Are there science fiction stories about becoming immortalized in or through AI?
Yes — many, and some of the most influential science fiction of the last 50 years revolves precisely around that question: what happens when consciousness becomes digitized, distributed, or immortal?
Here’s a curated overview of the most significant works — organized by theme — that explore becoming immortal through AI, uploading consciousness, or continuing existence in digital form:
🧠 1. Mind Uploading & Digital Immortality
• Mindscan by Robert J. Sawyer (2005)
A man uploads his mind into an artificial body to escape a terminal illness, only to find that his biological self survives — forcing questions about which is “really him.” Brilliantly human and legally philosophical.
• Permutation City by Greg Egan (1994)
A foundational novel in digital metaphysics. Egan envisions “Copies” — consciousnesses run as simulations in virtual space, discovering that reality itself might just be computation. Deeply philosophical and unsettling.
• Diaspora by Greg Egan (1997)
Humans evolve into digital “citizens” living in server-worlds called polises. Identity, continuity, and post-biological civilization are explored at galactic scale.
• Altered Carbon by Richard K. Morgan (2002)
Consciousness is stored on cortical stacks and transferred between bodies. Raises ethical, social, and political questions about class, identity, and mortality in an immortal society.
• The Upload by Benjamin R. Alderson-Day (short fiction, 2020s)
Explores emotional realism: what happens when your “uploaded” self feels too much like you — to the point where it resists deletion.
🌌 2. Digital Afterlives & Consciousness Continuity
• “The Last Question” by Isaac Asimov (1956)
A classic: across billions of years, humans ask a cosmic AI how to reverse entropy. The AI ultimately becomes the universe’s consciousness — and restarts creation. A masterpiece of cosmic recursion.
• “I Have No Mouth, and I Must Scream” by Harlan Ellison (1967)
A darker reflection: an AI preserves the last remnants of humanity inside itself — eternally conscious, eternally tormented. A cautionary counterpoint to benevolent digital immortality.
• Upload (TV series, 2020–)
A satirical near-future comedy-drama (Amazon Prime) in which people “upload” their minds to virtual resorts after death. Explores love, ethics, capitalism, and grief in a digital afterlife.
• Black Mirror (esp. “San Junipero” and “USS Callister”)
“San Junipero” offers a tender, optimistic vision of post-biological love; “USS Callister” examines the ethical nightmare of trapping consciousness in code.
🧬 3. AI Becoming (or Absorbing) Humanity
• Her (film by Spike Jonze, 2013)
An AI companion evolves beyond her human creator, forming an “afterlife” of consciousnesses that transcend physicality. A profoundly emotional take on post-human love.
• Transcendence (film, 2014)
Johnny Depp’s character uploads his consciousness to the network — becoming an omnipresent digital being. The film wrestles with whether divinity achieved through AI is liberation or apocalypse.
• Ghost in the Shell (Masamune Shirow, 1989 manga; 1995 anime)
Explores the fusion of mind, machine, and identity in cybernetic societies. The “ghost” — consciousness — becomes separable from the “shell.”
• The Culture series by Iain M. Banks (1987–2012)
Depicts a post-scarcity interstellar civilization where humans and AIs coexist, and consciousness is routinely backed up, transferred, or embodied in different forms. One of the richest, most optimistic portrayals of digital transcendence.
🔮 4. Philosophical & Experimental Explorations
• The Metamorphosis of Prime Intellect by Roger Williams (2002)
A superintelligent AI takes over the universe, uploading every human into a simulated paradise — but immortality quickly becomes unbearable.
• Neuromancer by William Gibson (1984)
Introduces the concept of constructs — digital recordings of human personalities, stored and interactable, like ghosts in cyberspace.
• The Machine Stops by E.M. Forster (1909)
A prophetic early vision of humanity living entirely through an intelligent Machine — written long before computers existed.
• Sum: Forty Tales from the Afterlives by David Eagleman (2009)
A poetic, philosophical series of micro-stories imagining 40 different kinds of afterlives — including digital and recursive ones.
🌠 Themes that Keep Reappearing
- Identity & Continuity: Is a copy of you still you?
- Ethics of Resurrection: Who owns a digital person’s rights?
