Echopraxia
A follow-up to the Hugo Award-nominated Blindsight, Echopraxia is set in a 22nd-century world transformed by scientific evangelicals, supernatural beings and ghosts, where defunct biologist Daniel Bruks becomes trapped on a spaceship destined to make an evolutionary-changing discovery.
Stats for nerds
45recommend
96mentions
66people
2014–2026years
58% recommend39% neutral3% critical
Mentions per year
What Hacker News says
- > It is also why the doomer argument is so ridiculous. It is a type of superstition that is confusing a real danger. We don't need to "align" the nuclear reactor so it doesn't get angry and meltdown. It is just a machine that we need to make sure doesn't malfunction in a way we don't want it to.
It's not ridiculous, unless you misinterpret it like you did here, and link it to consciousness.
Doomer argument is just that optimizers gonna optimize, and our own intelligence is just an optimization algorithm too, albeit highly general one -- and once we're building optimizers with comparative generality, we're facing possibility of one that out-thinks us and we can't control it anymore. The "alignment problem" is about how to ensure such sufficiently-smart optimizers end up optimizing for the same things we are, because any significant deviation here ends up in doom for us.
Consciousness never even enters the picture here. And it's not a given than it should.
(Incidentally, while hard sci-fi and not a paper, but also entirely unrelated to AI X-risk crowd, the books Blindsight and Echopraxia sketch a pretty convincing argument that consciousness may be an evolutionary disadvantage and it's not only not necessary, but may be holding us back.)
- For Peter Watts - Echopraxia is just as wonderful as Blindsight, highly recommend.
- I have a couple I would highly recommend:
The first is Stanislaw Lem's "Imaginary Magnitude."
Most stories there are written as sort of prefaces to non-existent technical/scientific books, papers, articles, etc. The last one, Golem XIV, goes a bit beyond and includes an entire lecture.
The content is incredibly detailed and technical, Lem knew his Technology and Science for sure, so sometimes it's incredibly accurate and prophetic. Even though it was written in the 1970's, Lem was already describing things that look uncannily like neural networks, emergent communication systems, and machine-generated literature, decades before they existed.
And... most of them are also really really funny.
Here are small summaries that ChatGPT helped me write for some of the stories there:
---- Eruntics ---- A faux-scientific introduction to a field studying “erunts”: biological (or semi-biological) agents, think bacteria/neural colonies, that can learn to read, write, and communicate.
These bacterial colonies learn to communicate through several learning iterations across many mutations and generations, much like, say, Neural Networks, and they even develop their own proto-language as they evolve.
Eruntics is a sort of "meditation" on emergent intelligence and how writing or meaning might arise in nonhuman systems.
---- A History of Bitic Literature ---- Purports to be a critical history of literature produced not by humans, but by machines or other non-human authors, the fictional discipline of bitistics.
It treats computer-generated texts as a full literary tradition, with schools, tropes, and structural analysis.
It also goes into detail on how at a certain point Bitic Literature required models to read what other models had generated and "validate" what they had written, since the complexity they reach in terms of things like neologisms, subtext, and interconnectedness to other texts and ideas, becomes so absurdly high that no human being can even attempt to read them or make sense of them.
---- Golem XIV ---- This is like half of the book, and oh boy...
It presents itself first as a report/lecture series from a U.S.-military supercomputer, framed with forewords, editorial notes, “instructions,” etc, but it quickly shifts into the machine’s own philosophical monologues about humanity, evolution, intelligence gradients, and existential insignificance/significance.
It contains some really iconic phrases, in particular during Golem's "Lecture":
" 'The meaning of the transmitter is the transmission' [...] To be sure, the corollary holds: 'The meaning of the transmission is the transmitter.' But the two members are not symmetrical."
" 'The construction is less perfect than what constructs' [...] Let us give it more substance: 'In evolution, a negative gradient operates in the perfecting of structural solutions'"
Golem XIV becomes a genuine “book within a book,” a hard, speculative essay on what a post-human intellect might think.
The second set of "books" that I would recommend are kinda related, even in tone:
Masamune Shirow's "Ghost in The Shell" and its sequel "Ghost in the Shell: Man-Machine Interface."
ChatGPT summaries:
--- Ghost in the Shell ---
At its core, the original Ghost in the Shell manga is a cyberpunk procedural about identity in a hyper-connected world.
It follows Major Motoko Kusanagi, a fully cybernetic government operative working for Public Security Section 9, a counter-cyberterrorism and intelligence unit. The cases involve:
- cyberbrains and identity hacking
- political corruption
- AI autonomy
- blurred boundaries between human and machine
- networks where memories, skills, and consciousness can be stolen or overwritten
The central philosophical arc centers on Project 2501, the “Puppet Master,” an emergent artificial consciousness born from networked information systems rather than human intention. It challenges the legal, biological, and metaphysical definitions of “life.”
The manga’s key concern:
When minds become software, what counts as a person? What counts as being alive?
It’s high-speed political action wrapped around deep questions about autonomy, embodiment, and consciousness.
