---
aliases:
  en:
  - AI intelligence explosion
concept_id: intelligence_explosion
preferred:
  en: intelligence explosion
  ja: 知能爆発
relations:
  relations:
    prerequisite:
    - artificial_intelligence
    - capability
    related:
    - recursive_self_improvement
    - technology_explosion
status: active
---

# intelligence explosion

A positive-feedback regime in which increasingly capable AI accelerates the research, engineering, or other inputs that produce still more capable AI, yielding a large and sustained acceleration in AI capabilities over many iterations. Recursive self-improvement is one possible mechanism, but the concept can also include distributed AI-driven R&D and does not by itself specify a discontinuity.

## prerequisite

- [artificial intelligence](../terminology/artificial_intelligence.txt)
- [capability](../terminology/capability.txt)

## related

- [recursive self-improvement](../terminology/recursive_self_improvement.txt)
- [technology explosion](../terminology/technology_explosion.txt)

## Documents: about

- [AI 研究開発の自動化だけによる自己加速は、実験用計算資源がボトルネックとなって頭打ちになる可能性がある](../documents/erdil_barnett.ai-r-and-d-automation-alone-may-hit-compute-bottlenecks.txt)
- [AI 研究開発が全面自動化されれば、ハードウェアを増強しなくてもソフトウェア知能爆発は十分に起こり得るかもしれない](../documents/eth_davidson.software-intelligence-explosion-remains-plausible-at-fixed-hardware.txt)
- [AI 研究開発の全面自動化後は、ソフトウェア知能爆発に至らなくても、しばらく非常に速い進歩が続き得る](../documents/eth_davidson.software-intelligence-fizzle-can-still-be-fast.txt)
- [再帰的自己改善は、自己改善能力への正のフィードバックを通じて知能爆発を起こし得る](../documents/yudkowsky.recursive-self-improvement-can-produce-intelligence-explosion.txt)

## Documents: mentions

- [スローテイクオフでは、超知能より前の AI がすでに世界を大きく変える](../documents/christiano.slow-takeoff-allows-large-pre-superintelligence-effects.txt)
- [クランチタイムではAIの労働力を防御用途に振り向け、移行を大幅に遅らせる](../documents/cotra.crunch-time-redirect-ai-labor-and-slow-transition.txt)

[Corpus index](../index.txt)