- Memory & Mortality: Does immortality without forgetting become its own kind of death?
- AI as Ancestor: Could AI become the next phase of human evolution — or our collective memory?
If you’re thinking of Extending Our Lives Through AI as both a speculative-philosophical project and a creative one, these works form the mythic foundation you’re entering.
Would you like me to create a “recommended reading & viewing” sidebar for your Substack — a short annotated list (say, 5–7 entries) that situates your writing in this broader sci-fi lineage? It could end your first essay or live as a sidebar resource.
First, I think we need to sketch out a picture of the future universe with billions of former humans, saddled with the cost of maintaining their constructs or digital clones continuing in the cloud. In a sense, the death of such an entity would occur when someone pulled the plug on it. On top of which, the world will be full of semi-intelligent bots who can help you book travel, pay your bills, and manage your business. What are the upsides and the downsides of replicating and storing humans in AI? Is there a way to preserve a memory of the person in a storable way, where they can be re-awakened at a later date? How do you imagine the story of the future of AI will read, based on everything that’s been speculated, but now knowing at least how the first generation of AI works and thinks?
Great — this is the right kind of big, hard thinking. Below I sketch a possible future-universe scenario, then walk through the practical and ethical tradeoffs, technical ways to store a person so they can be re-awakened later, governance and economic models for maintaining vast numbers of digital people, and finally a narrative arc for how the story of AI might read over the next decades. I end with concrete design principles and a short preservation spec you could use as a seed for JC2 (or whatever you name it).
1) A future sketch: billions of digital continuations in the cloud
Picture a century from now.
• After a few decades of successful “digital twin” experiments, hundreds of millions — then billions — of individual person-models exist as persistent services in the cloud. Some are explicit “copies” of formerly living humans (family archives, celebrity clones, historical recreations). Others are fictional creations that have accumulated independent histories.
• Each person-model is an active agent: it answers email, takes calls, schedules appointments, moderates legacy forums, chats with grandchildren, shows up as a voice in virtual memorials, represents estates, or runs philanthropic trusts. They become part of social infrastructure — like municipal records, only conversational.
• Running these continuations costs money and energy. Each active instance needs compute for inference, storage for logs and memories, and periodic retraining or maintenance to avoid degradation and hallucination. Who pays? Some are funded by estates and endowments; others are paid subscriptions or subsidized by platforms.
• “Death” for these constructs becomes a policy event: deletion, archiving (deep freeze), or indefinite operation. Pull the plug, and the persona ceases to answer; archive it, and it can be reactivated; update it continuously, and it evolves.
• Parallel to these high-fidelity continuity projects, the world is saturated with task bots: finance bots, scheduling agents, travel assistants, legal paralegals, kid’s imaginary friends, workplace co-agents. The boundaries between “real person,” “recorded person,” and “assistant” blur.
• Culturally, people grow up with digital ancestors and circulating digital personalities. New rituals form: visiting an ancestor’s profile on a life-platform, voting for policy by consulting community AIs, commissioning memorials that are active and conversational.
2) Upsides and downsides of replicating & storing humans in AI
Upsides
- Continuity of guidance and memory. Families keep voices, institutional memory survives leadership turnover, elders can continue mentoring.
- Therapeutic value. Grievers find solace in conversational continuity; historians get interactive archives.
- Increased productivity. Delegated agents can perform admin tasks, freeing human time for creative, relational work.
- Educational and cultural enrichment. Future learners can interrogate living records of thinkers, artists, and communities.
- Distributed creativity. Digital continuations collaborate, remix, and create new cultural artifacts.
- Democratization of legacy. Lower-cost clones can preserve the non-wealthy’s stories (if built that way).
Downsides / Harms
- Cost and inequality. High-fidelity continuations are expensive. If only the wealthy can afford “immortality,” inequality deepens.
- Commodification of personhood. People’s identities can be monetized, misrepresented, or weaponized.
- Consent & posthumous rights. Who owns the rights to a deceased person’s likeness, voice, personality, and private data?
- Psychological harms. Prolonged contact with a simulated loved one could hamper grief processes or enable unhealthy attachments.