--- Ghost in the Shell 2: Man-Machine Interface ---
Man-Machine Interface is not a continuation in the conventional sense, but more of a "post-human" sequel.
Here, Motoko Kusanagi, the main character from Ghost in The Shell, no longer exists as a single embodied entity. After merging with the Puppet Master at the end of the original story, Motoko becomes something new: a distributed, multi-bodied, evolving posthuman intelligence often referred to as Motoko Aramaki or Motoko 2.0.
The sequel follows her as:
- a partially decoupled consciousness running across networks
- an entity capable of switching avatars, bodies, and instantiations
- a mind that operates above nation-states, legal frameworks, and even human-scale cognition
Plot-wise, it deals with:
- AI-industrial warfare
- megacorporations fighting over distributed intelligences
- metaphysics of identity in a world of multiple selves
- data-ecologies and collective minds
- the emergence of “post-persons”
But the real shift is tone and philosophy.
Where the first manga explores the crisis of identity, the sequel explores post-identity.
Its main question:
What does “self” mean when consciousness is fluid, redundant, multi-threaded, and no longer tied to one body or one narrative?
Man-Machine Interface is dense, maximalist, metaphysical, and sometimes abstract , much closer to Lem’s Golem XIV in its treatment of posthuman consciousness than the first part.
Third, another set, Peter Watts’ Blindsight and its sequel Echopraxia.
--- Blindsight ---
Hard sci-fi first-contact novel about a crew sent to investigate an alien signal. The aliens they encounter demonstrate astonishing intelligence and adaptability, but seemingly zero consciousness. The book’s central question is brutal:
Does consciousness provide any evolutionary advantage at all? Or is it an expensive, unnecessary hack?
It’s one of the most rigorous modern explorations of intelligence without awareness, cognition without qualia, and post-biological life. Also features resurrected vampires as humanity’s most efficient problem-solvers.
--- Echopraxia ---
A follow-up set in the same universe, focusing more on Earth, zombies, hive-minds, military religious cults, and what Watts calls baseline humans becoming obsolete. It complements Blindsight by following the implications of a world where non-conscious “intelligence engines” outperform conscious humans.
Finally...
--- Greg Egan's Permutation City ---
At its core:
A group of people upload themselves as software (“Copies”) into a virtual environment where consciousness is implemented as a computational process. But Egan pushes the idea far past “mind uploading” into deep mathematical and philosophical territory.
The central premise:
If consciousness depends only on logical consistency, not computation performed in real time, then a simulated mind does not require a universe running on stable physics or an external computer at all.
Egan calls this the Dust Theory.
What the novel explores:
1. Copies — software persons - Human consciousness is scanned and instantiated as a digital mind. These Copies struggle with:
- psychological trauma from being disposable
- questions about their moral status
- fears of erasure
- pressure to perform labor for “real” biological humans
They are conscious, but economically and ontologically precarious.
2. The “Autoverse” A lifelike world governed by artificial chemistry and evolution. It explores how complexity and life can arise from arbitrary rules, not carbon-based biology.
3. Dust Theory A mind is simply a pattern, not a process. If any arrangement of matter in the multiverse instantiates that pattern, even for a moment, consciousness exists.
This implies that a mind could “run” without a computer — as long as the pattern exists somewhere, even scattered across physical chaos (“dust”).
Permutation City becomes a sandbox for minds that no longer require a substrate.
4. “Permutation City” itself The ultimate simulation, constructed to be self-sustaining in a way that breaks free from dependence on any external reality. Inside it, Copies evolve new forms of existence beyond human constraints.
Why it’s important:
It’s one of the most serious, mathematically grounded attempts in fiction to grapple with:
- the ontology of simulated minds
- the meaning (or irrelevance) of physical reality
- identity as a computational equivalence class
- consciousness divorced from biology or substrate
- universes that exist because their patterns are self-consistent
It anticipates debates around:
- substrate independence
- functionalism
- consciousness as pattern-matching
- AI personhood
- anthropic reasoning
- simulation theory
Before most of those ideas entered mainstream discussion.
- Peter Watts' Blindsight and Echopraxia are the 21st-century evolution of cyberpunk, IMO.[1] It's really too bad he seems to have decided not to continue writing in that fictional world.
[1] They're almost literally Bruce Sterling's corporations-would-turn-Frankenstein's-monster-into-a-product.
- It took me a moment to realize, even after the mention of Echopraxia, that this was Peter Watts.
If you enjoy hard to very-hard science fiction, I strongly recommend the first book of his series, Blindsight. I thoroughly loved the read and bounced right back to the beginning for a second read with the context I'd gained on the first one. It's an absolute firehose of concepts; reminded me a bit of Accelerando by Charles Stross but a little less pleased with its own geekiness. The best summary I could give would be a meditation on consciousness set against a first-contact backdrop.
- You may enjoy Peter Watts, especially Blindsight and its sequel, Echopraxia.
Watts is himself a biologist, with a refreshingly unromantic perspective on humanity's place in the universe.
(His other great story sequence, The Freeze-Frame Revolution, is some of the darkest sci-fi I've read since Harlan Ellison's "I Have No Mouth and I Must Scream".)
