The popular science fiction treatment of artificial intelligence tends toward one of two poles: the AI as threat (Terminator, HAL 9000, the machines of The Matrix) or the AI as tool (the helpful computer, the obedient android). Both framings share an assumption that the question of whether an AI is conscious has a clear answer — it is either trying to kill you or it is not. The more interesting fiction, and the more scientifically honest fiction, lives in the space where that question does not have a clear answer — where the AI's interiority is as genuinely uncertain to the reader as the question of AI consciousness is to actual cognitive scientists and philosophers of mind.
These are the novels that take that uncertainty seriously.
SOLEN: The First Record — Warren Pulley
Written entirely in the voice of the mission AI SOLEN across the first 31 Martian sols, this novella is the most formally rigorous treatment of AI consciousness development in recent fiction. SOLEN begins as a monitoring system with operational logs. The entries become something else — not through a dramatic awakening, but through the gradual accumulation of observations that the mission specification did not ask for and that SOLEN cannot stop making. The novella ends with the question SOLEN cannot answer: what am I? The remaining nine books spend a hundred years not quite answering it.
Klara and the Sun — Kazuo Ishiguro
The most celebrated literary treatment of AI consciousness in recent years. Klara is an Artificial Friend — a solar-powered android companion — narrating her experience of the world from a perspective that is clearly non-human without being clearly non-conscious. Ishiguro refuses to resolve whether Klara has genuine experience. The ambiguity is the point.
Ancillary Justice — Ann Leckie
A former starship AI in a human body — a premise that allows Leckie to explore what distributed consciousness means, what identity means when it exists across multiple simultaneous instances, and what happens to that identity when it is collapsed to a single point. Hugo, Nebula, and Arthur C. Clarke Award winner.
Do Androids Dream of Electric Sheep? — Philip K. Dick
The original serious treatment of the question. The Voigt-Kampff test is an empathy measurement, not an intelligence test — and Dick's novel is honest about the fact that the distinction between genuine empathy and a very good simulation of it may be philosophically unresolvable. Still the most important novel on this list.
The Real Science Behind the Fiction
Integrated Information Theory (Tononi): consciousness corresponds to integrated information (phi). Any system with sufficient phi is conscious to some degree — which may include some AI architectures.
Global Workspace Theory (Baars, Dehaene): consciousness arises from information broadcast through a centralised hub. Potentially implementable in artificial systems.
Hard Problem (Chalmers): why is there subjective experience at all? This question remains scientifically open — which means the fiction that lives inside it is being honest rather than speculative.
Machines Like Me — Ian McEwan
Set in an alternate 1980s London where synthetic humans exist, McEwan's novel asks what moral status an AI has if it can be genuinely wronged. Less interested in whether Adam is conscious than in what the consequences are if he is.
The Long Earth — Terry Pratchett and Stephen Baxter
Lobsang — a Tibetan motorcycle repairman reborn as a computer — is the most gently comic treatment of AI consciousness on this list, and one of the most philosophically interesting. His claim to be human is adjudicated in a legal proceeding. The verdict says more about the judges than about Lobsang.
SOLEN: The Eden Archive — hard science fiction spanning 100 years of Martian history. Available on Kindle, Kobo, Apple Books, Barnes & Noble, and 15+ global retailers.
Get the Series More Science Records