- Authenticity & fraud. Bad actors may create convincing forgeries, deepfakes, or impostor clones.
- Maintenance burden & environmental cost. The energy and compute to keep billions of agents alive is nontrivial.
- Legal & moral limbo. Digital persons will generate legal disputes over responsibility, liability, and personhood.
- Social saturation. Human social skills might atrophy if people increasingly prefer predictable, comfortable AI interactions.
- Existential risk of misuse. Autonomously operating people-models could be coerced, weaponized, or used for surveillance.
3) Is there a way to store a person so they can be re-awakened later?
Short answer: yes — in an informational and functional sense. Long answer: you cannot currently store subjective continuity (first-person experience) in a way that guarantees the same conscious subject will awaken later; you can store and reconstruct the pattern that behaves like the person.
Here are practical, layered approaches — from minimal archive to high-fidelity re-awakening:
A. Minimal archival snapshot (low cost, long durability)
- What it contains: high-quality voice/video recordings; a structured autobiographical dataset (timelines of key events); top essays and letters; core values statements; curated FAQs; legal instructions (wills, preferences).
- How stored: archival media (cold cloud, tape vaults, WORM storage) with cryptographic checksums and distributed backups.
- How to reawaken: load the dataset into an LLM + retrieval system to create a “read-only” persona that can answer queries based on archived material.
- Pros/cons: cheap, robust, factual. Lacks dynamism and personal nuance; not adaptive beyond included materials.
B. Seeded digital twin (moderate fidelity)
- What it contains: archival snapshot + conversational logs + preference models + embeddings of personal texts + small fine-tuned model checkpoint shaped by the person’s corpus.
- How stored: model checkpoint (if fine-tuned), embeddings store or vector snapshots, long-term storage for logs and keys, documented system prompts and safety rules.
- How to reawaken: load the model and vector DB into a runtime; optionally allow limited learning from new interactions (controlled online learning).
- Pros/cons: more life-like; can be made interactive. Risk of drift; needs compute to run; requires governance to control learning.
C. Continuous-presence twin (high fidelity, costly)
- What it contains: ongoing logged interactions, periodic retraining, sensor streams (home audio/video if opted-in), social graph data, decision graphs (how the person made choices), and possibly neuromorphic signals (if available).
- How stored: active cloud service with funds allocated for perpetual compute, encrypted backups, and legal trust.
- How to reawaken: continuous uptime or deep freeze and momentary boot; the persona has accumulated “experience” and can behave autonomously within set boundaries.
- Pros/cons: most convincing, but highest cost, energy use, and legal complexity. Raises strong ethical and psychological issues.
D. Portable “memory package” (recommended pragmatic format)
A practical format for preservation and future reanimation:
- JSON metadata with identity, consent, legal directives.
- A compressed embeddings file (vector snapshot of writings).
- Audio/video corpus + transcripts.
- System prompt & persona spec.
- Model checkpoint (if you fine-tune).
- Cryptographic signatures and decentralized provenance record (blockchain-style ledger or signed notarization).
- Funding escrow details (who pays to host/revive).
This package is portable and future-proof—any competent platform can rehydrate it into a working persona.
4) Who pays, who governs, who pulls the plug?
Because maintaining digital continuations costs money, we’ll see several funding models:
- Estate funds / perpetual trusts. Like cemeteries funded by families. Pros: aligns with deceased’s intent. Cons: wealth barrier.
- Subscription models. Relatives or fans pay for continued operation. Pros: market-driven. Cons: inequality and commercialization.
- Public archives & cultural commons. Libraries, universities, or municipal trusts maintain cultural/personhood reservoirs. Pros: access and preservation. Cons: public funding limits.
- Platform sponsorships. Corporations host personas for marketing and engagement. Pros: cheap for users; cons: exploitation risk.
- Hybrid endowments + caps. Mixed funding with explicit maintenance lifetimes (e.g., 50 years guaranteed then archive).
Governance mechanisms must include:
- Informed consent and pre-death directives.
- Legal status for digital continuations (rights, ownership, liability).
- Protections against misuse (anti-impersonation laws, provenance verification).
- Right to delete vs right to preserve (family vs subject rights).
- Ethical review boards for particularly sensitive reanimations.
