Superintelligence
An intellect that greatly exceeds humans in every domain — a concept whose most important argument is a 1965 speculation about what such a thing would build next.
Reviewed July 16, 2026Stable
When not to use it
- As a product consideration. Nothing you're building is affected, and invoking it usually replaces the specific question with a bigger vague one.
- As a synonym for AGI. AGI means human-level generality; superintelligence means well past it. The gap between them is exactly what's under dispute.
- As a settled premise. It's an argument from 1965 that has never been empirically tested, and treating it as established is a rhetorical move.
Reach for something else instead
- AGI is the nearer, more defined claim, and usually the one people actually mean.
- Specific capability thresholds — what can it do, at what cost — are measurable, which is why frontier regulation reaches for them.
- Concrete near-term risks — misuse, bias, security, labour — are happening and don't require any of this to be true.
Sources & further reading
- Good (1965), Speculations Concerning the First Ultraintelligent Machine — the intelligence-explosion argument in its original form.
- Bostrom (2014), Superintelligence: Paths, Dangers, Strategies — the modern treatment; orthogonality, instrumental convergence, the control problem.
- Chalmers (2010), The Singularity: A Philosophical Analysis — a careful examination of whether the explosion argument actually goes through.
Primary sources, listed so you can check the claims on this page rather than take them on trust.
Where people go wrong
- Conflating three separate claims: better-than-human at everything, recursively self-improving, and having goals of its own. They're independent, and arguments slide between them.
- Assuming the explosion argument is established. It's a 1965 speculation whose key premise — that intelligence is the bottleneck on producing intelligence — is exactly what the compute-hungry last decade calls into question.
- Treating it as unserious because it sounds like science fiction. It's a real argument with real weaknesses, and dismissal-by-vibe is the mirror of belief-by-vibe.
At a glance
FieldFoundations
Core argumentthe intelligence explosion, Good, 1965
Key premiseintelligence is the bottleneck on intelligence
Statusunresolved, load-bearing, unevidenced
DifficultyBeginner