Skip to content
Longterm Wiki
Back

The Vulnerable World Hypothesis

paper

Credibility Rating

4/5
High(4)

High quality. Established institution or organization with editorial oversight and accountability.

Rating inherited from publication venue: Wiley Online Library

A foundational paper in existential risk and global governance literature by Nick Bostrom; frequently cited in AI safety and biosecurity policy discussions as a framework for thinking about how emerging technologies can outpace societal safeguards.

Metadata

Importance: 82/100journal articleprimary source

Summary

Nick Bostrom introduces the Vulnerable World Hypothesis, arguing that continued technological progress may eventually produce 'black ball' technologies that enable catastrophic harm by individuals or small groups, threatening civilizational stability. He argues that avoiding such outcomes may require major changes to global governance, including enhanced surveillance and coordination, posing fundamental trade-offs between security and civil liberties.

Key Points

  • The 'urn of inventions' metaphor: technological progress draws from a set of possible discoveries, some of which ('black balls') could enable civilizational destruction.
  • Vulnerable world types include technologies enabling easy mass casualties, destructive incentives at the state level, and unmanageable global externalities.
  • The 'semi-anarchic default condition'—limited global governance, state sovereignty, and individual freedoms—may be incompatible with surviving certain technological developments.
  • Potential stabilizing interventions include global surveillance, preventive policing, and world government, each carrying serious risks to freedom and autonomy.
  • The hypothesis frames AI and synthetic biology as key near-term examples of potentially destabilizing technological capabilities.

Cached Content Preview

HTTP 200Fetched Apr 9, 20261 KB
# The Vulnerable World Hypothesis
Authors: Nick Bostrom
Journal: Global Policy
Published: 2019-11
DOI: 10.1111/1758-5899.12718
## Abstract

AbstractScientific and technological progress might change people's capabilities or incentives in ways that would destabilize civilization. For example, advances inDIYbiohacking tools might make it easy for anybody with basic training in biology to kill millions; novel military technologies could trigger arms races in which whoever strikes first has a decisive advantage; or some economically advantageous process may be invented that produces disastrous negative global externalities that are hard to regulate. This paper introduces the concept of avulnerable world: roughly, one in which there is some level of technological development at which civilization almost certainly gets devastated by default, i.e. unless it has exited the ‘semi‐anarchic default condition’. Several counterfactual historical and speculative future vulnerabilities are analyzed and arranged into a typology. A general ability to stabilize a vulnerable world would require greatly amplified capacities for preventive policing and global governance. The vulnerable world hypothesis thus offers a new perspective from which to evaluate the risk‐benefit balance of developments towards ubiquitous surveillance or a unipolar world order.
Resource ID: d4d570ab807ed938 | Stable ID: TRjom2SN7A