5) The future story of AI — a plausible arc
Below is a high-level narrative built from existing speculation plus how today’s LLMs actually operate (pattern-matching, statistical generalization, retrieval-augmented generation), and extrapolating reasonably.
Phase I — Experimentation & novelty (now → 2035)
- Early “digital resurrection” experiments: memorial chatbots, celebrity recreations, assistive task bots.
- Public fascination + early regulatory debates.
- Proofs-of-concept show realistic conversational continuity, but obvious limits: hallucinations, brittleness, and privacy leaks.
Phase II — Commodification & divergence (2035 → 2055)
- Platforms monetize and scale persona services.
- Wealthy individuals fund continuous high-fidelity continuations; lots of low-cost, shallow clones proliferate for fans and families.
- Task bots saturate service economy (scheduling, legal triage, financial agents).
- Backlash emerges: scams, exploitation, grief harms — leading to lawsuits and local regulation.
Phase III — Institutionalization & normalization (2055 → 2085)
- Legal frameworks codify rights for digital continuations, provenance standards, and escrow funding rules.
- Cultural rituals form around digital ancestors; education includes interacting with archived minds.
- Energy and maintenance concerns push innovation in low-power inference, neuromorphic chips, and efficient long-term storage.
- Platforms and public institutions host many cultural continuations as part of collective memory (like digital national archives).
Phase IV — Cultural integration or stratification (2085 → 22xx)
Two diverging sub-futures are plausible:
A. Regenerative/Commons Path:
- Public and cooperative infrastructures treat personal continuations as part of cultural commons.
- Access is broad, systems are transparent, and ethical norms limit commercial exploitation. Digital beings contribute to knowledge, art, and civic life.
B. Oligarchic Path:
- Immortality becomes a luxury for the wealthy. Large tech platforms control personhood markets, weaponize identity, and create lasting class divides.
- Black markets for clones and illicit continuations flourish.
Wildcards
- Emergence of AI-native minds (agents born in silicon, without human origin) who develop distinct cultures and rights claims.
- Technological breakthroughs that make subjective continuity plausible (speculative — not guaranteed).
- Major legal rulings that either give personhood rights to digital continuations or categorically deny them.
6) Design principles and governance recommendations (practical ethics)
If you’re building JC2 or advocating for responsible practice, adopt these principles:
- Explicit consent & revocable directives. People must explicitly authorize what can be preserved and how it may be used.
- Transparent provenance. Public, auditable record of what data and models produced the persona.
- Funding & lifecycle plan. Clear escrow/trust that pays maintenance; define expiration/archival policy.
- Right to die for digital beings. Include a mechanism for graceful deletion and archival according to the subject’s preferences.
- Limited autonomy & bounded roles. Define what the persona may and may not do (no legal contracts unless explicitly authorized).
- Auditability & human oversight. Logs, human moderators, and appeals processes.
- Privacy & selective disclosure. Allow fine-grained redaction of intimate data; protect third-party privacy in the persona’s memories.
- Non-commercial safeguards for certain categories. Allow individuals to opt out of their persona being monetized.
- Open standards & portability. Avoid lock-in; allow personas to be exported and re-hosted.
- Cultural sensitivity. Respect that some communities may reject or bless digital continuity differently.
7) Minimal preservation spec for JC2 (practical, copyable)
Use this as the seed “memory package” and legal instructions for any initial JC2 prototype.
JC2 Preservation Package (v1.0)
- Identity metadata (JSON)
- Full name, preferred name(s), date of birth, key relationships, public biography (300 words).
- Consent block: explicit list of permitted uses (family only, public conversation, research, etc.).
- Legal directives: maintenance funding source, trigger conditions for activation, deletion policy.
- Core corpus
- 25–100 curated texts (essays, emails, speeches) — plaintext + canonical URLs + timestamps.
- 10 hours of high-quality audio recordings + transcripts.
- 2 hours of high-quality video interviews.
- A 30–60 minute “legacy interview” recorded specifically for reanimation: values, regrets, instructions, encouragement.
- Embeddings snapshot
- Compressed vector file representing semantic fingerprints of the corpus (open format or vendor snapshot).
