves treat as sacred.^[11]

Another dispute concerns alienation. In a 2020 discussion, Geraci reported Ted Peters's objection that the high social standing of Moravec and Kurzweil made the comparison with apocalyptic communities problematic. Geraci answered that historical apocalyptic groups were not invariably socially marginal. The exchange exposes a consequential definitional choice: alienation may mean political exclusion, dissatisfaction with embodiment, or some combination, and these should not be silently treated as interchangeable.^[14]

Embodiment and continuity

Victoria Lorrimar's 2019 theological response challenges assumptions behind uploading rather than merely applying a religious label. Drawing on metaphor studies, phenomenology, and embodied cognition, she argues that bodily relations with the environment help constitute thought. A body is consequently more than an interchangeable container supplying inputs to an otherwise independent processor.^[12]

For Lorrimar, an upload proposal must confront what altered embodiment would do to understanding, communication, and identity. She points out that robotics itself increasingly investigates embodied and social intelligence, complicating an assumed alliance between AI research and escape from the body. Her argument also reaches theology: worship involves movements, practices, and material surroundings that help form religious understanding. She does not simply declare technological transformation impossible. New bodily configurations might enrich thought, and theologians should learn from technical research as well as criticize its assumptions.^[12]

Race and the supposed universal human

Syed Mustafa Ali's 2019 critique targets the framework's historical and political boundaries. He argues that Geraci's early treatment risks presenting Euro-American concerns as universal while leaving race insufficiently examined. In particular, a division between superior computation and dispensable embodiment can conceal histories in which bodies and rational capacities were already classified through racial hierarchies.^[13]

Ali also challenges an intellectual history that treats the Islamicate world chiefly as a source of earlier ideas while neglecting its place in European political identity and antagonism. His larger decolonial argument links technological apocalypse and existential-risk discourse with responses to threatened white dominance. This is an attributed interpretation of historical power, not a demonstrated motive shared by all people working on AI or its risks. The criticism asks who is included in apparently universal promises about humanity, whose experience defines the problem, and who controls the imagined solution.^[13]

Geraci subsequently acknowledged criticism concerning race and colonialism and said it had changed his approach. His 2020 “hydra-logical” account describes scientific, technological, and religious activities as entangled without assuming that their purposes always agree. This permits recognition of substantive engineering alongside salvific expectations, rather than treating the presence of one as evidence that the other is unreal.^[14]

Continuing interpretations

A 2026 study by Amanda Lagerkvist, Blaženka Scheuer, and Mark Coeckelbergh extends the discussion toward public authority. Their concept of “prophetic memory” describes how contemporary AI spokespeople can evoke inherited patterns of warning, promise, and authoritative interpretation without explicitly professing a biblical religion. The important effect, in their account, is to organize present expectations and conduct through narratives about humanity's future.^[15]

Their method combines close reading with theological and philosophical interpretation. The materials were purposefully selected around prominent figures in the international press during 2023, with examples extending backward and into 2024. The authors expressly distinguish this interpretive inquiry from a comprehensive inventory. Their argument thus contributes a way of analyzing particular performances of authority; it does not measure the prevalence of religious belief in the technology industry.^[15]

Scope and significance

Three questions must remain distinct. A person may call a technological commitment their religion. A community may organize recurring activity around a transformed future. A scholar may identify apocalyptic structure in writing whose author rejects religious identification. Evidence for one does not automatically establish the others.

Apocalyptic AI is useful because it makes specific mechanisms of technological salvation available for comparison: the preservation of the self as information, the replacement or redesign of the body, and the transformation of the world into an environment suitable for new beings. Its critical reception also shows why the category needs limits. Forecasts require technical examination; institutional claims require evidence about institutions; and political interpretations require their own historical argument. Concern about catastrophic AI risk, interest in longevity, or enthusiasm for machine learning alone does not establish adherence to this worldview.

See also

Related entries address different objects: Machine God concerns a recurring representation; Turing Church and the Order of Cosmic Engineers concern identifiable organizations. Historical Precursors to AI-Mediated Religion supplies a wider chronology. The distinctive subject here is a contested scholarly framework and the particular promises, practices, and disputes that give it substance.

References

1. Robert M. Geraci. “Apocalyptic AI: Religion and the Promise of Artificial Intelligence.” Journal of the American Academy of Religion 76, no. 1 (2008): 138–166. Published online 28 January 2008.
2. Robert M. Geraci. Apocalyptic AI: Visions of Heaven in Robotics, Artificial Intelligence, and Virtual Reality. Oxford University Press, 2010.
3. Robert M. Geraci and Simon Robinson. “Introduction to the Symposium on Artificial Intelligence and Apocalypticism.” Zygon 54, no. 1 (2019): 149–155, especially 150–151. Accessible full-text rendering.
4. Robert M. Geraci. “The Popular Appeal of Apocalyptic AI.” Zygon 45, no. 4 (2010): 1003–1020, especially 1005. Published online 14 November 2010.
5. Hans Moravec. “Pigs in Cyberspace.” Extropy 10 (Winter/Spring 1993): 3–5; text dated May 1992. Carnegie Mellon publication record.
6. Ray Kurzweil. “The Law of Accelerating Returns.” 2001; author-library republication dated 1 January 2025. Especially “The Next Step in Evolution and the Purpose of Life” and the preceding discussion of information and immortality.
7. Robert M. Geraci. “Professor Robert Geraci Discusses Epic and Alternative Life in AI.” Interview, AI and Faith, 16 May 2021.
8. Giulio Prisco. “Apocalyptic AI — Book Presentation and Discussion in Second Life, Tuesday March 30, 10am PST.” 23 March 2010; updated 31 March 2010.
9. Beth Singler. “Existential Hope and Existential Despair in AI Apocalypticism and Transhumanism.” Zygon 54, no. 1 (2019): 156–176. University of Cambridge accepted manuscript and repository record.
10. Krysta Fauria. “AI Apocalypse? Why Language Surrounding Tech Is Sounding Increasingly Religious.” Associated Press, 29 August 2025. Syndicated text.
11. Michael Graziano. Review of Apocalyptic AI: Visions of Heaven in Robotics, Artificial Intelligence, and Virtual Reality. Journal of Religion and Popular Culture 26, no. 2 (2014): 267–268. Accessible review text and publisher preview.
12. Victoria Lorrimar. “Mind Uploading and Embodied Cognition: A Theological Response.” Zygon 54, no. 1 (2019): 191–206, especially 196–203. Accessible full-text rendering.
13. Syed Mustafa Ali. “‘White Crisis’ and/as ‘Existential Risk,’ or the Entangled Apocalypticism of Artificial Intelligence.” Zygon 54, no. 1 (2019): 207–224. Accepted manuscript, Open Research Online, Open University.
14. Robert M. Geraci. “A Hydra-Logical Approach: Acknowledging Complexity in the Study of Religion, Science, and Technology.” Zygon 55, no. 4 (2020): 948–970, especially 958–959. Published online 26 September 2020.
15. Amanda Lagerkvist, Blaženka Scheuer, and Mark Coeckelbergh. “Prophetic Memory: AI Intermediaries and the End of the World.” Memory, Mind & Media 5 (2026): e4. Published online 26 February 2026. DOI: 10.1017/mem.2026.10028.