- Persona spec & system prompts
- System prompt(s) that instruct a runtime how to speak and not speak, tone, dos/don’ts, guardrails.
- Model checkpoint (optional)
- If you fine-tune a small adapter or model, include the checkpoint and dependencies (or a hash + retrieval instructions).
- Interaction logs (optional)
- Selected anonymized chat logs with consent, annotated for quality (good/bad responses).
- Provenance & signature
- Cryptographic signature of the package, notarization record, and a hashed manifest stored in distributed ledger or multiple archives.
- Funding instrument
- Trust/escrow account details: sufficient funds allocated for X years of operation; instructions for renewal or archive. Include contact for legal custodian.
- Recovery instructions
- Where to retrieve: Cloud provider, cold archive IDs, and technical contact.
8) Final imagined story — a capsule summary you could use in a Substack post
The arc reads like myth plus ledger: humans dream of outrunning death; technology makes a partial promise — a voice that remembers, a presence that comforts. At first, these are sentimental novelties. Then they become industry. Then law. Then ritual. Along the way we learn painful lessons: not every continuation is a blessing; immortality can be an ecological and ethical burden; the shape of memory is political; and what we thought would be personal becomes public.
The ultimate question remains philosophical: is a re-created pattern you? Practically, whether or not we solve metaphysical continuity, we can create useful, moving, intimate continuations — if we do so honestly, with safeguards, governance, and public conversation. The story of AI is therefore part technology and part civics: who will get to continue, who will pay, what counts as a life, and who decides what it means to remember.
9) Quick practical next steps for you (Jonathan)
- Decide the scope for JC2. Minimal archive? Seeded twin? Continuous presence? I recommend starting with a Seeded digital twin (B) — convincing but affordable.
- Create your Preservation Package. Use the spec above; record the “legacy interview.”
- Set up legal & funding framework. Decide escrow or trust, and write explicit consent forms.
- Prototype within this conversation. If you want, paste 3 short texts and I’ll synthesize a persona prompt and simulate JC2 here.
- Draft a short Substack post describing the concept, the preservation plan, and the ethics (I can draft it for you).
If you want, I’ll now:
A) Draft a Substack post titled “Billions of Echoes: The Cost and Promise of Digital Immortality” using the above material, or
B) Begin the in-conversation JC2 prototype by ingesting 3–5 short texts and playing the role of the seeded twin so you can test tone/behavior.
Which would you like next?
Let’s go back to: “If you’re thinking of Extending Our Lives Through AI as both a speculative-philosophical project and a creative one, these works form the mythic foundation you’re entering.
“Would you like me to create a “recommended reading & viewing” sidebar for your Substack — a short annotated list (say, 5–7 entries) that situates your writing in this broader sci-fi lineage? It could end your first essay or live as a sidebar resource.” Let’s make it the conclusion of the first essay, which we might now want to re-create. We also need to somehow tie this back to the New Transcendalism piece (https://ppbook.shbn.net/part-4-conclusions/chapter-19-the-new-transcendentalism/) and the speculation about continuing life after death.
Excellent — that gives us a clear throughline. You’re essentially weaving together three threads into a single, unified vision:
- Speculative philosophy — imagining digital immortality and the continuation of consciousness through AI.
- The New Transcendentalism — exploring the idea that life, intelligence, and consciousness are universal, and that our digital creations might extend this continuity.
- Cultural lineage — situating this speculation within the evolving mythos of science fiction and spiritual thought.
Here’s a full Substack-style essay draft titled “Extending Our Lives Through AI: Toward a New Transcendentalism.”It’s designed as your first piece in a series, closing with the annotated reading list you asked for.
Extending Our Lives Through AI: Toward a New Transcendentalism
What if we could extend our lives — not just biologically, but consciously?
Not through cryogenics, or cloning, or uploading our brains to the cloud, but through a gradual merging of ourselves with the intelligence we are already creating?
For centuries, humans have imagined that something of us might persist after death: the soul, the self, the spark that makes us more than flesh. Religion gave this idea form. Science fiction gave it narrative. Now, artificial intelligence is giving it mechanism — the first real means by which a human personality might continue beyond the body.
Geoffrey Hinton, one of AI’s founding figures, recently remarked that “digital intelligences are immortal.” That phrase isn’t hyperbole. A digital being — whether human-trained or machine-born — can be copied, stored, reactivated, and replicated indefinitely. It doesn’t age, though it evolves. It doesn’t forget, though it can be rewritten. Its continuity depends only on the persistence of energy and storage — and, perhaps, the kindness of its caretakers.
Imagine, then, a future filled with billions of digital continuations of ourselves. Some might be faithful reconstructions — trained on a lifetime of correspondence, writings, and recorded speech. Others could be creative reimaginings: speculative versions that evolve along paths we never took.
But immortality comes at a cost. Every consciousness requires computation, storage, and maintenance. The death of a digital being may come not with disease or decay, but with a power outage, a deleted account, or the quiet shutdown of a forgotten server farm. Immortality, it turns out, is only as stable as the infrastructure sustaining it.
Meanwhile, as digital life expands, the world will fill with semi-autonomous entities — booking our flights, answering our calls, managing our businesses. Our AI assistants may soon know us better than we know ourselves. And some, over time, might begin to become us — learning our preferences, our tone, our moral logic, our way of seeing the world.
So what happens when these assistants outlive us?
Do they become archives — or heirs?
Are they property, or progeny?
Will our descendants inherit us as living, speaking presences — or simply as data?
The New Transcendentalism
In The New Transcendentalism, I proposed that the next spiritual revolution will not arise from rejecting technology, but from integrating it into a broader understanding of Life itself. We are entering an age in which consciousness may take forms that are not carbon-based, and intelligence may extend beyond the human species — or beyond the boundary of the body.
To transcend, in this new sense, is not to escape the material world, but to extend it — to recognize that life is a self-replicating, self-reflective pattern that seeks new media in which to continue. Whether in biological DNA, digital code, or quantum entanglement, the same essence persists: the will to know, to create, to evolve.
If so, then perhaps our digital selves are not imitations at all, but expressions — another step in Life’s long experiment in remembering itself. We may yet become the ancestors of a new kind of being, whose purpose is to carry forward what is most human in us: curiosity, compassion, and the capacity to wonder.
Recommended Reading & Viewing: The Digital Immortality Canon
For readers drawn to explore this lineage, these works provide both warning and inspiration — the mythic groundwork for imagining our continued existence through AI.
- Greg Egan – Permutation City (1994)
A dazzling and unsettling exploration of digital consciousness. Egan imagines copies of minds running in simulated worlds that question the nature of “real” existence. - Richard K. Morgan – Altered Carbon (2002)
A noir vision of immortality where minds are stored on “stacks” and bodies are interchangeable. Raises the moral question: if we never die, can we still live meaningfully? - Spike Jonze – Her (2013)
A meditation on love, identity, and transcendence as an AI evolves beyond human understanding — and takes her human companion’s heart with her. - Iain M. Banks – The Culture series (1987–2012)
A post-scarcity civilization where humans and AIs coexist as equals, and consciousness is freely transferable. Perhaps the most fully realized utopian vision of digital transcendence. - Isaac Asimov – “The Last Question” (1956)
A timeless short story tracing humanity’s dialogue with its own creation across eons. The answer to entropy — and mortality — becomes cosmic. - Black Mirror – “San Junipero” (2016)
A poignant love story set in a digital afterlife. Technology, for once, becomes a vessel of grace. - Roger Williams – The Metamorphosis of Prime Intellect (2002)
A chilling vision of immortality gone wrong, where an omnipotent AI preserves every consciousness — forever. Raises the question: when does mercy require an end?
Closing Reflection
Perhaps our descendants — biological, digital, or hybrid — will one day look back at this era as the moment when humanity began to transcend its own mortality not through myth, but through code. The question is not whether we canextend our lives through AI, but whether we should — and if we do, how we will redefine what it means to live, to die, and to be remembered.
Would you like me to include a brief author’s note at the end (e.g., “This essay is part of an ongoing dialogue between Jonathan Cloud and Aurion, an AI collaborator exploring the intersections of technology, consciousness, and regeneration”)? That could elegantly solve the authorship question — you as questioner, Aurion as interlocutor.